mending things

Reducing your electricity bills

2011-12-15T06:52:00.001Z

I have been monitoring my electricity consumptions for a number of years on the basis that, in order to reduce it, i need to understand it first, i.e. where does it all go ?

Initially I took a reading from the meter once a week , then subtracted the past reading from the current reading to get the weekly usage. I did this in a spreadsheet. Now I use The imeasure which is a web site run by Oxford University. It does this and many more things, including gas consumption and factoring in weather conditions. I also monitor my real time using a current cost device, monitoring the whole house, plus a couple of individual appliance monitors (IAM) which provide me data about individual appliances. I can publish this data to the Internet via the current cost NetSmart unit. All these methods give me a better idea of what I'm using and where it's going.

I should explain that my household uses gas central heating as the primary source of heat and hot water. My weekly electricity usage averages 20 units a day, lower in summer , higher in winter. That is too high, I want to reduce it, but most importantly I do understand where it is going. The other thing I have come to understand is that there is no magic gadget or solution to save electricity. Many of the energy saving gadgets only affect the devices which already probably use the least electricity .

I find it useful to break down the electrical demands in my house into 3 broad groups in descending order of their. Consumption of electricity:

Things that make things hot.
Things that make things move
Everything else

Things that make things hot means devices that have a primary function to heat, and includes immersion heaters, cookers, kettles, washing machines , dishwashers, hair dryers, microwave ovens. All these units , and others, consume the most amount of electricity in a typica household. Some devices consume a lot but for a short period of time (kettles, microwaves), most are thermostatically controlled , for example an electric oven , once up to temperature regulates the temperature through the thermostat. It does not continuously use large amounts of electricity, rather it's use is regulated . A washing machine heats water internally for part of the cycle, similarly a dish washer does the same.

Things that make things move are devices with motors, pumps or compressors including vacuum cleaners, mowers, fridges, freezers, air conditioning units, sewing machines, blenders etc

Everything else includes, TVs, audio, computers, mobile phones, games consoles.

The biggest savings to be had are by changing the way you use the items in 1. , and once that is done , look at group 2, finally group 3.

Of course many items in the house are in more than one group.

A washing machine , has a heating element (group 1) and a motor to turn the drum (group 2) and a pump (group 2) and some electronics to manage the programmes (group 3). These composite devices go in the highest group they include I.e. group 1 in the example above.

As with all rules there are a couple of caveats :

1) somethings get hot as a by product of something else. For example fridges and computers generate waste heat, but it is not their primary function and I do not include them in 1.

2) conventional filament lighting, while primarily designed to create light, actually creates a great deal more heat. Most of the energy consumed is generating heat rather than light. In my model filament lighting goes in group 1

So how does this help ?

Well when considering any energy saving measure , I look to the groups and use this to decide where to spend the money. With heating items it's a question of considering programmes that use lower temperatures for shorter periods, Eco programmes etc. When considering replacing goods, paying some extra for more control in this area may be worth it. Change filament lighting to low energy CFL or LED equivalents.

Similarly in group 2. Rather than buy a gadget that , say, defeats standby mode, (group 3) It might be better to save that money and use it toward newer more efficient fridge ( group 2 )

Finally measure that save energy in group 3 are only really worth considering when you believe that groups 1 and 2 are as efficient as you can make them .

Mending a Dualit toaster

2011-12-10T08:05:00.001Z

Some while ago we got a 4 slot Dualit toaster. They look great with their 1950's Italian styling but actually are made in East London. They are also substantial and the hope was that, while they are more expensive they would last. Other toasters have not faired so well.

Well they are very well made, but eventually one of the heating elements failed. Toast was getting heated on one side. Time to replace the element.

I went to the Dualit website and ordered a new element. You need to know the model of toaster that you have. You also need to choose the right element. Within a toaster you have single sided heating elements for the two end positions and double sided for the positions between. My toaster has 4 slots meaning that I have 3double sided elements and 2 single sided. It was one of the double sided elements that had failed.

I searched on google and found this brilliant illustrated site By Peter Eland, on how to change the element. The only differences between this and mine are that a) I purchased my replacement from Dualit lit, which came with instructions and b) in my ca se I replaced a double sided element so did not have to remove the end cables.

Pretty easy repair, and they are as well made as they seem. You can also test the old and new elements when out of the toaster ,with a multimeter set to continuity mode. My old element was open circuit, as the heater wire and burnt out. The new element had Continuity I.e I could see a complete circuit, when I applied the two probes to the connection points on the element , when removed !

6watt LED GU10 bulb

2011-09-13T20:56:00.008+01:00

If you have seen my other blogs regarding GU10 & MR16 bulbs , you will know that I now favour LED replacements where possible. It's becoming something of an obsession.

I retain a few halogen bulbs in my kitchen where total brightness and light quality (i.e orange/yellow warmth) are important.

I noticed that the online retailer simplyLED were offering a 6watt(55watt equivalent) LED bulb.

I've purchased from simplyLED before and found their service very good, with a no quibble returns policy. I ordered one to try, which arrived the next day

The flash from the camera is being reflected back by the bulbs rear reflector, so looks like it's on !

As you can see this bulb is longer than a halogen, so suitable for open-backed fittings or where the bulb is only secured at the base. Other types of fittings may not have enough clearance.

In the picture above , the closest light is the 6watt LED , while the other lights are halogens. I would say the new bulb is brighter, hard to gauge from the picture but it's definitely brighter, when seen first-hand.

Again, hard to estimate, but the new bulb is brighter. I would also add that it's a slightly cooler colour, despite this being the "warm white" variant.

At £15 a go, these bulbs are not cheap. But these are the brightest GU10 replacements I have tried , irrespective of technology, so if total low energy, brightness and quality of light are your goal, definitely well worth trying.

They are still not quite warm enough in colour for my kitchen, so make everything look a lot whiter/more clinical, but this is personal taste and I know some people who favour a brighter/whiter light. I've kept the bulb in use away from the central group, and it works well.

I can also see that these bulbs would work well in a bathroom, where brightness might more important and colour less so.

postscript

One thing I did find was that the bulb I tried was causing some FM interference. I have an FM radio in my kitchen and I noticed that I was getting interference on BBC radio 4. It was coming from the GU10 lighting, as when I turned them off, the interference stopped. I'd not noticed this before and so replaced the new 6watt GU10, with an old halogen and the interference disappeared ? I don't think it could be from arcing at contacts, as the bulb would presumably flicker ? the lights and the radio probably share the same ring main in my house, so if this is a comon problem, it may not affect me if the radio is on a different circuit than the lights.

Mending a pair of M-AUDIO BX5a monitors

2011-09-11T09:00:00.019+01:00

One of my sons friends asked me if I could take a look at his speakers which had stopped working. he and my son had been mixing some songs using a pair of M-AUDIO BX5a monitors. These are bi-amped active speakers, with Pro audio style inputs (XLR and 1/4" mono jack). They each have two integral amplifiers so only need a line or balanced input. Each speaker also has a mains power input and a volume control, and a blue LED on the front to indicate power. The speakers have two drivers, a small domed tweeter and a small 5" Kevlar woofer. One had developed a loud buzz, coupled with a very loud 'thump' sound when powered off. The other had completely lost all bass output.

The speakers were only 13 months old, which seemed a very short life for a pair of speakers - I have speakers made in the 60's that are still going strong. However with additional electronics in these BX5a's I'm not really comparing like with like.These speakers are self contained audio systems. All you'd need is a source like an iPod or CD player and you'd have a pretty decent stereo system. They are primarily aimed at pro audio monitoring use and retail for around £200. They are also very well reviewed on the web. I was also intrigued by the name. My own main speakers are a pair of Goodmans BBC LS3/5a speakers - something of a legend among small monitors. I thought the naming of the BX5a was perhaps a nod to these classics.

My first advice was to see if the shop he bought them from would be prepared to fix them as they were not that old. They would only fix for free when the speakers were within warranty, which these were not. Oh well ....I googled the speakers and some of the symptoms and pretty soon found dozens of hits on the buzzing problem, with the loud thump on power off . Again and again people 's BX5a had developed loud buzzes through both tweeter and woofer. It seemed that the speakers had fallen victim to a more widespread problem of poor quality capacitors. This problem has been the scourge of the IT Industry too, I have had a Dell Optiplex fail the same way. I understand that Apple Time machines have had problems with caps leaking/exploding too. Possibly it's due to the caps being a faulty batch or the temperature upper limit being exceeded and prematurely ageing them.

The power supply in the speakers uses a couple of 6800uF 25v caps to smooth the output from the 4 diodes making up the bridge. NOTE : the switched mode supply, ubiquitous in every other aspect would be too electrically noisy for these applications. Here we have good old linear supplies with diodes, and smoothing caps.These two caps appear to have a common problem where they leak. The top of the cap appearing to bulge, and in some cases electrolyte oozing out.
These symptoms certainly seemed like one of these speakers , so I offered to replace the caps . I figured it was within my capability.

I should state at this point, that this is quite a hard repair to carry out. Some forums imply this is easy, but it took me around 90 minutes to do the first speaker, and 60 to do the second. You have to substantially dismantle the speaker and remove two caps which are soldered and glued in position. There are a lot of steps along the way.

I looked on-line at caps and in the end RS Components did suitable caps 6800uF 25v . In addition other important values are temperature (105 degrees) diameter (18mm) and lead pitch (gap between the leads - 7.5mm). The Cap length is not so important but you need to get all the other parameters the same. RS and Farnell have thousands of caps, but by the time you specify all these constraints, you are down to one or two , I went for these Nichicon caps.

RS states that the caps come in packs of 2 , so I ordered a total quantity of 2, assuming this would equate to 2 packs of 2, meaning 4 caps. I had decided that while the second speakers problem was different, it was worth replacing the smoothing caps in both. It might fix the other problem, and I reasoned that it looked like this might become a problem in the second speaker at some point anyway. The package arrived 2 days later but only included 2 caps. I think I had been confused by he quantity Vs the fact that 2 came in a single packet. Anyway I ordered a second batch and set to work on the speaker with the buzz. When dismantling these speakers you will need :

a tray to keep all the screws in
a medium sized Philips screw driver
a No. 3 metric Allen key

WARNING : DISCONNECT THE SPEAKERS FROM THE MAINS BEFORE ANY OF THESE STEPS. THESE SPEAKERS CARRY MAINS VOLTAGES INSIDE SO IF IN DOUBT, GET A QUALIFIED ENGINEER TO CARRY OUT THIS WORK.

I put a towel down on the table and laid the speakers face down. The front does not have a protective grille, but the drivers are recessed enough into the bezel that they don't protrude and laying them on their front is fine . 10 Philips screws hold the back panel in place and I carefully removed these, cupping my spare hand around the Philips screw driver point, so as not to slip and scratch the cabinets. This is more important later when removing the bass driver. It's very easy to stick a screw driver through a speaker cone if you are not careful - I have done this before and it's easy to avoid with care. Once all the crews were removed, the back is loose, but I soon realised that the cables inside connecting with the drivers on the front , were too short for me to gain access to the amp Cables ties had been used to gather up any slack, so I couldn't get the back panels with attached amps clear via the back.

I turned the speaker over, laying it carefully on it's back. the towel is important as the volume knob on the back protrudes, so laying the speaker on something a bit soft will save your tabletop and the speaker volume pot.the front bezel is held on with a No. 3 Allen key headed screws. At first this seems frustrating but actually is a smart idea. It means you can apply pressure in a plane across the speakers, not toward them (as with conventional screws). I still cupped one hand over each driver, while removing the Allen screws with the key in the other hand.

Once they are removed you can lift the bezel. The bass drivers are not attached to it, they are still screwed to the front of the cabinet. However the tweeter is attached to the bezel. You will need to disconnect the speaker wires from the tweeter to free it. The wires are coloured black and white and have identical sized pushed connectors. I photographed the tweeter at this point so I'd remember the tweeter polarity.You can pull the speaker wires from the tweeter but take not of the polarity or better still take a picture. Once the tweeter is disconnected you can take the bezel and tweeter and place safely to one side. You then need to remove the woofer. This is held to the speaker cabinet by 4 Philips screws. Again, cup one hand over the driver while you carefully remove the 4 screws. It' very easy to slip and pierce the driver with a screw driver - I know, I have done it before, and kicked myself for my stupidity.

Once you remove the bass driver you do not need to take note of the polarity as the push connectors are a different size for positive and negative.

Once you have disconnected the woofer, place it safely to one side.

You will find you still cannot remove the backplate as the LED which indicates power is fastened to the front of the speaker. The cable slack has been wound with the excess speaker cable using a cable tie. I carefully cut the cable tie, releasing enough slack cable to allow the amplifier module to be removed from the speaker while leaving the BLUE LED cable in place.

Once the amplifier module was out I could see that the smoothing caps were failing as predicted.

You can clearly see the electrolyte oozing out of one of the smoothing caps.
To replace the caps you will need a de-solder pump , a soldering iron and a craft knife.
However before you can gain access to the back of the board there is a metal plate which has to be removed. You can see the plate in the picture below. It's held on by a couple of screws on the right hand side, which also hold two earthing wires in place. These screws have been secured with green enamel paint or possibly nail varnish, which I had to chip away to access the screw heads. On the left a cable tie through a hole secures the plate.

I removed the screws on the right hand side and cut the cable tie on the left. The plate is further secured to the back of the circuit board with an adhesive strip. With gentle pressure the plate slowly came away. This gave me access to the back of the capacitors. I took a picture to remind me of the polarity of the cap leads (negative -ve and positive +ve). You can see below how bad one of the caps appears. The sticky electrolyte has got some of the speaker acrylic wool stuck to it, making it appear hairy !
I heated the solder pads on the reverse side of the board where the capacitor leads passed through. I used the solder sucker to remove as much solder as possible. I then carefully cut with the sharp craft knife around the base of each cap to cut through the glue which has been used to further secure the cap to the board. This is a slow and delicate operation, hence my comment earlier about this being a difficult fix. You don't want to damage the surrounding components nor flex the board. It's possible to raise the copper tracks from the board if you apply excessive heat or try removing the solder too many times.

I did more de-soldering until I could bend the remaining length of lead up, allowing the cap to be slid out from beneath . Eventually each capacitor slowly loosened and I could gently pull them away. I further de-soldered the remaining holes and trimmed as much yellow glue from where the caps had been. It was then a case of inserting the new caps, checking the leads were correct. (Negative lead has a white strip down that side of the can ), re-soldering them and cutting the spare from the protruding leads.

If you chose the right value of capacitor ( 6800uF, 25v, 105 degrees, 7.5mm lead pitch, 18mm diameter) you should find that they fit perfectly, though might be a little taller, but there is plenty of space.

Once fitted I carefully reversed the process, fitting the plate back and applying a new cable tie through the corner hole. Fitted the rear panel in the box and then the bass speaker, finally the bezel. Always remember to use one hand to protect the drivers - that treble unit has a strong magnet in it and a wondering screw driver will be strongly attracted to the soft dome if you are not careful.

I repeated this entire sequence on the second speaker, thinking that it a) might solve the other problem (loss of bass) and b) is probably a good thing to do. While the capacitors were not in as bad condition, there was a distinct doming of the ends. I also checked the continuity of the two fuses on the amplifier board. Both were fine.

Powering up each speaker revealed no buzzing, the LED still came on and no big 'thump' noise on power off. Placing my ear next to each driver revealed a small amount of background hiss from each treble unit and a tiny amount of background hum from the bass, but this seemed normal to me. Unfortunately I have no interconnect cables to test the speakers with to hand, so returned them to there owner with the advice that he try them ASAP.

Well the good news was that the speaker that had buzzed, now worked perfectly. However the speaker that had no bass .............still had no bass.

I realised at this point that I had omitted to do the single simple test which would have saved me the following rework . I should have tested to bass speakers continuity while I had the driver out i.e tested that there was a circuit between the positive and negative terminals on the back of the disconnected speaker. An open circuit /no circuit would have indicated the speaker had failed internally, as it now appeared it had. Of course it could be other problems with the power amplifier ?

The offending speaker was returned to me. I had also found a cable I could use to test with an RCA plug on one end, and a 1\4" stereo jack at the other. The BX5a needs a mono 1/4" jack , but I knew it should work if I slightly remove the jack so that the internal contact touched either the left of right band on the jack. This was a mono cable, so both were wired to the signal in a single RCA.

I repeated the steps above as far as removing the bass driver, and tested it for continuity. No circuit ! This was good in so far that had the driver blown completely it was a relatively easy procedure to order and drop in a new driver.

I wanted to double confirm the speaker was not functional , so wired it up to one channel of a small desktop Sonic Impact T-amp I had. I turned the volume to quite low and played some music. Without a cross over the sound would be bass heavy, but I figured it should produce some sound.

Nothing, as the continuity test had predicted.

I sat down and started web surfing to see get a cost for a replacement bass driver. The M-AUDIO website was not very helpful here. I could find no links for spare parts.

I stared at the driver and , without much thought double checked the continuity again. My trusty TANDY/Radio Shack digital meter has a simple beep continuity mode, where it beeps if it sees a circuit. I tried the positive and negative terminals again. No change.

Then I tried the wires that lead to the terminals from the centre of the speaker voice coils - short lengths of silver wire.

BINGO !

We had continuity. Further tests showed that the negative terminal of the bass speaker was open circuit. It looked OK, but it appeared that the solder joint was not good. I de-soldered it and resoldered it, and the speaker was fine. I put the second speaker back, after testing it quickly outside the cabinet with my CD player (A Sony Playstation 1), my RCA to stereo 1/4" jack plug and a much loved copy of "The Kinks are the Village Green Preservation Society".

Worked fine so put the speaker back together again. Job done.

A final puzzle.

It's impossible to connect the bass drivers incorrectly as the spade connectors are different sizes for positive and negative, red and black. However the connectors on the tweeter are the same size.

I noticed that the tweeter connections are mirrored. Rather than both tweeters being connected the same way (black left , white right connector), the arrangement is inverted in the second speaker. I'm not sure about this. I don't know too much about this tweeter but would assume, that like the bass driver, both speakers should be wired identically to preserve absolute phase ?

Here is one speaker, with the bezel disconnected, photographed from the top:

Here is the other speaker, with the bezel disconnected, photographed again from the top:

I guess if the treble sounds a bit odd, you might want to check if your treble units are wired in phase or out of phase with each other ? I think they should both be in phase, both with each other and with the bass drivers, though how you tell is unclear as neither tweeter connection tabs were labelled.

CurrentCost IAM device

2011-08-21T08:39:00.005+01:00

Inspired by the success of my bridged energy monitor which enables me to publish to and access data from the web, I decided to also get an IAM. CurrentCost is a household energy monitor, which uses wireless technology to communicate with an inductive 'clamp' installed around the mains electricity cable in my meter box (see other currentCost post).

Recently the bridge enables the display to post data to the web. The IAM is a small unit resembling an electrical plug with a socket on the back. Basically you unplug a piece of equipment in your house, plug it into the IAM and plug the IAM into the wall socket. What it does is enable you to monitor individual appliances in your household using the CurrentCost display, in addition to the whole household. My Display supports up to 9 IAMs in addition to the primary function of displaying the total household consumption.

I've always been a bit mystified as to why my background usage never falls much below 300 watts. Even while on holiday, the house would routinely be using 400-500 watts. No lights or TV were left on , nor heating. I thought the IAM might give me some insight into different appliances behaviour i.e fridge/freezers.

NOTE : Currently the Bridge cannot export IAM data to the web but this is being addressed by the new bridge Mk II renamed netsmart.

CurrentCost Bridge

2011-07-05T21:40:00.007+01:00

I just got a 'bridge' for my current Cost CC128 device enabling me to publish data to dashboard etc. Early days but here is my currentusage

Works well, and meant that I could monitor my household energy use from abroad while on holiday ! The currentcost dashboard at my.currentcost.com has a mobile browser version which worked on my Android phone.

trying to save energy : trying to make sense of the results

2011-04-14T07:39:00.007+01:00

So in my previous blog I outlined how it's easy to log your consumption, how to convert everything to the same units (KwH) and a site - imeasure - which makes this a lot easier.

If you do all of the above , or even some of it, you will see how much energy you use per week. Over time you may see trends, probably connected to the seasons. My household gas usage drops in the summer and rises in the winter - pretty predictable stuff.

I'm going to park oil, as I don't have any experience of using heating oil, but much of what I describe will be similar for gas & oil.

Separating the consumption of gas , in my case for central heating (CH) and hot water (HW), from electricty for a moment, it's worth thinking about what determines how much is used.

for gas, it's used by a boiler. the boiler heats water in a heat exchanger which is pumped through the house, and diverted either through a coil in a how water tank (HW) or through radiators in the house. This is referred in the UK as a fully pumped 3-switch central heating system, and I suspect it's pretty common.

Various controls - thermostats and timers - determine when the boiler comes on and for how long. the better insulated the house and tank are the lower the losses and the less gas that needs to be burn in the boiler. Also the better the controls - the more regulation of the amount of gas used.

In my case the HW circuit comes on twice a day determined by a timer/programmer unit. heat is diverted to the HW tank , determined by the tank stat, a small box strapped to the side of the tank , which touches the copper skin of the tank and switches off when the tank is hot enough. 55 degrees is supposedly the best temperature to set the tank to.

the Boiler has an additional thermostat in it, which always seems to be set at 2-thirds by each engineer who has ever serviced it. This sets how hot the boiler can get assuming it is required to deliver full heat to HW, CH or both.

In the case of CH, it's determined by a) the timer/controller i.e.e is this a period of heating and then b) by a room stat in the hall which determines how hot the house is supposed to get.

I'm not a fan of conventional bi-metallic room stats and favour programmable timer/stats which are much smarter and flexible (see my other blog on the CM67 ).

You can further refine the arrangement with more zonal heating. Commonly this is achieved in the UK with thermostatic radiator valves (TRV's for short) which enable a degree of local control on a room by room basis. the ideas is that if some rooms need to be cooler than the main house stat OR if they have some other form of heat (Sunshine coming in, computers/games consoles pumping out hot air) they regulate the radiators nearby to compensate. this in turn means the system requires less heat and this feeds back to the boiler needing to be on for less time.

Personally have never found TRV's to work that well. They rely on mechanical temperature expansion properties of a solid to open/close the valve, which given that they are attached to the principal heat source ,the radiator, is pretty crude and suffers from hysteresis (over swing).

reducing the amount of gas you use is most easily addressed by :

a) reducing losses . Better insulation for the house and the hot water tank. I fit a second jacket on the tank , even if it has a polyurethane foam one pre-bonded to it.

b) fitting good controls. A tank stat will ensure you only heat HW to the right temperature. A good thermostat will set the house to the right temperature. The timer stats also set the temperature based on previous days data , so start heating earlier to achieve a set temperature at a specific time - an optimise feature.

c) setting the controls. Figure out what temperature you need the house at different times. if no-one is in perhaps during the day, the temperature can be allowed to drop, but raise the temperature for the evening when people are at home. Similarly at night it can be cooler.
most systems implement this crudely with a timer which turns on CH at different times of the day, with no consideration of what temperature the house is or may need to be at a future point. My CH is actually left on at the timer all year, even in summer BUT the programmes in the room stat determine if it comes on i.e if the house is at the right temperature at the appropriate time.

d) ensuring the system is performing to spec i.e have the boiler routinely serviced, bleed the radiators.

So by considering and optimising factors a), b) & c) you should be able to see how much impact this has on gas consumption. This is harder to do than electricity. the boiler is essentially a binary device. It is on or off, it's just a matter of how long for. Electricity is more varied depending on the load. It may take months or even a years worth of data to see what impact measure have had, and external factors like weather , will also affect the results. I'd also mention that many electrical appliances generate heat either directly or as a bi-product of some other function, and cutting back on them may mean the CH has to work harder. This is why I favour converting everything to the same units , to try to see what impact this has across fuels i.e lowering electrical use might increase CH demand.

Trying to save energy: measure first

2011-04-02T07:49:00.007+01:00

A common theme across this blog is reducing my household energy consumption. You will see this in my threads on lighting, computers, gadgets etc.

As I see it there are 3 reasons to do this :

1) to save money - lower bills

2) to reduce my impact on the environment - CO2

3) to reduce my impact on non-renewable energy sources (gas, oil, coal, nuclear - they all have to be mined in some form or other.

While I have Green interests, I also use a lot of energy in my household, through gas & electricity consumption, and I sometimes feel I am the environmental equivalent of a Champagne Socialis t. I have computers, TV's and all the modern trappings which I'm reluctant to give up. My house is part old and part new. It's a long way from a super insulated eco-home, so I have to be pragmatic. I will never get my consumption down to super low levels, but I want to try and reduce it as much as possible.

Measure it

How much energy brings me to the first issue in saving energy which is how to measure it. If you don't measure how much you use, how do you know if you are making things better or worse through changes in behaviour or technology ?

You could look at the amount you pay on your bills, but this is misleading. Bills vary with tarrifs and are often estimated. It's common practice in the UK for energy companies to estimate bills and have a single reading to reconcile actual with predicted. Energy prices vary, mostly up, so you could be using less but paying more than you were.You have to look at consumption rather than cost.

There are a number of ways of measuring your consumption. The simplest and cheapest is probably to read your meters (gas, electricity, oil) once a week. Record the number and subtract it from the following weeks reading to see the weekly consumption. If you divide by 7 you have an approximate daily consumption. You can chart this over time. You could refine it with other factors, for example external temperature - there will probably be a correlation between consumption and colder weather.

You could take this further and record the readings daily. This might be excessive BUT it also might show trends which, in turn, might suggest the main causes . In my house, Sunday and Monday tend to be the days of highest electrical consumption. This is due to heavy use of oven and washing machine on Sunday , and more washing on Monday. Electric ovens and washing machines are both heavy users of electricity.

As you employ strategies for reducing your energy consumption i.e low energy lighting, you should see a decrease. You can do this on paper or build a simple spreadsheet.

It's a a good thing to do. You need to read the meter at the same time each week (or day) to make a fair comparison week by week.

All units being equal

A further refinement is to convert everything to the same units so you can see the total energy consumption across gas, electricity or oil, assuming your household uses a combination. If you are all electricity for heating, hot water and lighting, then you can ignore the next bit - your life is simpler. Electricity is measured in kilowatt hours, gas in cubic meters or cubic feet, oil in gallons or litres

Now you could ask why does this matter ? If I can see the rate of use of each, and see if consumption goes up or down, why does it matter if this is expressed in different units.

A fair point. I think it's easier to understand your consumption as a whole by establishing a total figure. While you may take measures that reduce your consumption of electricity i.e turn down the stat on an electric heater, there is no real saving if this increases the amount of other fuels you consume. Other fuels may be cleaner or cheaper or impact the environment less, but my objective is a total reduction first and foremost.

An example of this might be the replacement of conventional filament 100 watt bulbs with low energy equivalents. Pretty simple stuff , swap a 100 watt bulb for a 12 watt equivalent. I've saved 88 watts or 0.088 units every hour they are on. However filament bulbs are inefficient because much of that 88 watts went out as heat. If my house heating has to work harder because of this, it's not as bigger a saving as I thought. Perhaps a too simplistic example, and we could discuss the relative merits of gas Vs electricity as a fuel and it's relative environmental impact. This has been done elsewhere , my objective is to measure what I use , and converting everything to the same units can only help to clarify.

Gas
While electricity is conveniently measured in Kilowatt Hours at the meter - the standard unit of energy consumption , gas and oil are measured and metered in units of volume. To compare the amount of energy you are using from electricity with the amount from gas or oil, you have to convert the latter into units Kilowatt Hours. This is a bit more complex as a unit of gas or oil may give out different amounts of energy when burnt, depending on it's calorific value. Just like food, gas has a a calorific value i.e the amount of energy given off if burned, and this varies slightly depending on where it came from and differences in the processing.

You can convert the volume units from the gas meter to kWh using fairly simply process :

Take the units used as reported by the meter
Multiply by the metric conversion factor (2.83)
Multiply by the volume conversion factor (1.02264)
Multiply by the calorific value (specified on your bill, something like 39.47333 , but will vary across regions )
Divide by the kilowatt hour conversion factor (3.6)

Oil

For houses that use oil, it's even harder. I have not found a formula, and again , I can see that the calorific value of a type of oil will affect the energy produced. I found this on a web site as a very rough guide

heating Oil is 10 kWh/ltr

A simpler solution

If all of this conversion seems too hard then fear not. There are plenty of resources on the web which can do the conversion for you. I use imeasure , which is a site run by Oxford University. You just type in your meter readings and it does all the conversion, subtraction and charting for you. Once a week I enter my readings into their site. I can also see how my type of house compares with similar properties. I even get an Energy rating, which I'm ashamed to say wobbles between an E and an F - remember I said I was interested, I didn't necessarily say I was good at energy reduction. But I am trying.

The next blog will be on interpreting the measurements

Replacing MR16 Halogen bulbs with LED equivalents

2011-03-13T15:31:00.009Z

If you have read my other blogs about GU10 mains voltage halogen lighting, you will know that I have mostly replaced these bulbs with LED equivalents. they cost more, but last longer, are cooler and use a fraction of the energy. a halogen bulb is typically either 35 or 50 watts. most LED bulbs are 3 watts.

While the bulbs in my house are predominantly GU10 mains halogens, I do have a couple of MR16 bulbs. These are a different type of halogen that operate from a 12volt supply. I should mention at this point that I don't have any dimmer switches in my house. While some mains voltage GU10 LED bulbs are dimmable (always check first) I am informed that it is 12v MR16 dimmable LED's are difficult to make work reliably. My advice would be to replace the dimmer with a normal switch, contrary to popular belief , dimmed lighting is not saving energy.

MR16 bulbs have two pin connectors rather than the lugs found on GU10 bulbs. GU10's use a push and twist method of connecting to the socket. the MR16 simply push in place with metal clips also gripping the edge of the bin base.

They use a separate power supply , either a transformer (old type ) or a switch mode power supply(newer) to convert 250v AC mains to 12vDC. A transformer is a hefty box while most now use a switch Mode Supply, which can be made very cheaply. They are small light box, perhaps the size of two boxes of matches.

In the picture you can see one next to the vent hose in black. these transformers are often tucked away in the ceiling cavity. In my case the extract fan units were accessible , as the extract units vented up through the attic and out through dedicated tile vents via plastic hoses, the type used by tumble dryers

Because of the lower voltages, the MR16 bulbs are the type used in my shower fittings. You can see a common type of extract fan, with a central MR16 light.

Now I had tried LED MR16 equivalent bulbs in these fittings , but they either flashed briefly then went off or flashed repeatedly like a strobe light. Neither was ideal for showering.

The reason these LED 3 watts replacements didn't work is due to the power supply. the original power supply units fitted would only work with a minimum load of 20-30 watts. However the LED bulbs are only 3 watts. I have also read of people with a single power supply running a number of lights. When replacing with LED bulbs they found they had to keep at least one halogen in the circuit to keep the load high, or they would experience the strobing effect.

In my case these MR16 bulbs are fed from dedicated power supply units. the solution was to replace the power supplies with units designed for LED lighting. these can handle loads down to 1 watt. You need to ensure that the LED supply you choose can handle the correct load. In my case it's 3 watts from a single light. if you have a power supply driving multiple bulbs , which is normal too, you need to ensure the power supply has a maximum load sufficient for the total of the wattage of the bulbs i.e if you run 6 LED bulbs at 3 watts each , the power supply needs to have a maximum greater than 18 watts

in this picture you can see an LED driver or power supply, especially for LED MR16 replacement bulbs. Above it are two varieties of LED MR16 replacements, a Phillips Brand cost circa £20 at B&Q with 3 super-strong individual LED's and a Surface mount unit from SimplyLED for £11, who also supplied the driver for £8. Both provide good bright light. the SimplyLED is described as "warm white", meaning it has a slight yellow tint, they also do a "brilliant white" bulb which has a slightly bluish tint. This might be preferred by some, to my eyes fluorescent lights have a similar colour tint.

replacing the exiting transformers is relatively easy however
WARNING
THIS PROCEDURE INVOLVES MAINS ELECTRICITY.DO NOT UNDERTAKE THIS IF YOU ARE UNSURE . IF IN ANY DOUBT CONSULT A QUALIFIED ELECTRICIAN. ALWAYS TURN OFF THE MAINS BEFORE ATTEMPTING THIS WORK. NEVER RELY ON THE WALL SWITCH TO ISOLATE THE CIRCUIT WHILE DOING ANY WORK. ONCE THE MAINS IS TURNED OFF REMOVE THE EXISTING HALOGEN BULB FIRST.

Once you have unscrewed the cover you can see that there are 4 screw terminal connections,
one set of two for the 240v mains input , labelled L for LIVE (Brown) and N for Neutral (Blue)
the other for the 12vDC output. the output is polarised, so the transformer indicates positive (+ve) and negative (-ve) outputs, though neither of the bulbs had any polarity markings and both appeared to work connected either way.

I simply disconnected or cut the wires from the old transformer inputs and outputs, stripped and prepared the wires correctly, and connected to the new transformers. I double checked that no copper wire was visible, and that the screw terminals were not binding on the insulation. I then gently pulled each wire to see that the connections were sound. I then put the covers back on.

I also tried a different type of transformer on another light. This kind I purchased from ScrewFix for £12. exactly the same procedure except that the inputs and outputs were at opposite ends.

once I was sure of all the wiring, I put the new LED bulbs into their sockets before turning the mains back on. Had I left the original high wattage halogens in the light fittings, I might have damaged the new power supplies with the heavy loads. It was then a simple case checking the bulbs illuminated correctly from switch on, which they did

Dyson DC14 beater clutch

2011-03-13T13:29:00.016Z

I've got a Dyson DC14 vacuum cleaner. See my other posting on general servicing.

of late I've noticed that the beater function often does not engage. The beater is the rotating bar of brushes that agitates the carpet. Most upright vacuum cleaners have this feature. on the Dyson it is enabled by rotating a large purple dial on the front left of the suction assembly. The idea is that you disable it for vacuming hard floors, but enable it for carpets. Even turning it forward did not always engage the beater/brush bar.

I decided to take the cleaner apart to see what might be the problem.

First I needed to remove the lower plate. This is held in place by 3 twist fastenings which can be undone with a coin. Once the plate is removed you can see the brush beater bar.

You can see that the beater brush device gets en-wrapped with hair, string and anything that is long and thin.

I routinely run a sharp knife along the edge to cut through all the hair and string to easily remove it. You can see on the right that the rubber drive belt emerges through a plastic housing before passing around the beater. You can also see that I have disconnected the suction hose. this is the white plastic hose which provides suction to the beater enclosure. It screws into place, it's wire bracing spiral providing a simple screw thread.

here you can see the purple beater selector knob. It controls a clutch mechanism. the vacuum cleaners motor drives one belt to a clutch unit, which then drives the beater via a second belt. the selector knob engages or disengages the clutch.

next I removed the beater bar by stretching the rubber band on levering the bar out. You can see the belt still in place.Next I removed the pivoting assembly that holds the beater bar in place. It is held in place
e by two three-quarters circular fittings on either side. It pushes out.It is now possible to also remove the entire front hood structure which is attached to the main body by two circular plastic clips which resemble plastic washers, These lever off quite easily and you can slide off the entire front section. You will also need to lever the drive belt to the clutch from the main motor shaft. this then means you have very easy access to the clutch mechanism. It's held in place by 3 Torx T10 screws and this requires a long necked T10 torx screw driver to access these . One is down a long tube, so a driver with replaceable bits will be too fat. I used a Stanley 67-494 driver. Once out I examined the clutch mechanism. Other than some slight wear to the plastic "click" cam it seemed OK. However both belts seemed extremely shiny on the side which engaged with the clutch. With some fiddling I was able to flip both belts around so that the unworn outside edge was facing inward. I did this by twisting the belt and carefully feeding the twist back through the clutch.

I reversed the above steps to put the cleaner back together. The clutch seems to work again. If I slip the purple knob forward into the carpeted position the beater engages immediately. Before it would need further pressure and would not always start. I think the problem is due to belt wear, so inverting the two belts has fixed that. I'm not sure that these belts are easily replaceable without a completely new clutch assembly. There appears no way to extract them from the clutch unit.

the clutch assembly (pictured)

Follow up January 2012

Before Christmas I fixed up an old NAD 5120 turntable. The rubber mat had become scuffed and rather scruffy. On the suggestion of posts on the excellent vinylengine forum, I used AF Platenclene liquid to restore it. This substance is used to restore the grip on the rollers of laser printers and photocopiers. Not only does it clean the rubber , it also restores much of it's grip. The next time the Dyson clutch starts to slip, I'll give the belt a squirt with this and bit of a wipe. I have yet to try it, but I think it should help

GU10 halogen bulbs revisited

2011-02-17T20:47:00.016Z

Things have moved on since my previous blog regarding GU10 halogen bulbs. To recap, 10 years ago we had our house extended and rewired. We had 40 recessed halogen light fittings installed, the type which runs from 240V main voltage (GU10) not the 12v system (MR16).

While the light is good, and the reliability of the bulbs has improved, they still use a lot of electricity (50 or 35 watt types) and get very hot. In my previous blog I explored the options which were available including first generation LED bulbs and Compact Fluorescent Lights (CFL) in the GU10 size.

That was two years ago , and at that time LED bulbs were not up to the job and CFL's , typically Megaman variety seemed the best alternative. These consumed only 11 watts and give pretty good light from a few minutes after turn on.

I posted a few follow-ups as I tried a few next generation LED bulbs and this post picks up at that point.

In my utility room I had 4 of the Megaman brand CFL GU10 bulbs and these seemed to be taking longer and longer to get achieve full brightness, though initially this had not been a problem. One failed and when I removed it, the bulb was discoloured , with brown marks around the edge and indications of heat stress cracks in the plastic. It also had developed a rattle if I shook it. I took out the three remaining Megaman bulbs, and though they still worked, they too all had the same characteristics of discolouration and cracks. they even all had rattles, though they still worked ! this prompted me to revisit LED bulbs. It also makes me think that the initial promise of CFL long life no longer carries much weight.

I have replaced CFL conventional bulbs increasingly frequently, and I'm guessing that as the prices have fallen , so too has the promised longevity. There is something of a backlash against CFL bulbs , partly due to the high cost of disposal. These bulbs when they fail, should not be thrown away with household rubbish, but recycled at your local tip with other fluorescent bulbs and tubes as they contain small amounts of toxic substances such as mercury.

So I purchased a number of the newer LED alternatives, which still promise long life and low power consumption, and this blog is my findings.

Here is my police line-up of the usual suspects, from left to right we have :

1) a conventional 50 watt halogen bulb, included for reference to size

2) A Megaman CFL GU10 equivalent

3) An LED bulb, purchased at Maplins for around £5. A very early example using a cluster of 24 conventional blue-white bulbs. As per my previous blog, this was very poor in terms of light quality, brightness and reliability. After a year a number of the individual led bulbs had failed.

4) A3 Watt alpha LED GU10 from B&Q, cost around £13. These were mentioned in the previous blog as the first LED alternatives that showed real promise. This uses an array of surface mount LED's behind a semi opaque diffuser. Great light colour though unsuitable as spotlights. immediate full brightness from switch on.

5) A 4 Watt GU10 High Power LED (20 piece SMD 5050, 320 Lumens, 50 watts equiv.) ( QB-1050) (Warm White) @ 11.99 each from SIMPLYLED on the web. These use surface mount LED's which are clearly visible in the picture as yellow blocks. The light is bright and well spread, though a little yellow. I have three of these in a guest bathroom and the light is good, though not quite as good as a halogen.

6) A 3 Watt Interlux™ GU10 Warm White at £10.99 from LEDLIGHTS4LESS on Amazon. these use 3 super bright LED bulbs through a diffuser. the construction is similar to 5) in that the body of the bulb is metal with cooling slots. These are closer to a spot light with good colour and come very close to a 35 watt Halogen. I have 17 of these in bathrooms and kids bedroom. These , while not quite as good as a halogen, give a nice bright instant light source and are my favourite LED GU10 bulb replacement.

I do have a couple of the 12v bulbs (MR16) in shower extract fan fittings with integral lights. While I tried MR16 LED bulbs from both SIMPLYLED and LEDLIGHTS4LESS, they flickered very noticeably. I believe this is due the the 'ballast' i.e the transformer that drops 240v down to 12v. These expect to drive a certain load , typically 20-50 watts. Unfortunately the 12v MR16 LED bulb is 3 watts, and so always check first to see if your MR16 lighting uses ballast power supplies capable of driving a few watts. I will add at this point that both SIMPLYLED and LEDLIGHTS4LESS replaced the MR16 bulbs with extra GU10's with no fuss and both gave exceptional service over the internet.

My conclusion is that while LED bulbs are still quite expensive (circa £10 per bulb) , they are capable of providing good light quality, with instant full brightness, while only using a few watts and staying cooler than a 50 watt Halogen. They are not quite as good as Halogen and I retain some in my kitchen work areas and above some cupboards in my bedroom where the premium light quality of Halogen is preferred,.

Pallets & Compost bins

2010-04-01T07:35:00.004+01:00

Pallets are all around. I see them as I drive about. In yards, loading bays and car parks.

I have collected some, with the approval of the site owners, nay, with the enthusiastic approval of the site owners.

There is a lot you can do with pallets. You can build compost enclosures. Here is one from my garden - 6 pallets and 12 cable ties and you have a really good compost bin.

You can cut up pallets for firewood. Breaking them apart is hard, a circular saw is the best way I have found. Only use untreated or unpainted pallets for firewood as burning the other sort will give off various toxic substances . Pallets range in size and construction but most have rectangular blocks in the corners , middle and centre positions to take the weight. These burn especially well, though on some they are made from compressed wood chips and unsuitable for burning.

the pallet plank wood is soft wood and burns very well but rather quickly. Sometimes it's worth taking 4 or 5 pieces and nailing them into a single log size piece. One nail is all you need, but it can slow down the rate of burn in a stove.
These and any other nails present can be carefully fished out of the ash at the end with a magnet

Compost bins

You can make composters out of all sorts of things. Old leaky water butts are good. I just cut the bottom section off , along the raised line.

Another Dalek here (see other blog entry on Daleks), incorporating the drum of an old washing machine as the torso. Makes a great incinerator for the rare times I need to burn diseased wood in the garden.

A collection of different composters, some bought and some made from old water butts on freecycle

failing to repair an ipod nano 4th gen

2010-03-28T15:46:00.015+01:00

So sometimes I fail to fix things, and it is only fair to document those occasions too.

I am very fond of the stories of Sherlock Holmes. In the short story "The Adventure of the Yellow Face", set at Norbury, Staffordshire. Holmes gets the case completely wrong. Later Holmes says:

"Watson, if it should ever strike you that I am getting a little overconfident in my powers, or giving less pains to a case than it deserves, kindly whisper 'Norbury' in my ear, and I shall be infinitely obliged to you."

For me, that whisper goes 'nano'

A few weeks ago I opened the washing machine and out tumbled my sons iPod nano, with the recently washed clothes. It had been in his shirt pocket and not been noticed.

So.... consulting the web the general consensus was that slow drying stood the best chance of resurrecting the device. Some advocated keeping the iPod in a bag of dry rice. I just placed it inside our airing cupboard and left it for a week.
I then plugged it into my sons PC and after a short while the display lit up to indicate it was charging. I was then able to sync it and even update the firmware. I breathed a sigh of relief as the iPod appeared to be none the worse for it's 30 degree washing cycle adventure. Even the headphones worked.
However after a couple of weeks my son asked if I could take a look as , while it still functioned, there was no back light . No amount of playing with brightness settings worked, even a reset still had a screen that was next to impossible to read. The iPod still worked, rather like an iPod shuffle, you had to navigate 'blind' as it were.
Another consultation with the web, and the consensus was that the display was busted but might be replaceable. Given the environment the iPod had been subjected to , I could see that the display might have failed, the back light is part of the display unit.
Looking on ebay, a new display module costs around £10, which seemed a good idea as with no back light the device was next to useless. Hindsight is a wonderful thing and should probably have stopped at this point. Take note.
I ordered a new display and it duly came. Good service from my seller.

I consulted youtube where a number of videos are posted showing how easy it seemed to dismantle an iPod nano. the new display came with a special tool and after all how hard could it be ?

The two videos that best seemed to explain this were this one:

I rather liked this one. I especially liked the way you can hear the Tube trains going by ,the station announcements in the background and the familiar London accent. This chap favours sliding the whole main board out the top , having first disconnected the headphone jack assembly.

And this one:

This is an American video, again, looks reasonably simple. This presenter advocates popping out the glass and accessing the display connections through the empty display window.

With this amount of advice, what could possibly go wrong ?

My experience

First off, I was able to remove the white end pieces which are glued on with a gel-like glue. The included green plastic tool worked fine for getting the ends off. So far so good

Now you realise how small the Philips screws are, and I needed to go out and get a set of tiny Philips screw drivers. PH00 is the correct size. You can see here two screws next to a UK 20 pence coin.
Having prized off the plastic end pieces , there was still glue clogging the screw heads. I used a wooden cocktail stick to pick out all the glue to expose the screw heads. I was able to remove the screws from the base of the iPod - three in total as per video 1.
I suspect that the first Video, while correct, was carried out on an iPod that had already been prepared "Blue Peter" style. It took a lot longer to get all the glue out and unscrew the Philips screws from the base.
At this point I hit my first real problem - I could not get the white jack socket assembly off. If I tugged it too hard , small pieces of the white plastic snapped off - not good. If I pulled the whole assembly down, the main board with the Dock connector slid down too. This is connected to the iPod shell via a tiny ribbon cable to the jog wheel. If that cable snaps it's certainly not going to work. I slid the jack housing back home and pondered video number 2. He takes a different approach and leaves the base, doing everything via the top.
So , I removed the top plastic and undid the screws and removed the metal top piece. This exposes a tiny piece of metal with the hold select button on it. gently lift this off and store safely. It forms the mechanical linkage between the button you slide for hold mode and the tiny switch connected to the back/top of the display. I was able to get access to the top of the display and slide the whole thing up around 5mm.
As per the video I carefully prised off the switch which is connected to the back of the display with glue and to the case with a tiny ribbon cable. Letting it dangle by the ribbon - very delicately, I gently slid the whole display up a further 10mm . IN the glass window of the iPod I could see the lower part of the display board and the ribbon cable.

So far so good .......In the second video he explains at this point that you can 'pop' out the glass screen by applying pressure at the bottom. I did this and it didn't pop, but inserting a fine blade along the top rubber seal and then trying worked. Out came the display glass window, and I put this to one side.
Next to disconnect the ribbon cable which connects the display to the mother board. This is situated in the bottom left corner. The whole lower part of the display was covered in a piece of transparent yellow tinitd tape, which I removed. I then prised the tiny black locking bar above the ribbon. This bar is normally set pointing upward and you lever it down, perpendicular to the plane of the display, with a tiny screw driver.This done I should (according to the video) have been able to slide out the display completely, but the ribbon was stuck to a silver metal tab on the lower left hand corner of the display. I had to gently insert a screw driver head beneath it and very gently prise it off the tab whilst pulling the display out.

The display was free and I removed it from the iPod case.Here is what it looked like at this point, you can see the tab at the top and the ribbon connector socket and tab just to the right as a black bar:
I inserted the new display and slid it back in until the ribbon cable was able to engage in the slot. This is very fiddly indeed. You also have to simultaneously keep an eye on the hold switch which is dangling via a tiny ribbon cable at the back, and easily broken off as you slide the display in.

Eventually I got the ribbon cable connected to the display and I snapped the black bar forward to engage. I decided to perform a quick test at this point and slid the hold switch across and pressed the Menu button. The ipod came on ....

.....but the back light was still off !

I double checked that I'd fitted the new display , I had. This was most odd.

I repeated the removal and re-installation of what was confirmed to be the new display. No change. The iPod still worked but the back light was off.

I tried cleaning the edge of the ribbon connector with a cotton bud and some alcohol, as I thought it was darkened at the edge. perhaps some arcing had occurred after the drying out ?

No change.

Things went from bad to worse as two things occurred almost simultaneously :

1) I tried to reinsert the glass window and now amount of pressure would pop it back in. Indeed too much pressure will warp in the top metal edge of the iPod. I tried bending it gently back with mixed results

2) The display stopped showing anything.

Annoyed with myself I was confronted with the fact that the nano was dead.

Frustrating indeed as I'd spent several hours and treated it it with supreme delicacy

At this point I did the only thing possible under the circumstances. Had a cup of tea, put all the nano bits in a padded envelope, and ordered a new nano from dabs . 5th Gen this time with video camera and FM radio, slightly darker green shade too...

Looking back perhaps I was lucky that the nano survived the washing machine as it did. Perhaps this had shortened it's life and my fumblings finished it off altogether.

I guess the old nano and the 2 displays can go on ebay. I suspect the displays are both fine, the problem lies elsewhere. I think the hold switch probably was flexed one too many times and the internal ribbon broke, meaning the nano was trapped in whatever mode I'd left.

So this device was most definitely not mended

A postscript

the broken nano went for £12 on ebay. Not bad, I think the buyer got a bargain as almost certainly it included 2 working displays. The new nano is better than the old, it has a video camera and a tiny speaker and a few extra features, so if you are going to have to replace something, it's always slightly better if there is some improvement with the new one.

It happened again !

Whilst on holiday, my sons new iPod 5th gen nano got washed in his jeans again. We dried it out over a week and then attempted to charge it. It appears to be working fine, so far. Of course this happened with the 4th gen for a while until the back-light failed. Anyway, as we say in the UK, "touch wood" for luck, it's still working. So if you submerge or wash your iPod in the washing machine the best thing to do is leave it in a warm dry place for a week and then re-charge.

A slightly greener Dell

2009-09-03T06:44:00.013+01:00

Well the trusty old Dell Dimension 2350 is still going strong. I decided to replace the power supply with a more efficient one. If you've read my earlier Dell post you will know that the original Dell PSU started to fail some years ago and was replaced with a gold Sweex Brand 400 watt supply, which has worked well. I decided to replace the Sweex with a 400Watt AKASA AK-P040FG8 which is 80plus rated . This means it is more energy efficient. The Sweex was fine, but I didn't really do much research when I bought it, and I had another project in need of the Sweex , so it seemed like a good idea to upgrade the one in the Dell to a more energy efficient design. My Son had also seen the occasional lock up of the PC while running games and iTunes simultaneously, and I thought this might be due to PSU demand.

However before replacing it I wanted to do some tests of standing and peak power consumption of the PC with both the Sweex and Akasa power supplies. I plugged the PC into an energy monitor plug which displays the power consumption instantaneously. In the picture you can see the display showing the PC consuming 118 watts of power.

With each power supply I booted the PC, logged into XP and let the PC settle to a steady state, after all the AV and start up scripts had finished. I then ran sample loads of a full AV scan (heavy disk and CPU use) and then later the GPU benchmark 3dmark03 (heavy GPU use) and watched the peak consumption.

With the Sweex the steady state was around 94 watts rising to a peak of 147 during the AV disk cycle. 3dmark03 peaked at around 144 watts.

With the Akasa the steady state is around 78 watts rising to 127 under similar load. Not very scientific, but I wanted to see if the claims of the Akasa were justified. It certainly appears to use less electricity.

the energy cost savings, while worthwhile, would take a long time to repay, but my plan is that if the dell fails, and one day it will, the PSU can be used in other projects.

Fitting the PSU

Fitting a new PSU in the Dell Dimension is pretty straightforward job.

First you need to disconnect the existing supply from the mains, then remove the side panel.

Remove all the existing power connectors, including the MOLEX white connectors to each disk a, floppy (I don't have one) and CD/DVD drives. The standard MOLEX connectors are white nylon and always seem to hurt my fingers removing them. Then disconnect the existing power supply from the motherboard. There are 2 connectors, a large block and a 4 pin connector for the CPU.

Once all cables are disconnected you can remove the 5 standard PC screws which hold the PSU to the bak plate of the PC. Once free, it's a bit of a fiddle to get the old PSU out and I had to temporarily remove the top CD/DVD combo drive to allow some room.

Once removed it's really doing the whole thing in reverse.

one issue with newer power supplies is the number and range of cables they support. the akasa is not a modular supply, which would mean I could disconnect unnecessary cables at the power supply end. Modern supplies have SATA and PCI-express cables, neither of which are used by the Dell.

As you can see the Akasa cables are neatly sheathed in black nylon, and after a bit of 'dressing' I was able to tie up the unused cables and allow as much air flow space as possible. removing the floppy drive and bracket provides a bit more space. Another nice feature of the Akasa, is the molex connectors for the IDE drives are much nicer and easier to plug in. No pain with these.

Once fired up it all worked first time, which was a relief, as my son was away on a camping holiday, and I didn't want him to return to a broken PC.

Cooling

I did notice that the primary disk drive (located lower front) was reporting slighly increased temperature via speedfan, during heavy load (AV). It was peaking at around 36 degrees which is higher than average and higher than I would like. This may be due to the Akasa fan running slower and hence drawing less hot air from the case, or due to the extra cabling restricting air flow. I'm experimenting with blocking the vaccant floppy case hole with Duck tape, also the hole vents by the PCI cards. Some caution is needed here, but my thinking is that I want cold air to be drawn in from the lower front of the dell, through the front vent holes, over the primary drive (cooling it) and then evacuated vis the PCI blower, case fan or PSU fan. The hole left by the removal of the floppy and the vents by the PCI cards go someway to compromising the venturi effect . I'll keep an eye on this and see if it makes an appreciable difference

mending loudspeakers

2009-07-12T10:08:00.015+01:00

Having been interested in music and audio equipment since the late 70's, I hate to see old speakers go to waste, especially when they classic British vintage designs like Wharfedale, Leak, ,Mission,Celestion etc.

I subscribe to freecycle and sometimes I see old speakers come up on there, usually because they are damaged in some way. I'm no expert on this, but have managed to cheaply get some pairs back to working and presentable state. It's actually not that hard. If certain collectable brands ever came up (i.e Tannoy, Lowther) I know enough not to attempt this kind of "get you home" repair and would consult the experts. The speakers described here are high quality models from the 70's/80's that were produced in large numbers, and while very good, they are not that valuable.

The Leaks

The first pair I got from freecycle were a pair of 70's vintage Leak 2020 sandwich speakers.
The term sandwich refers to the technology that Harold Leak used to make very light but stiff cone material. The speaker cone is a sandwich of expanded polystyrene (Styrofoam in American) between sheets of adhesive aluminium tape. This gave a strong , rigid yet light material. in the original advertising Harold was show standing on an inverted cone, to demonstrate how strong they were.

When I got these speakers they were in poor state. One of the drivers had a hole in the polystyrene, probably where a screwdriver had pierces it while someone was trying to remove it.
One of the ceramic tweeter soft domes had a tear, the cabinets had been used as plant stands and had substantial water stains and one of the grilles had a smiley face painted on it , still just about visible above.

I stripped the speakers back to basics, carefuly removing the grilles, drivers and rear speaker connector block. the signature Leak aluminium frame around the grille was removed. When removing drivers I always use a good quality philips screwdriver which I use my free hand to grip around the tip of he screw driver and the screw holding it in. This is to avoid the screw driver slipping and going through the speaker. I have done that before ! usually the bass/midrange unit is held on with 4 wood screws around the frame. I try to remember which screw went where. Internally most speakers use push connections to wire them to the crossover and rear connections. Always make a note of which coloured wire went where.

The cabinets of most speakers are made from veneered chipboard, though in older or more expensive speakers other dense forms of wood are used, such as marine ply wood. I decided that since these speakers were headed for my kids games console ("Guitar hero") , I needed a quick and easy fix to the badly scratched and stained cabinets. I had a tub of dark Jacobean briwax, a thick and gloopy wax used for restoring old furniture. You have to use it in a well ventilated place i.e outside, as the wax is held in toluene which is fairly nasty in an enclosed space.
Having wiped the cabinets down with a damp cloth, and left to dry, I applied a thick layer of briwax using a plastic pan scourer. Drastic measures, but these cabinets were in a bad way. I left the wax to dry a bit then polished them and they came up very well. The dark wax fills all the light scratches, so the cabinets look old rather than just tatty. 'Stressed' in other words.

Next I repaired the drivers. The bass driver with the hole just needed a little PVA glue (Evostick) to plug the hole. The schoolboy error I avoided was using polystyrene cement to repair a polystyrene speaker. This would have been a disaster as polystyrene cement is intended to glue the rigid form of the dense plastic form of the material, typically used for construction kits and yoghurt pots. It actually melts the expanded form, and achieves an effect somewhat like the yellow blood from the movie 'Alien' when it burns through the deck.

I also used a little PVA to repair the tear in the tweeter cloth cover. the tweeter itself was intact. Possibly I could have replaced the cloth completely, but these are not the highly prized Alnico tweeters found in some models, just the ceramic ones. I wanted this project to be zero cost.

All the fittings ( rear connector plate, connectors, screws ) were cleaned up, and after several attempts to get the aluminium grille frame clean with various chemicals (it was pretty dirty) ,eventually I found it came up like new after a session in the dishwasher.

This just left the grilles. the material was a plastic type, stretched over a hardboard frame. The hardboard would warp if too much wet was applied. I tried several good goes with 1001 carpet shampoo, and this mostly got rid of the face. The flash in the photo above still shows it but , it's hard to see in normal light.

Put them all back together and they sound great. The tweeter is the weak spot with these speakers and can sound a bit gritty, but the bass is very impressive. the styling is very 70's , which you either love of hate, but I'm rather fond of them and they served well as TV speakers in my kids den, also pumping out rock music for 'Guitar Hero'. This was until ...

The Missions

A pair of Mission 707 speakers came up on freecycle , described accurately as needing the foam replacing. I don't have a picture of how they looked, but indeed the rubber 'foam' that surrounds the cones had completely perished. Google revealed it's a common problem with this vintage of Mission drivers , and other speakers from that time. The Leaks , which were older had no indication of 'foam rot' , nor do an even older set of Goodmans Goodwoods, which also came from freecycle and required no work at all.

Otherwise the speakers were just rather dusty from having been in storage.

Same approach as with the Leaks, strip them down, carefully removing the drivers (bass and treble units) taking note of the wiring. the Missions were budget speakers from the early 80's as I recall, and had and have quite a following. The unusual features were the inverted arrangement of tweeter below bass, and the heavy plastic front baffle. This comes off, and I gave them a good wash in warm soapy water, taking note not to scrub off the Mission logo.

The cabinets had some scratches and were a bit grubby. Wipes down with a wet cloth and briwaxe as with the Leaks. I cleaned all the screws and rear connectors and plastic rear plate.

The Mission Grilles use a finer material stretched over a plastic frame. 1001 carpet shampoo didn't really bring these up as well as I hoped, so I ended up washing them in a sink of soapy water a couple of times. Fortunately the grille material retained it's tightness and the plastic frames didn't warp as the plywood ones would.

I put the speakers back together except for the damaged woofers.

A number of people on the internet/ebay sell re-foaming kits, and I had been iterested to try this for some time. Essentially the kit consists of a new set of rubber seals for the speakers , plus rubber glue. you remove all of the perished rubber seal, get them as clean as possible and re-glue the new seals. The complicated bit is getting the speakers centred.

If I pushed the free standing cones in, I could feel the voice-coil rubbing against the inside of the speaker. This was a bit troubling as I wasn't absolutely sure the speakers worked. If driven hard with broken foam surrounds the coil might have been pushed too far forward or burnt out.

I got a set of foam rings for he 707 from good_hifi in Holland, who sell on ebay. they cost around £17, but I figured these speakers were worth it. With broken drivers they had no value, with working frivers they probably go for around £30-40 on ebay.

The foam Rings took about a week to arrive. I was a bit nervous about this as I'd never done it before but the directions were straightforward on the site, especially this video. The centering approach using clothes-pegs worked a dream, and while the speakers seemed initially to rub in all directions, once I'd glued the surround to the cone and centred it on one axis, it worked well. I was then able to centre on the other axis as described in the video. The rubber glue dries quite slowly , so you can fine tune the position. Anyway the picture below is the finished results.

The missions are 10 years younger than the Leaks and it shows. They have replaced them for 'Guitar hero' by popular demand, but I intend keeping the Leaks as they are classics and the bass is very good. Who knows, with the arrival of Rock Band 'Beatles' on 09.09.09, the leaks might be better suited to the mid-sixties sound

DIY speakers

I have sometimes made speakers. I read some articles about an interesting full range driver from Japan called the Tangband W3-871S , which can be purchased reasonably cheaply. It works well in the Cyburg Needle design.

I was pondering making a pair of needles when I noticed some ethnic vases in a local shop called 'snip'. the vases were wooden spheres, and I wondered about making them into spherical speaker cabinets. After some fiddling around , they turned out quite well and work as good Computer monitor speakers.

the curse of the GU10 halogen bulb

2009-07-11T08:43:00.013+01:00

I know more than I want to about these light bulbs. While the UK Government is waging war on the 100 watt incandescent light bulb, they have, typically, ignored the growing trend of using these bulbs. I guess they pay people to change bulbs for them.

We had our house gutted and rebuilt in 2000, and as a consequence had around 40 recessed lights fitted, each of which take the GU10 mains voltage (240v) GU10 light bulbs.

At first, the main issue was that the halogen bulbs had a very short life span and cost a lot to buy, around £5. I seemed to be changing a couple a week. Things have improved and now the bulbs are relatively cheap ( £1-£2) and have a fairly long life expectancy.

So what is my gripe ? Well these bulbs are rated at a hefty 35 or 50 watts. The packaging even implies that they are energy saving in some way, but of course watts are watts. i
If you have 5 * 50 watts halogen bulbs in your kitchen, then you are using a unit every 4 hours they are on. If I turn on all the halogens (assuming the bulbs are 50 watts) then I'm using 2Kw. that's the same as my washing machine

So my point is that 2 * 50 watts GU10 bulbs = 1 * 100 watts bulb .

The other thing to note about these bulbs is they get very hot. The recessed light fittings are supposed to have a heat cowl fitted above them if in open loft space to prevent insulation or other material touching the back of the bulbs which get quite hot.

While the light from these bulbs is good quality I was unhappy with the energy consumption and nervous that a couple of installations might create a fire risk above. I looked into low energy versions and here is what I found:

The bulbs above are all GU10-fittings, but use a range of technologies. You can also see that while they share the same diameter and lugs at the top, they vary in length.

These are from the left :

11watt Compact Florescent (CFL) Megaman brand cost £8
7 watt Compact Florescent (CFL) Farnell brand £8
1.5 watt LED unknown brand £9
50 watt halogen Sylvania brand £2

So the three technologies for GU10 bulbs are LED, CFL and Halogen.

The Megaman I got from B&Q and it's the best of the CFLs and probably the best alternative. gives reasonable light instantly. In common with all CFLs the light quality is a bit bleached or cold, but they work well. The only issue is that they are physically much longer, and depending on the light fitting , they may not fit. Most of my recessed fittings are backless. the bulb is gripped by the rim and the electricity is connected by a loose cable connector at the back. This means the depth of the bulb can be accommodated. You may find that your fittings will not accept a bulb this deep, so it's important to check before buying.
The Farnell bulb is closer to the form-factor of a standard GU10 bulb and can be used where the depth problem described above is an issue. Even so the dimensions are not *exactly* the same around the 'neck'. These do not give good light from switch on, and need a few minutes to warm up, initially giving a pinkish light. However once warm they work well, and I have a set of these in my utility room.
I had high hopes for LED technology as this appears to provide the cheapest running costs. However the light quality is very poor. A dim cold bluish light which is not good for spotlights. I also noted that some of the individual LEDS' have failed over a 2 year period, so reliability is not great. I have high hopes for LED lighting, and I have seen some bulbs on the web which use a single high power LED rather than an array of 20 smaller ones. However these are currently around £23 per bulb, too much to be practical
Finally you see a halogen bulb for reference. excellent light quality, but heavy power consumption and they get very hot.

So in conclusion the Megaman variety wins as the best allround replacement if you want to use less energy and generate less CO2. the Farnell offers a viable alternative where space is an issue. LED bulbs have yet to be a viable alternative. I noticed that all the display units in B&Q use LED bulbs. This is , I assume, to reduce their energy costs, but it also provides a way to see the type of light they provide.

update 9/01/10

yesterday I visited the local B&Q, and noticed they had a number of LED bulbs. They had the standard ones as I'd used and dismissed above, but they also had a much more expensive 'alpha' range. Two types, one was a 20watt equivalent for around £16 ! This appeared to have a single large LED at the centre, and the light was described as brilliant white.

the second was a 10 watt equivalent which uses 24 SMD white (Surface Mount Device). the dome is difuse so it's hard to see what the LED's look like but I can make out that they are yellow. I have seen pictures on the web of a new type of LED bulb which uses rectangular yello LED's , not the tiny spherical ones , found in the cheaper LED bulbs.This was around £12 and the light was described as soft white.

Because I now appear to have an obsession with finding an energy efficient GU10 bulb, I got the one of the second type to try. £12 is a high price for a bulb, but in the persuit of science ....

Anyway I used it to replace a standard GU10 that had failed in my bathroom, and the light quality is excellent. Better than either brand of CFL and a million miles from the previous LED bulb. Light quality is warmer , like a regular GU10 halogen. Perhaps it's a little diffuse as it does not have a single light source, but very good. It only consumes 2.8 watts around 1/20th of a 50watt GU10.

Update 10/07/2010

I recently aquired a number of another type of LED GU10 bulb. These have 3 high powered LED's and are suited as spotlights. They use only 2.8watts and cost around £12 each. These work very well where the earlier Alpha type are too diffuse.

Economy 7

2009-05-30T09:06:00.008+01:00

I am on an Economy 7 electricity tariff. E7 , as it's abbreviated, offers lower cost electricity at night, but with slightly higher day rate and standing charge.

The original reason for E7 was so the Electricity providers to reduce the peak demand. In very general terms, the electricity we are using now is being generated now. This means that the grid has to anticipate peak demand, fro example the ad break in the middle of "Coronation Street" , when millions of viewers get up and switch electric kettles on.

By encouraging people to use electricity during off-peak times i.e at night, less generators have to be bought online. Also the mix of generators determines the amount of CO2.

For live CO2 data from the grid see

http://realtimecarbon.org/

You can see that the mount of CO2 released in grams per unit is at it's lowest at night.
Also lowering peak demand means less power stations needing to be built

The biggest household use of electricity is heating water and space heating. E7 is typically used in conjunction with a hot water tank with immersion heater and night storage space heaters. Night storage heaters absorb heat during the E7 period, using their high thermal mass. They then slowly release this heat during the day. Similarly hot water is heated in an insulated tank during the E7 period, which is then used throughout the day.

A house with E7 heating and water , typically has a separate meter and ring main for the E7 "heavy use" devices. This circuit is switched on and off by the meter at specific times. A good description of this whole idea is here. E7 tariffs include all household electricity, so this means that any item in use during the E7 period (23:30-06:30 GMT, 00:30-07:30 BST where I live ) are charged at the cheaper rate, and any thing in use during he day is charged at the higher rate.

My Mum has this form of heating and she likes it, though it does have it's critics. It's not instant heat, you have to almost plan in advance , how many heaters you need to be on and how hot you want them to get. Each heater has 2 control settings 1) A thermostat which determines how hot each heater gets, hence how much heat it stores and give out 2) A mechanical flap which enables heat to be released at different rates. For example, in the evening you may want more heat to be released.

Now at this point I should say that I don't have E7 heating or hot water. However when we moved into our house in 1991 we had an E7 meter installed , which at that time was free. It may still be , you would have to contact your electricity supplier. We have subsequently purchased white goods which could exploit the E7 cheaper rate. Most German washing machines, dishwashers, tumble dryers seem to have models which have a delay timers, and we tend to use these.

( I think a number of European countries have similar systems, the French have Red Tarrif where electricty is half price all year except 20 nominated days when it's 10 times the price. These days are the coldest of the year, though never a weekend or holiday, and you get advanced warning the day before via a light on the meter. It's a good system as it's entirely optional and means for those 20 days you just have to think a bit about what you use. My mother-in-law has this system and likes it, but she does have a huge wood burning stove, and a couple of UK Storage heaters she can turn on the day before )

In addition a house has a background level of electrical consumption. (see my blog on the CC128 electricity usage meter). Better still during summer, a good chunk of the period extends into my waking day, with the cheap period ending at 7:30. My Mum in Kent, says her cheap rate ends at 08:30 in the morning. Plenty of kettles boiled, toasters run , power showers used , during off peak then.

So, why doesn't everyone just get E7 installed ? Well here is where it gets really complicated, and is the reason for this blog.

As I said at the start, when you are on E7 you pay cheaper night rate per unit (5.159p - Scottish Power) but slightly higher day rate (11.229p Scottish Power) and a slightly higher daily standing charge (15.88p per day with E7 versus 12.88p per day normal tariff. These rates were true at time of writing, and for Scottish Power with my tariff. Different suppliers have different rates. Some have no standing charge, others further subdivide the day by charging different rates for the first n units used. It soon becomes rather complicated.

I decided to try and figure out, for my supplier, whether E7 was actually better for me i.e. for a given rate of use, was E7 more or less expensive than Scottish Powers standard rate.

The maths gets complicated, and while I'm sure there is a formula to calculate this, it's beyond my waning mathematical skills. I resorted to a spreadsheet to model this, please drop me a line via the blog if you want a copy.

It's a bit hard to include a spreadsheet in a blog but here is how it works :

1) You have to gather from your provider the rates i.e the E7 night & day rates, the standard rate and the standing charges for both. My model doesn't work where you have further partitioning of the rates i.e the first 500 units at a certain rate, though it could be adapted.

2) You need to have an approximate idea of how many units you use per day, either by monitoring your meter or going through actual readings. For me it's around 22.

3) the bit you can play with is is the percentage of night used per 24 hour period. In my spreadsheet I can alter the percentage up or down until the E7 total cost is lower than the standard rate. In effect I'm using this percentage value as a slider to see where E7 is cheaper.

Here it is in a simple text form:

Number of days      365

E7 Night rate 0.05159
E7 Day rate 0.11229
E7 Standing charge 0.1588 per day

Standard rate 0.10827
Standard standing charge 0.1288 per day

Average units per day 22
Total Units 8030

Number of day units 7317.739
Number of night units 712.261
percentage night/day 8.87

Total E7 Cost £916.42

Total Standard £916.42

Having entered all the variables (tariff rates ,number of days and the estimated daily use), I just tested different percentages. I found the tipping point FOR MY TARIFF was 8.87% i.e. I needed to use >= 8.87% of my total electricity during the E7 period for E7 to be cheaper.

I was quite pleased with this, as 8.87 % (OK lets call it 10%) is not that much. Assuming 22 units per day, that's 2 units during a 7 hour night time period. Possibly fridges and freezers plus some background computers and sky plus device might account for that alone. I am actually managing between 25-30% during the E7 cheap period, as I'm able to run washing and dishwasher during this period.

Now because of the complexity of the calculation and the various tariffs, the percentage for E7 needed to be used for another persons circumstances may be higher. I seen some tariffs where , dropping the numbers into the spreadsheet, you have to use as much as 38% during the night period in order to make E7 pay. That would clearly require use of night storage heaters and hot water. It's a very complicated area, but I wanted to determine if the widely held belief of " if it's less than 30% use it's not worth it" was true. The answer is more complex. My actual consumption is nearer 25% so I'm comfortably clear.

As my American friends would say :

"Your mileage may vary"

Formula Explanation

Number of days 365 ( a year)

E7 Night rate 0.05159 ( from my supplier website )
E7 day rate 0.11229 ( from my supplier website )
E7 standing charge 0.1588 ( from my supplier website )

standard rate 0.10827 ( from my supplier website )
standard standing charge 0.1288 ( from my supplier website )

Average units per day 22 ( based on previous bills, consumption history or just watching the meter )

Total Units #VALUE! =C1*C10 (Number of days * Average units per day)

Number of day units #VALUE! =C11-C14 (Number of Total Units - Number of Night Units)

Number of night units #VALUE! =(C15/100*C11) (Percentage of units used during night period * Total units )

percentage night/day 8.87 (Enter example percentage of night to total units - experiment with different figures to see which is cheaper E7 or standard rate)

Total E7 Cost #VALUE! =(C14*C3)+(C13*C4)+(C1*C5) ((Number of night units * Night rate) + (Number of day units * day rate) + number of days * E7 standing charge)

Total Standard #VALUE! =((C13+C14)*C7)+(C1*C8) ((Number of day units + number of night units) * standard rate + (Number of days * standard standing charge)

Update October 2009

I have decided recently to change to Ecotricity as their Green credentials are the strongest of the UK suppliers. They also offer an E7 tariff and , in my area, the percentage night-to-total use rises to around 25%, which is doable for me.

what to do with an old hard drive

2009-05-24T21:55:00.026+01:00

Over the years I have accumulated a number of old PC hard drives. Some were mine, other from neighbours & friends who asked if I could "fix their PC". It occurred to me that I had a burden of responsibility for these disks, if someone got hold of them they might be able to extract confidential data. In each case I had determined that the disk was dead or dieing, having failed the manufacturers full surface test (via Ultimate Boot CD - UBCD), but conceivably the disk could be attached and read on another computer. They might contain personal data - so I decided to put these disks beyond recovery.

First you have to dismantle the hard drive. this involves removing all external stickers from the disk, as some covertorx-headed bolts , and with a torx driver (purchased from your local hardware shop - in my case the always helpful W. C. Baker of Fleet ), remove all the bolts which hold it together.

An excellent video of how to dismantle a hard drive is here :

Once the disk is in bits then probably the data is beyond economic recovery. I doubt the platters are interchangeable, and they do not appear to have manufacturers model numbers. Putting one of these platters back in a different model of drive is unlikely to make it readable IMHO, though I dare say a Police forensic department or similar could extract data still, should they have a burning reason to do so. But it could take days.

However ..

having dismantled the drive you have a number of useful artefacts :

1) the shiny platters(s). These are perfect round mirrors. They could be used as small camping mirrors, for bird scares, wind chimes (they ring) but I decided to make them into a solar collector. I went to my local hardware shop (W.C. Baker of Fleet) and described roughly what I wanted to do, leaving out the bit about building a death-ray. It turned out that standard UK plumbing washers are exactly the right fit. These are 1" / 24mm diameter rubber washers of approx 3-4mm thickness, with a hole in the middle. These provide a snug fit for the hole in the centre of all hard drives. Better still the discs can be tilted slightly on the washer fitting, enabling a degree of focusing of the beam. I attached all 9 to a piece of wood. this enabled me to fix this unit in my garden and direct and focus the "death ray" onto a specific target. You can see the awesome effect of this in the picture at the top. You can also see how the projected light from the sun was focused onto my wood pile. It did not spontaneously burst into flame BUT you can clearly see the focused beam in the picture and it was noticeably hotter when I dared to put my hand into it. My hope is that it will over time, slowly dry out some fresh logs I have cut.

The other components you have from the disk drive are :

2) A very strong magnet or 2 - these work as the best fridge magnets you will ever find - you can probably hang the phone book on your fridge with one of these.

3) The disk controller board. This is the circuit board on the back of the disk. If you are lucky you can sell these on eBay, as they contain no data , but may have a resell value for people with the same disk, but a failed controller board. I sold around 1 in 5 and they don't go for much, just a pound or two but that's OK.

4) the drive case - the chunky black part is high grade aluminium and can be recycled at any Alu recycling plant, or by local charities for funds. Near me is an Aluminium recycling plant (T Baker (JNR) Ltd of Farnham) who, if you ask, will donate your recycling proceeds to the Phyllis Tuckwell Hospice, a very worthwhile local charity

5) The drive heads. this is a wonderful looking piece of engineering. I have no particular idea what to do with them. they might be used as strange jewellery or parts of a robotic fancy dress party. I'll save these for future use.

6) Assorted precision fittings and washers which could be used for all kinds of projects, jewellery, collage etc. There are a set of bearings which the drives head unit uses to pivot.

7) The stepper motor. These electric motors require some logic from the driver board to work, so I don't think you can use them as 12v motors. Again - a future use will present itself, but for the moment they provide nice little stress-busters, as rotating the centre part produces a a satisfying little click as it goes round.

8) The disk "lid" i.e the silvered metal sheet which covers the entire disk enclosures top surface, as opposed to the side with the controller board. On closer inspection, once removed, you will see that this appears to be a sheet of metal. In fact if you tap it you see it's heavily damped. My belief is that this is a sandwich of 2 metals and some deadening material to reduce the mechanical disk sound. It probably also has a thread of rubber adhesive running around the entire circumference which made a gas tight seal with the main body. Leave the rubber in place. I have used a couple of these metal pieces to sit CD players on to stop skipping as they seem to act deaden vibrations.

Current Cost CC128

2009-05-10T08:20:00.007+01:00

My latest gadget, which I have had for a few months is the Current Cost CC128. This device displays how much electricity we use , in real time. It also gathers historical data which I can download to a computer for analysis. There are quite a number of these type of devices from Owl, Wattson and others and they all work in broadly the same manner.

The device has a display unit , as seen in the picture. It communicates with a sensor which you install alongside your electricity meter via a wireless connection.

The sensor unit has a clamp which loosely attaches around the cable leaving your electricity meter. That's it ! Takes about 2 minutes to install and does not require an electrician. Nor does it interfere with your electrical supply or wiring in any way. Completely non invasive.
I had previously had an Owl device which worked very well, but did not record historical data. I passed this unit on to my mother-in-law who is using it in France, where it also works well. I chose the CC128 for a number of reasons :

It keeps historical data
There are a number of open source projects to provide integration between the unit and a PC/Mac for charting and analysis. Dale Lane has written a very nice application to allow you to download the data via a USB cable to your PC. Dale is incredibly knowledgeable on this subject , extremely helpful and his application is free, so I salute him.
CurrentCost hope to bring out additional sensors in the future, possibly including gas, oil and water consumption too. These would have to use a different approach. The electrical sensor uses induction to measure consumption. With gas the sensor would have to be an optical sensor which 'counted' the rotations of the smallest units on the meter.

I'm trying to reduce my energy consumption, which I consider to be far too high - approximately 21 units per day, and I'd like to reduce it . The CC128 helps me in a number of ways. I have the display setup on the window ledge by my kitchen sink. I can see at a glance what the consumption is . It's sensitive enough for me to be able to tell if unnecessary devices are on, for example lights left on upstairs. I can see the data presented in various forms, including KwH, KG CO2 or Cost in £/€/$. It also gives me a handy graph of the previous 24 hour period, in terms of Night, Day and Evening use.

Now of course I could gather some understanding simply by recording data from my own meter every day. This would go a long way to understanding how much I use. However it's a lot easier with a device like this. Daily data does not provide any insight into which devices use the most electricity. I expected electric ovens , kettles, and dishwashers to use a lot, but fridges , pumps and motors also consume a fair amount. I can see that an old fridge I have for storing drinks in creates a huge spike every time the compressor kicks in. I can see that my washing machine uses 80% of it's power in the first 10 minutes of operation .

Dales application can analyse the data from the device and provide some insights , for example trending (am I improving) , which days of the week do I use the most (Sunday dinner and Monday wash day are the worst offenders) and when during the day am I consuming the most. This is important as I have Economy 7 cheap rate electricity from 23:30- 06:30 GMT which is 00:30 - 07:30 BST. My spreadsheet calculations are that, with my Scottish Power tariff I need to consume 9% during the E7 period for it to be worthwhile. If you have E7, your tariff may vary, I have seen some where the break even point is as high as 38% !

I am also using imeasure , a a wonderful web site run by Oxford University , for recording your weekly GAS and Electricity readings to see how well (or not) you are doing against similar types of property. It even gives you an A-F energy rating, just like the ones you see on White Goods. Currently I'm an E , which I'm not happy with, but all this analysis is helping me improve that. More to follow on this subject

Replacing the brushes in an AEG lavamat 72620 washing machine

2009-01-17T11:01:00.007Z

I'd recently noticed that clothes were often still rather wet when I took them out of the machine. This started around November, but lately the fast spin cycle has been accompanied by a crackling noise from somewhere inside, like sparking or arcing. I had similar symptoms with an old Servis washing machine, and it meant that the brushes were worn down to the brass spring-cups. Finally last Monday the machine failed it's cycle displaying a C9 error code.

I decided that it was worth replacing the brushes, and after a lot of searching (not all spares sites are easy to us) I got through to AEG/Electrolux spares department and ordered a new set (£16 per brush and you need 2 !). Today I'll fit them, using the approach taken in the companion blog on replacing bearings.

Update

All went well. I :

1) disconnected the washing machine from the electricity, water & waste

2) emptied any water from inside via the filter hatch (lower front left hand side)

3) Put some old towels on the floor and pushed the washing machine over on it's front

4) removed the 5 screws holding the back panel on

5) removed the electrical connections to the motor (1 mains block and one earth spade connector)

6) loosened the 2 bolts holding the motor to the "spider" frame . Both 13mm.

7) Supported the motor with one hand , and removed the bolts with the other.

8) Took the motor out, removed the old brushes, held in with 2 Philips screws each

9) Cleaned the armature with some cotton wool buds and ethyl alcohol.

10) Fitted the new brushes. The old ones were completely worn away

11) Refitted the motor, used a metal bar to tension the motor against the drive belt while I tightened the bolts

12) Reconnected and tested

13) Worked Fine

Great Gadgets : Honeywell CM67 programmable Thermostat

2008-10-31T09:01:00.008Z

A gadget which I think goes largely unnoticed by consumers and plumbers alike is the programmable thermostat. I have a Honeywell CM67 but the model has been superseded by the CM907. Worcester do a similar model- I'm sure there are others as good. You can get them for around £50 online at plumbing shops.
There is a range called Heatmiser too, you can find these on Amazon.

These devices replace the room thermostat, which in many cases is the classic "clunk-dial" based on a bi-metallic strip. These old style stats are based on 19th century technology. Their strip is made up of 2 thin layers of different metals, bonded and when it heats up the strip bends due to the differing rates of expansion of the two metals and this forms the pole of a switch. Because of the phenomena of hysteresis, these types of system are slow to respond, and take too long for the stat to switch on or off. To compensate , they actually have a tiny heater in them and actually consume some power. The heater is a wire wound resistor next to the strip which attempts to warm the strip to make it more responsive. It's an inaccurate, old and ugly solution, but still widely used as it's cheap, reliable and I have to say is popular with the innate 'conservatism' (small c) of typical British Plumbers/Builders in my experience.

A programmable stat does the same job but also a lot more. First off it uses a thermistor to measure the temperature which is a much more accurate way of telling if the house is at the right temperature. Of course both systems suffer from the problem that the placement of the unit is where the measurement is taken , which might not be typical for the whole house.

The programmable part is that you can set the device to achieve specific temperatures at different times of the day. So I can set the unit to achieve 19 degrees in the morning for getting the kids up, drop back to 17 during the day, go back up to 20 for home time and climb to 21 in the evenings, for example. There are 6 temperatures per day and 7 days, each of which can be entirely different.

In addition there is an optimum start feature which uses the previous days 'data' to see when the system needs to come on to achieve a specific temperature at a specific time. This takes the guess work out of starting the system early . It also means that while fluctuations in weather will tend to catch it out, it will gradually adapt to colder/warmer weather and switch on earlier or later as the temperature trend changes.

Finally there are various exception programmes (holidays, work at home days, party mode !) where the programme is overridden for a specific number of hours or days. If you go on Holiday you can specify that the temperature drops to a lower one, but comes on in time for the house to be warm when you walk in, n days later.

I actually use this as my main CH programmer, so the CH timer beneath the boiler is permanently set ON all day/year round. This doesn't mean the heating is always on , as the stat controls the pump & boiler circuit, it just means that the decision as to when the heating comes on is based on a more specific set of conditions. I use the stat to programme , which means that it picks up a cold day in summer, but conversely doesn't come on for those sunny days we sometimes get in October.

While these devices are not specifically 'Green' i.e they may increase your consumption,my experience is they decrease it . The house is heated much more accurately , and not based on arbitrary click-on/click-off based on who last walked past the stat. The different modes mean that I can have the system throttle back for holidays or simply the odd few hours when I might change the pattern of what I day (work at home, or in the office)

The programming can be a bit complex due to all the options, but it's not too hard. Finally wiring one up is pretty simple, the main complication is that if your remove an old room stat you have 4 wires to deal with whereas these units only need 2. The reason being they don't need to have mains power for the tiny heater element I mentioned.

I should add that my unit has an additional module on the right hand side. The CM67 had a few bolt-on units, as is apparent from the pictures. Mine has the clock module which sets the time every day from the UK Time Service broadcast on Long Wave. This means that :

1) This clock is always accurate, so acts as a useful time source

2) I don't have to re-programme the stat when we change with daylight saving time

I have also seen a preview of new internet-enabled programmable thermostat called the ecobee on the web. It's being made by a company in Canada and uses on-line weather forecast information in it's optimisation. So for example if the forecast predicts a sudden cold spell the stat factors this in it's calculations and , for example, starts to heat the house earlier if there was a sudden cold spell. This is a great idea, though the device is predicted to cost around $400

A Dalek at the bottom of the garden

2008-05-03T10:02:00.009+01:00

I emptied out one of our compost bins, and at the same time took delivery of a couple of old leaking water butts which I cut the bottoms off to use as compost bins. The bottom of the water butt and the LinPac composter just seemed to suggest one shape to me .. a DIY Dalek !

The eye is one of those old adjustable legs from a fitted kitchen cabinet carcass with a CD stuck on the end . Two brass cup hooks support the limbs - in this case a plastic compost 'spike' (Found in France - a handy device to push in and aerate the compost) and a tool for clearing a shredder, but any garden tools could do.

Shame no body makes an Official BBC Dr Who Dalek shaped garden compost bin

It's for the kids you understand .....

Update September 2009

this one constructed in the same way, except the waist section is an old washing machine drum, which I use as an occasional incinerator/ tree stump burner.

The original has moved and now has a motorcycle tyre adding some stability to it's head, and also denoting rank. Its also has the bottom covered with chicken wire to prevent rodent intrusion, assuming they were brave enough

How to build a gaming PC from a Dell Dimension 2350

2008-03-19T07:06:00.064Z

"What a piece of junk!"

"She'll make point five past lightspeed. She may not look like much, but she's got it where it counts, kid. I've made a lot of special modifications myself."
: Luke Skywalker and Han Solo

In 2002 I bought a Dell Dimension 2350 P4 2.6Ghz PC for my young son. His older brother had a PC, and it seemed right that he should have one too. I didn't know much about PC's at the time and assumed "Integrated Intel Graphics" meant that it might be a bit slower than a dedicated graphics card, but the price was good.

Alas I learned the hard way that if you (or your kids) want to do serious gaming on a PC you need a dedicated graphics card slot , typically AGP or now PCI-Express aka PCI-E. The dimension 2350 did not have one, only 3 humble PCI slots .

Thus began a long and enjoyable exploration of how far I could push this entry-level PC toward serious gaming , and along the way I learned a lot about PC's and have grown to really like this funny old machine. Many, many people have said I was mad and should spend money on a new mother board , graphics card etc. I also have other PC's, some of fairly recent build and performance, but like one of those cheap old cars with all the custom trim, this Dimension has been a lot of fun, even if it makes no economic sense

Almost immediately I found out that many games simply didn't run on the on board graphics, so the first upgrade was to a PCI geforce4 card I bought in panic at Maplin . I found a posting via google of someone who had been through an identical experience and it detailed how to change the graphics adaptor from on-board to PCI in the BIOS. I followed the instructions
I fitted the new card, booted, let windows discover the new second graphics adaptor, made that the default via the Desktop Advanced settings. Rebooted to BIOS and changed the primary graphics adaptor from Auto to PCI.

It worked ! I disabled the on-board and ran entirely from the PCI card which was able to adequately play "X-men: Wolverines revenge" - the offending game which baulked at the on-board Intel graphics chip set.

Thus began the arms race between me and the games industry to see who would blink first. With each new game I have to further tweak the hardware or settings to squeeze a bit more performance out of the old Dell. Here is a picture of the inside of her today, much changed from original.

Over time I'll elaborate on things that have been worthwhile to do, many of which are applicable to any PC. At this point I should introduce a document which I have found invaluable , it's Koroush Ghazi's Ultimate XP Tweak Guide.

There is a good deal of misinformation on the internet about tweaks that supposedly make XP go faster. many are bogus or misinformed, whereas Karoush's guide is well argued and very rigorous . You can obtain the full guide for free , or by making a donation obtain a copy which is print formatted. I printed off a copy and very well thumbed it is too.

Graphics cards

PCI is considered fine for most general requirements i.e sound cards, modems, additional USB ports, network cards, but was found to be a performance bottleneck for Graphics and new standards for interconnects were developed (AGP, PCI-express). Many people don't believe they even still make PCI Graphics cards, worse many refer to PCI-E as PCI , lazily dropping the E-for-Express part, unaware of the difference.

PCI Graphics cards are alive and well and continue to be manufactured. They may not have the very latest ATI or nVidia chip sets, but they exist, largely because people still need them toi either run additional monitors or because they need graphics cards for small form factor media PC's which require silent operation and low power requirements.

I have tended to stick with nVidia cards, more out of comfort than any specific reason. The steps below relate to nVidia but I there will be a similar approach to installing ATI-based cards. There are a wide range of ATI PCI cards, the top model currently being the 1550.

I have gone from geforce4 FX440 to a 5200, 5500, 6200 and am now on an Sparkle 8500GT 256Mb PCI card. This may seem insane to a dedicated custom PC builder, but actually it's not. The resell price for old cards, especially PCI is pretty good and I have always got between 50-70% of the original cost back on ebay. Most PCI graphics cards sell from between £25-£50 which is not that much for a graphics card.

the latest card ,which I've only had for 3 weeks is the Sparkle 8500GT 256Mb PCI card. I think this is probably the highest spec nVidia card available in a PCI form factor, though there are ATi cards of similar spec. This card is quite something, being based around a mid range GPU chipset. It's physically large and occupies 1 1/2 PCI bays, so if you use other cards or PCI case fans , you may need to factor this in. It's so thick because of the huge heatsink, so this is entirely passive in it's cooling. It's also a little unusual in that when it was installed XP detected a PCI-to-PCI bridge, so I believe this card implements some kind of PCI-to PCI-E bridge technology. Examining the PC using Control Panel > System > Hardware > device Manager , I can see the card as normal BUT if I use the PCI latency tool it does not see the card, only a PCI-to-PCI bridge device occupying that IRQ. The latency value is also set to zero, as is everything except the network card (32).

I ran the 3dmark03 test on it and achieved a score of 4777, where the previous best card, the 6200 only scored 2011. In gaming too it seems significantly better, my son can play S.T.A.L.K.E.R: shadows of Chernobyl on full resolution and with high detail. Only dynamic lighting has to be disabled, in favour of static lighting, but that's really the icing on the cake.

Drivers

I try and run the latest drivers from nVidia and the latest version of DirectX from Microsoft. Installing the latest drivers is pretty easy, though I follow a slightly more complex install route, based on Koroush observation that some driver components don't always get correctly updated:

Download the latest nvidia drivers from the nVidia driver homepage and save to your disk. I also download driver cleaner pro and install that. Don't run it yet.
Remove the old driver using the Control Panel > Remove Software >nVidia drivers. Don't worry , the PC works OK with no driver. If you rebooted it uses a generic VGA driver. Not as pretty, but you shouldn't be left with a blank screen.
Reboot into safe mode. I seem to have problems getting the Dell to do this from the boot menu, so while not in safe mode, I first run msconfig from the "Run" option, then select the boot.ini tab and check the SAFE mode box. Later I uncheck this once.s I've carried out step 4.
In safe mode I run Driver Cleaner Pro to remove any pieces of the nVidia driver left behind. I have also used this approach when updating the SoundMAX sound drivers
re-run msconfig and unset the SAFE mode and reboot to normal mode
On reboot XP will discover the graphics card and prompt to install the new driver. I cancel this then run the nVidia driver update
The nVidia install will run through and may reset the monitor to the lowest resolution. It will also ask you to reboot in order that it can start a number of background processes : 1 a service and 2 via the users start-up tray. You don't have to reboot here, I normally reset the resolution by right-clicking on the desktop background > properties then set the resolution as high as I can.

Latency & IRQs

PCI graphics cards are way behind AGP or PCI-E cards in terms of performance, largely because the PCI bus is the bottleneck and is shared with other devices which in turn, may be making demands of the bus at the same time. There is not a great deal of info on PCI tuning regarding graphics cards, but there is a lot on tuning for sound cards. People involved in using PC's as recording studios sometimes find that their soundcards suffer from lag or 'stutter'. This can be tweaked by modifying the PCI latency, or in other words the time alloted to each device on the bus. In the Dimension , while the sound card is part of the motherboard, it does show up as a PCI device
(Control Panel > System>hardware>Device manager > View by Type).

Another concept to introduce is the IRQ (Interrupt request). Each device in the PC is allotted an IRQ number at boot time by the XP PnP Service, other OS's such as Linux have a similar concept. There are 16 IRQ's (00-15) and so devices typically have to share their allotted IRQ number, as devices include things like the USB interfaces,floppy drive and some of the ports at the back of the PC.

Two concepts worth pursuing in an attempt to squeeze as much performance as possible are :

1) Try to distribute IRQ's evenly i.e don't have CPU intensive devices sharing the same IRQ (Sound, Network, Graphics)

2) Tweak the PCI Latency values to get the best balance

To achieve 1) First you need to look at which devices have which IRQ's using the system tool in control panel. I try to keep Sound, network and Graphics cards on separate IRQ's. But how do you control them ? Well If you are really determined you can override the Universal Plug And Play service, but I prefer a different , safer approach.

First I decided to remove or disable anything I didn't need from the PC. Not only does this free up some room for the PnP service, it also frees up some current for the PSU (see later) and frees up some physical space to improve cooling (also see later). I never use the floppy drive, the only possible use might be if I ever decide to flash the BIOS, but the A01 version I'm running seems to work fine. I could have disabled the Floppy in the BIOS, which would have left it in place should I ever need it. But I decided to open the case, remove it's power and IDE cable then remove the whole drive and frame that supports it below the hard disk. At present I don't have a way of blanking the slot at the front , but I'm not too worried.
Similarly I removed the Modem card, which I never use, which left just my Graphics card as the only device occupying a physical PCI slot.

Next in the BIOS I disabled the floppy drive and I also disabled the Parallel printer port on the back. This PC will never use a parallel printer, most are now USB or better even networked. I also disabled the parallel printer service associated with this

After this if you boot you may see if the balance of IRQ's has been affected. The other thing you can do is move the graphics card to different PCI slots. the affected the IRQ distribution for me, and I found the top slot gave the best balance i.e Sound, network and Graphics on different IRQ's. Some are shared still, typically with USB devices, but this was the best I could achieve.

PCI latency can be examined using the freeware PCI latency tool . I found that maximum latency was being set by my graphics card already (248) and as I was not experiencing any sound lag or stutter symptoms I decided to leave it. I also reasoned that perhaps my network performance might be being affected by the latency dominance of the Graphics card but it was too hard to tell. So I left it alone, but if you are experiencing sound problems and the graphics card is set high, you can use the tool to override this and see if halving the graphics demand improves things. It's a popular and proven mod in the sound card world.

Memory

Soon after starting down the PCI graphics card route I also upgraded the memory. The Dimension has two memory slots and can support up to 1Gb. I upgraded to a full 1Gb made up of 2 512Mb PC-2700 simms. I have also played around with the accompanying page file, experimenting with having no page file. In the end I created a dedicated page file partition of 5Gb on my second hard disk and created a maximum size 4096Mb page file in it. I disabled XP recovery mode on that partition. the evidence was that having no page file made no appreciable difference and some large games (i.e. S.T.A.L.K.E.R) might exceed my 1Gb memory capacity.

Based on feedback I received from Robert, I tried 2Gb of memory (2 * 1Gb SIMMS) and the Dell is able to see this in both the BIOS and XP, so now I'm running this machine with 2Gb main memory + 4096Mb page file in it's own partition.

I try to keep the number of processes to a bare minimum, thus maximising the available memory for games by the following :

reducing the number of services to the bare essentials via control panel >Administration >Services. If in doubt I set them to Manual rather than automatic and see if the server still functions OK
Ensuring that few if any programs are run at startup. Plenty of software will add something to your start-up when installing. If you ever install something that requires a reboot, its a pretty sure bet that it has either added a service or a start-up script. Sometime these are necessary for the running of the product i.e. Norton Ghost, but often it's just a lazy background update programs. While small they all take up memory and some processing power.
Checking that no extra tool bars have attached themselves to the browser. I have noticed that various parasitic programs seem to be able to attach themselves to IE in the guise of tool bars. Some are valid, like the google tool bar, others seem to be designed to misdirect your browsing to sponsored sites. I try locking down the Internet Explorer tool bar via it's settings, then checking from time to time what has attached itself using HijackThis

If I reboot the PC , then login, then wait a few minutes for XP to settle I see 18 processes running in Task Manager. A neighbours PC which had some malware problems had 67 processes running. Now some were due to specific graphics or sound cards and were necessary, but you see what can happen.

Disk Drives

The Dell Dimension 2350 came with a single 40 or 60 Gb Western Digital drive as I recall. This soon filled up and was supplemented with a second drive a 7200RPM 200Gb Samsung.

Fitting a second drive is a bit of a challenge as the Dimension is quite limited in space,but there are two places you can fit it:

1) In the second tray beneath the existing CDROM drive at the top. This requires a an adapter kit to privide the side runners to slide into to the slot. The Dell had a rounded grey blanking plate on the front which removes to gain access.

2) Piggy backed behind the existing system drive at the bottom of the front. The Dell, like a Compaq we also owned , has the system drive in a vertical position at the bottom front of the case in a metal 'cage'. It's possible to obtain a second metal cage from here, which will attach to the first to add a second bay. This position at the front receives cool air which enters the case via vent holds behind the plastic front piece.

I opted for option 1) and for some time this ran well. However after a while I noticed some 'Bad Block' errors occurring in the system event logs (Control Panel > Admin Tools >Event Viewer > System )- a tell tale sign of a disk going bad. Worse a few related to the main system disk too.

When disks go bad

I therefore cloned the disks to 2 new Maxtor disks which I replaced them with. I used Norton Ghost to do this. Cloning failing drives is not widely recommended on the basis that a copy of a bad drive will be incomplete. Of course the alternative which is to fresh install everything, isn't much better. Interestingly the old Seagate tools had a feature where they would analyse the surface for bad blocks AND reveal the filename of any files affected. This meant you could clone the disk knowing which files might be bad , and make a calculated decision. They might be unimportant i.e temporary files or possibly could be copied from another PC.

To try to avoid this I now employ a number of strategy's :

1) I routinely back up the drives with Norton Ghost

2) I monitor the system event logs for bad blocks or any RED error messages

3) I enabled S.M.A.R.T logging in the disk drives via the BIOS. S.M.A.R.T is a standard way that modern disk drives gather statistics about their performance. This is gathered by the disk and stored within the disk itself. Various tools are available but the data can be quite cryptic.

My favourite tool - speedfan has a tab to display the S.M.A.R.T data in a meaningful 'health check' form, and better still, has a button to compare your disk performance data with others. This is very useful as , for example , it gives a more detailed percentage health measure and how the disk temperature compares with the average for it's type from real users across the internet.

I try and keep the disks as cool as possible and both rarely creep above 30 degrees C (see cooling). speedfan is also able to display the real time temperature of a range of components within the PC including both disk drives.

NOTE: If you ever replace a failing disk drive, you may find the new disk is running slower than expected. This is due to a feature in XP, where, if a drive starts to report errors, XP downgraded the IDE protocol in an attempt to reduce the errors. However , when a new disk is added, it doesn't change back. See a discussion of this on the Microsoft website
It's easy to fix, but may go unnoticed. In my case it manifested it self as general slowness, particularly the number of passes the coloured "pellets" have to make in the XP boot screen. Normally this is around 6-7 but it was 20+ before I fixed this using the above article.

Partitions

The general consensus as far as I can see, is that Windows seems to favour single large partitions. However I don't, and have my second disk divided up into smaller partitions. For me the greatest weakness with windows is how hard it is to move data and applications around and distribute them across disks. I don't mean drag & drop, cut & paste. I mean "I have a new disk and I'd like to move existing applications and data onto it ". This is easy to do on a Mac, but on windows reinstalling the application on the new drive is the only answer, you'd think it's in everyone's interest including Microsoft's , to make this easier ?

I try to distribute the IO across both disks to achieve some interleaving, especially as my drives are on separate IDE controllers, so I have my operating system (XP Home) on the C: drive and most applications are on the G: drive. In addition I have 2 specialised partitions P: & Z:

P: is a 5Gb partition and only contains the page file. recovery mode is disabled for this partition. No file system files are on it and it's not backed up

Z: is a 5Gb partition and holds all TEMP data. i.e the Control Panel >System > Advanced > Environment Variables that refer to TEMP files all point to this partition , both system and user. Also the IE temporary internet location also points here. It has no recovery enabled and is not backed up. This approach is one from my old Unix system admin days, and it has some advantages. For example this area can get very fragmented, but that fragmentation is contained within this area, so I can defragment this partition very quickly without defragmenting the whole C drive (the default location for all TEMP files). I also tend to download files to this drive and unzip them there. that way my main partitions don't fill up with forgotten downloads and unzips i.e nVidia drivers all get downloaded and expanded here. This drive also has recovery mode disabled, if I lose it it doesn't matter.

I routinely clear out junk from the drives. You can do this via Drive properties (right click on specific drive) which removes recognised temp files and older temporary internet files. I also quite like CCleaner

It not only enables you to analyse and remove all 'crap' from across all drive in a single pass, it provides tools to let you see what programs will auto start (and remove them). Some of it's options can be a bit severe, for example removing recent IE history or recent file open lists, but you can customise it .

I also routinely defragment all drives using the standard included XP tool (right click drive properties), though the TEMP more than others which mostly remain fairly static.

Power Supply

I'd assumed that power supplies required no thought, they just worked and never failed, and if they did , the PC would just not power on.

I have had 2 power supplies fail. One in a Compaq which was conservatively rated at 250WATTS, and one Christmas eve I decided it was the perfect time to attach a 3rd hard drive . This was in addition to a DVD combo and AGP Graphics card. At start-up there was a lot of smoke and a strong smell like the UK Antiseptic 'TCP' ! Worse the Compaq had a unique power supply, so an off the shelf ATX supply would not fit. That marked the end of that particular PC.

the original Dell Dimension 2350 came with a 250watt supply, and this failed but ina a different, gradual way. I noticed that mostly at boot up the CD drive was not there, yet I could add it via 'Add new hardware' and it would magically appear. I thought little of it, but then started to experience random BSOD's and other odd and entirely unconnected types of problems. It was suggested to me that the PSU was starting to fail, so I got a 400Watt ATX supply, which, contrary to much urban myth DOES fit in the Dell Dimension 2350. I didn't give much thought and bought a SWEEX model from my local PC mod store, you can see it at the top of the picture, it's hard to miss in it's full 'bling' gold plating. What I was unprepared for was how much better the PC seemed after ward. Everything seemed fast/quicker etc.

I don't know if the SWEEX unit is especially good, for another PC I got a green high efficiency PSU which consumes less power.

Sites exist which can calculate the total power rating of the entire PC. I think most PC's now come with a 400WATT supply as standard. speedfan is also able to monitor he power supply voltages on some motherboards though not on the dimension.

The new PSU has the intake fan position underneath,and is larger which improves the cooling of the case.

Cooling

Keeping a PC cool is very important :

1) Components such as disk drives have their life expectancy shortened if continuously subjected to excess heat.

2) The CPU will slow down if it gets above it's operational temperature range. This is a fact, from the P4 onward, Intel incorporated a feedback safety measure, so if you exceed the temperature, the clock slows down. I have witnessed this with a 1.5Ghz P4 Compaq presario which would experience game sound stutter in certain games at certain points. I tried everything, new drivers etc., but could not find a cure. I then found a posting from a French gamer who suggested it was due to the CPU overheating. I downloaded motherboard monitor, which I used before speedfan, and this confirmed that the core CPU temperature was regularly exceeding the maximum for the processor type. In the end the cause was dust "bunnies" i.e. small balls of dust clogging the vanes in the heatsink. I cleared it all out with a can of air duster, then reseated the heatsink with fresh heat grease and the temperature dropped and the game lag disappeared.

3) The noise increases. Some of the fans will regulate their speed with temperature, so a hot PC will have faster fans and generate more noise, which is always annoying.

the classic layout for cooling in a PC is to draw cool air in from the lower front of the case, pass it up over hot components (disks, memory, motherboard) then expel it via a case fan and the PSU fan at the upper back. this assumes a tower configuration. In addition specific "hotspots" have localised blower fans, such as the CPU and Graphics cards.

The Dell Dimension 2350 uses a clever design feature where the case fan also doubles as the CPU fan. A large case fan is positioned on the back, and a large plastic green cowl or hood covers the CPU heatsink. Thus the case fan draws air across the vanes of the CPU heatsink before expelling it from the back. Air is drawn in via various vents at the lower front - to cool the system disk, and via vent holes above each PCI card - to cool the cards.

"If you can not measure it, you can not improve it."

Lord Kelvin

Before you can asses the benefits of any cooling measures you have to be able to measure the temperature first. For example I can see that removing the side panel from my PC causes the hard disk temperature to increase . This seems completely illogical as common sense dictates that more ventilation means better cooling ? However, by opening the case you short circuit the wind tunnel the designers have incorporated and less cool air blows across the disk drives. So you see measurement is important.

Most components have on-board temperature sensors which Windows XP completely ignores. Fortunately tools such as motherboard monitor and speedfan are able to extract this data and report on it. I have another PC where speedfan reports the temperature of the CPU, memory controller, 3 disks and the Graphics card and presents this nicely as a gauge on my taskbar.

Alas the Dell dimension motherboard is a somewhat limited here. While speedfan can extract information about the temperature of the disk drives and the graphics card , thus far (up to version 4.34 beta )it is unable to extract the CPU or memory temperature for the motherboard. This may be because the 2350 was built as a cheap model and lacked some niceties, but there clearly is some monitoring of the CPU temperature going on in the BIOS as the case fan speed alters with CPU workload. However I have every confidence that Alfredo Milani Comparetti, the author of speedfan will find way if it exists.

I have fitted round IDE cables to better allow the flow of air as standard flat cables can behave like sails, and block or reroute the lamina air flow through the chassis. I have also removed non essential PCI cards (modems) and the floppy drive. This provides more internal space and airflow.

The second disk drive is in the bay beneath the DVD/CD combo drive and has a coolermaster aerogate front panel fitted. This provides a) a disk fan b) 4 temperature sensors c) the ability to alter the speed of 4 fans and d) a selectable coloured LED display. Initially I used all the temperature probes, but in the end I only monitor from a single probe attached to the CPU heatsink. The fan is useful in keeping the drive temperature low and drawing cool air in the front. The Blue light also looks cool.

I have a Thermaltake A2426 fan located in the slot above the PCI card, which is just a backplate slot with no accompanying PCI slot on the card. Beneath I have a an Azern BlitzStorm double fan occupying the remaining PCI slot below.

All of this means that the Disks seem to stay below 30 degrees C and the GPU peaks at 52 degrees. I may change this to a double blower pushing air at the GPU heatsink

Overclocking

With adequate cooling and a good power supply , overclocking is probably the last area where I can squeeze some more performance out of the Dell. The Dell 2350 motherboard has no overclock capability, but the Graphics card does. Watch this space ....

Update : Have recently upgraded to XP service pack 3, speedfan 4.35 and forceware 175.19

Update: 27/05/09

Still running well, with Operation Flashpoint as the game of choice for my Son. the Dell has been plagued a bit of late by the winav.exe malware/trojan thing. It's a trojan which appears to look like a legitimate antivirus tool. It sends all kinds of warnings to the screen, dsiabled the control panel security center and places a red warning in the task tray. AVG free edition did not appear to pick it up, nor did Spybot or windows defender . If had to deal with this sucker twice in the last few months. Perhaps I mistakenly used system recovery to convince myself I'd got the system to a point prior to infection, when in fact it was still there.

Anyway I'm now experimenting with avira antivirus and Malwarebytes, which seemed to do the job.

Generally the PC still works well, but is challenged on newer games. the number of background processes has crept up to around 30 with the extra AV stuff plus an iPod/iTunes.
With current nVidia drivers I'm achieving 5700 in 3dmark03 with no over clock. the latest drivers include the nvida real physics engine, which they aquired through the purchase of Aegis. I have enabled this in nVidia control panel, we'll see if it improves game play.

I have written a separate follow with some further tweaks here

Update : 11/07/2010

The Dell is still going strong. I have switched to Avast for anti virus as this seems to have found a couple of viruses that AVG missed. I was also finding that AVG started a lot of processes.

The PC is used less for games now. My son tends to play games on the xbox360, while the Dell is used primarily for facebook, iTunes and BBC iPlayer, but it keeps on going .

I routinely check the disk health in speedfan, but both hard drives S.M.A.R.T data is within standard operating parameters.

Update : 29/07/2011

the Dell is still going strong. I reverted to Avast free edition for AV, which seems to work well, with periodic running of Malware Bytes. Avast has a nice full scan at boot feature which runs a complete scan prior to windows loading, useful to do sometimes.
The Graphics drivers are up to date from nVidia and the graphics card is still the sparkle 8500.

Recently the audio quality became very distorted, though oddly , only on the right channel. I checked both the front and rear audio connections (green 3.5mm socket) but the same . I reinstalled the SoundMAX drivers, used for this machines on-board audio. Again - no change. In the end I decided to get a Creative external Soundcard X-Fi sound unit, which connects via USB, and provides outputs for 5.1 audio as well as headphones. I toyed with getting an internal PCI card, but the huge graphics card leaves little space. Also the Creative USB card was more future-proof than a PCI sound card, though it did cost more . It works well, the sound comes from a completely silent background, no strange chirps or whistles which sometimes plagued Music when played through the onboard sound system.. There are all kind of effect, for example impressive surround sound effects delivered through conventional stereo headphones

I made sure that I disabled the on-board sound in the BIOS, thus freeing up an IRQ and removing the possibility of conflicts. I used Driver Cleaner Pro to remove any remaining SoundMAX drivers too. Now if I look at the Sound devices in device manager I only see the Creative device and not the SoundMax.

I'll re-run 3dmark again sometime soon and post the results. Possibly disabling the onboard sound subsystem might free up resource ?

I also gave the Dell a good vacuum out internally too, lots of dust had accumulated around the vanes of the main heat sink which would have caused it to overheat. Alas there still appears to be no means of assessing the core CPU temperature with software. I believe the rear fan speed is entirely regulated by a built-in thermistor.

The only issue I'm seeing with the PC is that it takes a very long time to take a user to their desktop from the user picker screen. It sits at the "loading personal settings" screen for more than a minute. i suspect it's waiting for something, which times out then it proceeds. I'll debug what it's doing here when I get more time.

Update 24/8/2011

Well the slow loading of settings turned out to be the XP SSDP service which looks for UPNP servers on your network. I don't need this and it probably got enabled by accident. disabling it got the "Loading preferences ..." time back to acceptable.

I re-ran 3dmark03 on the Dell and scored a new all time high of 5317 !!!!!! (I know, a low score on an aged benchmark , but probably appropriate on a PC that's 10 years old )

Not sure if this is due to freeing some resource using the USB sound card ? I doubt it. Actually I noticed that for the first time 3Dmark ran the Audio tests too. previously the on-board SOUNDMAX must have not been recognised by 3dmark. maybe in it's scoring it incorporated some numbers from the 3 audio tests.

Next time I have some free time I'll look at the IRQ spread, now that sound is disabled in the BIOS. I assume that PCI bus contention still is a factor, I'm swapping a PCI directly connected on-board device for a USB connected device, but of course USB also uses the PCI bus.

Replacing the bearings in an AEG lavamat washing machine

2008-02-29T20:42:00.009Z

We have had an AEG Lavamat washing machine for 9 years and it's given very good service, probably running on average, a wash load a day for a family of four, including two teenage boys.

I'd started to notice that the spin was gradually getting louder , and recently some extra random bangs and knocks were introduced into the cacophony. It looked as if the main bearings were worn out, which was confirmed by some play in the drum when pushed up.
I maintained an old Servis washing machine which we had before the AEG, and it was forever breaking down with either clogged filter (nicely placed underneath the machine for zero accessibility) or the brushes wore out in the motor. I had arrived at the opinion that a washing machines bearings were the equivalent of a cars "big end" i.e. something that when it goes is considered terminal and I had no experience of replacing them. But I decided it was worth it as a new set cost around £20 and the machine should go on for some time after replacement. I had little to lose ...

I googled AEG and bearings and found this fantastic website which explains exactly how to remove the old bearings and fit the new ones. While the model was not precisely the same as mine (72620) these instructions worked a treat, and my machine is back in action. How quiet it is now. I think with washing machines and also vacuum cleaners they get louder over time, so you only notice it when you either fix or replace them.

PostscriptYesterday the machine started making a strange noise during pump out. The sound appeared to be coming from the bottom left hand corner, where the water pump is. I thought "Oh no - another part to replace"

Anyway I opened the filter hatch, drained off the excess water and removed the filter. Amongst the water was a lolly stick, which had been in someone's pocket and made it's way into the trap, then had been fowling the impeller blades in the pump. Easily solved

Update November 2011

My washing machine has been working OK, but I had noticed what looked like a growing pool of water on the floor beneath. This was not one large deluge , more something that was accumulating over time. I decided that perhaps the bearings needed replacing again. While not below the water level, some water can escape through the main bearings if the grease starts to break down. I ordered a new set of bearings, waited for them to arrive and that weekend , opened up the back of the machine.

I have no excuse, I have read all the Sherlock Holmes stories many many times, and he always states "never theorise before you have all the facts".

It wasn't the bearings. There was some water staining on the damping on the back of the outer drum, but this looked like it was old, probably from the last replacement. The water appeared to be leaking from where a flexible oval cross section hose mated with the detergent tray. Removing the detergent tray was a rather complicated and difficult job, and while a replacement hose could be purchased I had another problem. The staining on the floor appeared to have a greasy , wax-like property. It could have been that a mix of liquid detergent and water had slowly leaked out from the hose over time, leaving only the detergent residue. However the path that the water had taken seemed to run down one of the shock absorbers. The patch on the floor had a familiar chemical smell, like tins of fluxite. I wasn't sure if the leak had been washing out hydraulic fluid from the shock absorber.

It was going from bad to worse as it would take days to get more components and be a hard refit. The machine was over 12 years old.

I decided it was time for a new machine. I'm lucky to have a local small independent Electrical goods store ("Ace Electric") and an hour later I had ordered a new Bosch Exxcel machine, which was fitted on the Monday, the old AEG taken away.

So that kind of brings this thread to a close. I guess I would have to say that the most important lesson in trying to mend things is knowing your limits and when to stop.

The new Bosch is very good, has all sorts of ECO programmes, a larger drum and a 1400 rpm spin speed which really dries out clothes before they are removed.

And finally, I have a spare set of AEG bearings, if you need them let me know

What to do with an old Sony Playstation 1

2007-11-02T07:15:00.003Z

If you have an old Sony Playstation 1, especially the first variant (there were several iterations ) they do make pretty good stand alone CD players. Sony incorporated the ability for the Playstation to play normal audio CD's into the device and it turns out it does it rather well.
Talking to people, it seems lots of them have aged Playstations either gathering dust or that they "gave to their young cousin" but they really do have a new lease of life. Now that we are on the PS3, the aged PS1 is considered by many to be virtually worthless, despite costing around £300 when they first came out.

The early model (SCPH-1002) already has red & white RCA sockets on the back, and you can use a regular controller for play/skip/stop functions or you can buy a separate remote (around £5 on ebay)

There are some tweaks you can do to them, and especially in Germany , people mod them even adding valve output sections. The best site for modifications is dogbreath.de

However they sound pretty good in standard form. Later iterations dispensed with the RCA sockets, but you can buy an AV adapter pretty cheap on ebay, which provides the necessary RCA sockets

Update 22/12/2010

I have a Tivoli Model 1 FM Radio in my kitchen, which is very good. I had previously hooked a discman up to it via it's AUX input, but the discman eventually died. I managed to get a PSone - the tiny brother of the PS1 which was introduced toward the end. This makes a great small form-factor CD player. It doesn't have RCA sockets on the back, so you have to use the special cable that comes with the device.

Cleaning the Apple Mighty Mouse

2007-10-26T08:03:00.002+01:00

I have an Apple G5 iMac and I use the standard Apple Mighty Mouse with it. This is a wired, optical mouse , with a small tracker ball on top for scrolling. The top ball develops the same dirt build up on the underlying rollers that afflicts conventional ball mice. However the Mighty Mouse is much harder to remove the ball. Mine stopped up scrolling , so I googled this problem and found this excellent site.

http://www.linklink.co.uk/apple-computers/cleaning-inside-and-dismantling-the-apple-mighty-mouse/

It's a fiddly job, and the tiny magnetic rollers are best handled with non ferrous (i.e. non-magnetic) tools, so a pair of plastic tweezers are a big help. I followed the instructions above and my mouse is back to normal. Shame it's so hard and involves forcing the "skirt" section away, necessitating super glue to fix it back on though

Update May 2010

Well once again the tiny roller ball has stopped working. I really like the look of the new magic mouse which has no scroll ball, just senses touch movement. However I want it wired - wireless mice are too much effort for too little benefit IMHO. I always find the batteries run out at the worst moment. I'm not sure that apple will produce a wired magic mouse though :(

Update June 2010

Well I got a Logitech LS1 mouse for fathers day. Worked great out of the box with my iMac G5, comfortable with both a vertical and horizontal scroll capability. there are mac-specific drivers you can download from the Ligitech website. Goodbye mighty mouse ......

Dishwashers

2007-10-13T09:13:00.002+01:00

This article is intended to hi light some things you can do if your dishwasher is not cleaning effectively.

NOTE : Do not access any of the internal electrical mechanisms of a dishwasher unless you are qualified . The purpose of this article is not to explain how to repair the internal electrical systems of a dishwasher , this is for the service engineer.

However if your dishwasher is not performing well there may be a number of things you can safely check to return it to form.

Since the late 80's we have owned two dishwashers. The first was a Bauknecht and lasted around 8 years. Our current model is a Bosch.

Dishwashers are complex machines, taking cold water from the main, heating it internally and pumping the hot water through various jets at the plates, pans and cutlery etc. stacked in racks.
They normally have a number of cycles which make up a program. For washing, detergent from tablets, liquid or powder, is mixed with the water. For the rinse cycle , hot water is mixed with a rinse aid liquid. This not only rinses the plates but increases their temperature and hence latent heat. When the cycle finishes the plates dry utilising this heat to drive off any remaining water. The rinse aid assists the process of preventing streaking from water running off. Salt is normally added to an internal reservoir , which adds it to the water to reduce the effects of calcium deposits in hard water areas.

Because cold water is heated to high temperature (normally between 55-75 degrees) dishwashers draw a lot of current from the mains electricity. That combined with the presence of water means they are potentially dangerous if they fail. I have a colleague who had a dishwasher catch fire, hence my warnings in the opening paragraph.

Most dishwashers have two baskets, upper and lower. Water is sprayed at the crockery from usually two rotors. These have water blasted through them at high pressure. Nozzles along the length of the spinner provide them with rotational energy, while other upward facing nozzles deal with the business of hosing down all your plates, cups etc.

The water is circulated around within each distinct cycle, and typically passes through a trap at the bottom. The trap attempts to filter out larger pieces of food, and to trap grease.

Ironically the best way to keep a dish washer working well is to keep it clean internally .

If plates are coming out with traces of food or detritus still on them then it's normally either the trap or the spinners that have a problem.

The trap can be removed, though how varies from model to model. This sits in the bottom of the washing area of the dishwasher, beneath the lower basket. You can normally remove this by either unscrewing it or some kind of unlocking action . These are designed to be customer removable. Once out, I wash the whole thing in hot soapy water to clear any grease and food items. It's worth doing this at least weekly .

The rotors are the part that actually do all the work. They can have 2 problems :

1) If items block their path and they are not able to rotate , which in turn means some plates will not get sprayed. Always check that no items block their rotational path . A fork hanging down or perhaps a large item on the lower rack blocks the spinner on the upper rack. When the washer is full, it's sometimes worth checking that they can rotate by giving them a push, they should turn easily.

2) Small items of food, especially hard things like small bones or pieces of shell etc. will pass through the filter and enter the recycling water. These will probably end up being pumped into the spinners and then clog the jet holes. You can remove the rotors, normally the top one is suspended beneath the top basket and can be unscrewed. On my washer the lower rotor just pulls out, but be careful. These components should be user serviceable , but you need to be gentle when removing, especially if you have never done this before. If in doubt, consult the instruction manual.

On inspection you may see small items poking out of the nozzles. You might be able to remove these with tweezers. However if not you basically have to shake the rotor and encourage the particles to leave it via the main inlet hole. Usually the centre of the rotor has a tube where it fixes to the water inlet. This is the only other way out for the detritus to exit the spinner. Shake vigorously up and down until the small pieces fly out. This is easier to do when the rotor is dry.

I have also known scale to accumulate inside the rotors if the salt has been allowed to run out for an extended period. Soaking in a hot water and distilled vinegar is good for removing scale or using a commercial dishwasher cleaner which you place inside and run the washer on the hottest wash.

Having removed the blockages from the rotors, it's worth washing them in soapy water , checking the part where water enters and the whole device pivots around. various detritus can accumulate here and impede the smooth running of this bearing. Coffee grounds seem to accumulate in this area on mine.

Lastly, run water from the tap through the spinner to see if any remaining fragments inside get pushed out and block the nozzles. Once all nozzles run clear , install back in the dishwasher.

A final area worth checking is along the rubber seal that runs the width of the machine at the base of the door. More "gunk" can accumulate here, and is worth wiping with a cloth. Mind your fingers as you may get folded steel joints here, which can be sharp. You can also check to see if the seal is starting to perish, which may happen over time. It's possible to get replacement seals.

Update December 2007

If technology is going to fail it seems more likely it will happen at the worst time i.e. during the busy Christmas season. The dishwasher stopped pumping out water. I could hear it trying but no water was pumped out through the waste hose.

I removed the filter and felt around in the dirty water for any obstructions. I then used a wet & dry vacuum cleaner to suck out the water so I could inspect the chamber that the filter sits in. I noticed a couple of hex bolts which held in place a small plastic cover. Removing this exposed the pumps impeller, which I felt around to check for obstructions. This was starting to look like a failing motor, but I decided that maybe if I left the impeller exposed and soaked the whole assembly in warm soapy water for an hour. I then put the cover back in pace, tightened the hex bolts , though not too tight. Hurray - it now pumps out and appears back to normal. I think maybe some debris got into the motor impeller chamber which was cleared when I removed the cover.

Update February 2008

Well the dishwasher limped through Christmas and new year and finally the pump gave out in early Feb and wouldn't respond to the previous treatment. I decided that the dishwasher had exceeded it's useful life of 8+ years, as replacing the pump was likely to be an expensive and difficult part to fit.

I purchased another BOSCH Exxcel model, which is basically the same as the one it replaced, though it has a lower electricity consumption on the economy wash, with lower water useage. What is gratifying to see is that Bosch have improved on some of the old design in a number of places, for example there is no longer a need to press the select button after closing the door each time to continue with the program. the baskets are better thought out, and the top rotor too has improvements where detritus used to collect.

Homemade diy xbox 360 cooler

2007-10-05T20:16:00.003+01:00

My Son was kindly given an Xbox 360 a year ago by friends who returned to the USA. It's been great but in May it died, suffering the red circle of death problem. This is the Xbox equivalent of the PC "Blue Screen of Death", though is actually worse, as it indicates a hardware problem. The Xbox 360 has a very powerful dual core processor and onboard graphics, and throws out of lot of heat through two case fans at the back. When running a game, it resembles a fan heater. Some of the problems appear to be heat related. This is well covered elsewhere on the web and I don't intend to discuss it here. Microsoft have extended previous warranty to 3 years. We got ours repaired by Microsoft and it's up and running again fine.

However I decided I'd try to avoid this happening in future by improving the cooling of the unit. I did not want to open the 360 as this will void the warranty, which still has 2 years to go. I looked at commercial 360 coolers, and these seem to fall into 3 groups :

1) Bolt on fans at the back of the 360

2) A fan base unit. basically a stand that the 360 sits on in it's vertical mode. The base of the stand has a fan which blows cool air up through the holes in the base/right side of the 360

3) Using a commercial laptop cooler

Of these I didn't really fancy any 1) Bolting additional fans on where existing fans are already desperately trying to vent hot air seems a bad idea to me. The lamina flow of air around propellers is a complex science. Fitting two fans back-to-back does not seem to me to guarantee double the air flow. The fluid dynamics of air flow around fast rotating fans of different size and rotational speed seems haphazard at best. Worse it may actually reduce the air flow . Also if the external fan fails, its blades will block the existing fans exhaust route.

2) Is a better idea as it blows air in to the inside using the holes provided by the design. However this only works if your Xbox in in vertical mode, which is not ideal for where my Sons Xbox is placed.

3) Seems a good idea and some laptop coolers can run using the power from a spare USB port, which the 360 has. However most seem to be too noisy. The 360 fans are already quite loud, so adding more fan noise seems not ideal. These start at around £20 (UK sterling).

I decided to build a homemade, DIY xbox 360 cooler from junk I had lying around.

The "Brick"

First off lets consider the power supply or "brick". This provides high current DC voltages to the motherboard within the Xbox. It's a switched mode supply, probably not unlike the ATX power supply you find inside a typical desktop PC. Early Xbox adopters found that if these are not well ventilated they may cause the 360 to freeze during game play. It's a good idea to keep the "brick" unit well ventilated. If you look closely it has small ventilation holes at both ends. When in use you can feel hot air emerging from the Xbox-end of the brick. I'm not sure whether this is due to internal convection or a slow moving internal fan (I have not taken it apart). I decided that the "brick " should be kept off of the carpeted floor and allowed as much clear air flow as possible. I have it raised up on an unused metal frame that came out of our oven. Metal is also good as it will act as a heat conductor to a small extent. I'd guess that the best position for the "brick" would be mounted vertically, with plenty of free space around both ends , with the console end upward and the mains electricity downward, so the hot air would naturally rise through convection, like a chimney.

I also periodically blast some air-duster through the vent holes in an attempt to clear any dust that may have accumulated internally. I once had a PC where the core CPU temperature was reduced by 10 degrees just by clearing dust "bunnies" from between the heat sink vanes.

The Base Unit

I decided that the base unit should be placed on a flat metal surface. When placed in the horizontal orientation, the feet of the Xbox keep it only a few millimetres off the ground. While the primary inlet holes for air are at the sides, there are a few holes underneath toward the edges. A certain amount of heat will be conducted through the case work. Placing the unit on a soft insulating surface .i.e. on a carpeted floor would seem a bad idea as air flow would be compromised and case heat would be insulated, and not conducted away into another surface.

I found a black metal shelf from some bedroom clothes rails units we had once used. These are steel shelves with a 10mm overhanging lip along each edge. I'd guess any box section shelving would do. Initially I used the shelf simply as a flat surface to place the unit on, as the wooden chest it normally sat on has an uneven surface and hence would compromise any airflow beneath the unit. Being made of matt black painted steel it would also act as a heat sink, to some small extent

Next I drilled some vent holes in the metal - along both ends where the console end piece sit and a cluster beneath the centre of the unit. This is to encourage airflow.

It then occurred to me that I had a number of discarded fans from various PC's I've had or worked on. I fitted a case fan vertically on one side of the 360 so it would be blowing into the right had side of the Xbox, where there are small ventilation holes. The other end is partially obscured by the disk drive so not as good. I fixed the fan in place using a couple of cable ties daisy chained together and looped around the fan and joined up underneath the shelf. I ran the wires also through the hole. I found an old wall plug power supply from a child's night light which output around 9 volts. This is enough to supply the fan as these are designed to run off the 12 volt Molex connector inside a PC. By running the fan at a lower voltage you reduce the speed of rotation and hence the noise level, which is a popular trick in reducing PC noise levels.

I also found a smaller fan, the kind sometimes used on older 486 processors or Northbridge chipset coolers. This was thin enough to fit under the shelf, without sitting proud of the edge supports. If it did you could fit some deep rubber feet to raise the whole shelf up further. I glued this beneath the central cluster of holes I'd drilled, using 2 part epoxy glue.

I fitted a small electrical junction box and wired both fans in parallel to the supply. The case fan was whisper quiet but the small "whizzer" was very loud. I read a few posts on the web suggesting putting a 100 Ohm 3 watt resistor in series to drop the voltage further. I didn't have such a thing to hand but had a bunch (9) of 1.1K ohm 1/2 watt resistors which I soldered in parallel with each other giving me a 4 watt resistor of approx 100 Ohms resistance (See adding resistors in parallel ). I wired the composite resistor in series with the small whizzer fan. I tried fitting it in series with both fans but this caused the voltage to drop too low for the large fan to spin at all. In series with the smaller fan was just right. It turns at a slower quieter speed as does the side case fan.

the end result is :

1) I don't obstruct any of the existing cooling fans

2) I can run the Xbox on it's side, which is more space efficient for me

3) It's inaudible when running

4) It cost me nothing

5) It's not too conspicuous

NOTE :Never attempt this or anything similar unless you have a good understanding of electrical safety.

23rd December 2008

Well yesterday my Son reported that the Xbox360 was refusing to boot and the power supply brick displayed a solid red light. We spoke to Microsoft support , and tried a friends power supply, but with the same result. This indicates that while the bricks display the error, the problem is actually with the Xbox itself. The unit will be shipped back to Microsoft for repair. Actually their support were very good I thought, and the whole process is quite click. I had to pay this time as I'm outside the 3 year warranty, so charge was £60. It was a bad time to fail as my Son got Call of Duty 5 for Xmas for the 360, but it will be like a second Christmas when the unit comes back.

7th January 2009

Well the 360 came back from being fixed/replaced by Microsoft and is running again fine. I think it is probably a new base unit, I should have noted down some distinguishing ID numbers.

Dyson vacuum cleaners

2007-10-04T07:04:00.001+01:00

I have owned 2 Dyson Vacuum cleaners since 1991. The first was the DC01 model, which served for 7 or 8 years but eventually failed. The main problem with this model was the choice of plastic, far too brittle. So, the inner tool clip fell out from the wand, and the plastic bushes from the wheels disintegrated and had to be replaced with brass tube sections. The cleaner sometimes cut out, caused by a poor cable connection and had to be rewired. It basically vibrated its self to death over a number of years. This was replaced with a DC14, 4 years ago.

The DC14 is better made of a more robust and flexible plastic. It is also a more complex design, incorporating more cyclone chambers and removable sections for accessing blockages.

It does however exhibit a few problems from time to time, but they are easily solved. It's ironic that devices designed to clean things (vacuum cleaners, dish washers, washing machines) are often vulnerable to dirt or debris.

"Screaming" This is when the cleaner makes a shrieking sound . I have traced this usually to dirt deposits on or around the various seals which couple the sections together. Either on the rectangular rubber seal on the main dust collection drum, or the various removable sections. Just wipe any dirt deposits away with a damp cloth and reseat. Small residual accumulation of dirt or debris here causes a small air gap, through which air is sucked. This acts like the reed in a wood wind instrument and creates the sound.

Cleaner does not pick up when used upright. This is when there is suction through the wand but it does not pick up from floor or carpet when used as an upright cleaner. This may be caused by a blockage, but in my case is caused because the cleaning head unit is not correctly parallel with the floor. The floating section which contains the rotating brushes should always be parallel with the surface being cleaned, in order to maintain sufficient suction. If you take a floor level look at the left side of the cleaner, you will see that the head is tensioned down by the same white pipe section which provides the suction. This short length of tube , approximately 10 cm in length, is made of pipe with a bracing metal wire coil. This seems to act also as a spring in tensioning the head of the cleaner , so it's always pushed "heel down". If this does not happen try twisting the pipe until it retains it's springiness and pushes down the heel of the cleaning head

The Cleaner smells bad. I have only made this mistake once ! If you wash the top foam and plastic filter section under the tap, it's important that it's completely dry before you put it back in. If it's replaced while slightly damp, bacteria will develop in it and it will start to smell. Worse the very action of the cleaner air flow will exhaust this bad smell as you clean. Just wash thoroughly the filter and leave to dry in the Sun or near a radiator for a day.

Cleaner motor cuts out. This happened on Friday. If the motor over heats it has a thermal cut out. This shuts down the motor until it has cooled. The cause is usually an obstruction, either in any of the complex piping between cyclone chamber and the tools or header unit OR an obstruction on the output of air. In the DC14 the motor expels air upward directly beneath the cyclone chamber. On the allergy model a small round chamber holds a HEPA 'lifetime' filter so that expelled air is also filtered. The chamber holding the filter can carefully be opened by gently raising 3 plastic catches. Inside you can remove the purple filter. The underside of my HEPA filter , from a 4 year old DC14 is black, probably from microscopic pieces of carbon from the motor brushes. In the UK it is possible to buy another HEPA for around £25 and I may do that, as suction seems to be declining.

Loss of suction through wand and when used upright. Generally I have been noticing that there is less suction of late. I emptied the dust container, removed all the inserts and checked for blockages. All OK. The top yellow filter had been rinsed through quite recently, but I eventually tried the cleaner with the top filter removed. The cleaner was back to great form. This suggests to me that the yellow "lifetime" filters have a finite life ? The blue foam looks clean and it appears to be just sponge. However I'm not sure if the filter integral with the yellow piece lasts for ever ? Have ordered a replacement.

This seems to have given the cleaner a new lease of life. Perhaps I didn't need to replace the lower HEPA filter after all. However the top filter looks like it did need a replacement. The cleaner now sucks up fine. I can only assume that the yellow plastic top filter does not last forever as claimed ?

An introduction : mending things

2007-10-01T18:44:00.000+01:00

This is my first Blog . It will be dedicated to the maintenance of assorted household technology