Not Quite Toast [Dell laptop]

OK, here's a fine tale. Believe it or not - this story, and the accompanying photos, are absolutely true and unaltered.

In another life, I was an I.T. Support engineer in an academic institution. Although it wasn't strictly part of the job, I would often undertake minor repairs to the staff kit, such as mending frayed power cables or fixing broken connectors. There are quite a few tales there too, but I digress.

One day, a staff member came into the Helpdesk office bearing a carrier bag and a very sheepish expression. He had, he said, had an accident with his laptop, and it wasn't working, and did I think I could fix it? He opened the bag to reveal this mess:

After I stopped laughing (cruel, I know) I had to inform him that there was a chance that it might be beyond even my legendary abilities to resurrect this unfortunate item, and that he should bid a sorrowful farewell to his files and seek laptops new. This suggestion was accepted with a degree of rueful acknowledgement that in truth, he hadn't held out much hope anyway.

The full sequence of events was related. He had needed to go out of the house for a short time, and being conscious of the possibility of burglary he had looked around for a suitable place to conceal the laptop in his absence. As he was in the kitchen, the oven seemed to be a place that any thief would never look for a laptop, so in it went. Shortly after that, his wife came into the kitchen and proceeded to prepare the dinner, consisting in part of a nice apple crumble (you can see where this is heading). I suppose a black laptop does look a bit like a baking tray, but anyway for whatever reason she didn't realise the situation and switched on the oven to heat up while she mixed. After a while she noticed a distinctly un-crumbly smell and opened the oven....

Credit where it's due. Although the oven temperature clearly exceeded the melting point of most of the plastic components and turned them into a Dali-esque sculpture, this Dell has an alloy frame and top and bottom covers - these were surprisingly unaffected but the rest was, well, toast. The screen had melted and all the keys had evaporated from the keyboard, and everything was in a sorry state. I took the above photos just for the record because it's a sight you don't see often (although with Lithium Ion batteries becoming ubiquitous in cheap kit and charging not always sensibly controlled, we may see more incendiary incidents).

My colleague having gone off to throw himself under a handy train, I put the remains on a shelf and got on with something else.

Some time later I was poking at the pile of ash and I began to wonder what would happen if I connected a power supply to it. "Nothing" was the most I expected, although "lots of smoke and sparks" was also a possibility. Anyway, I went ahead and with the fire extinguisher close at hand, I connected the power unit and switched on. To my utter amazement, this is what I saw...
 

Phoenix or what? Way to go, Dell!

 Not only did it not spark or catch fire, it actually booted as far as the command line and (rather superfluously) informed me that it couldn't see the hard drive, it couldn't understand the power supply parameters and it might not feel like charging the battery. It told me to strike the F1 key to continue but as you can see, there no longer was any F1 key. In fact if the key switches had not melted, it may have got further, but in the event I thought that it had already gone far beyond the call of duty, so I switched off.

The upshot of all this was that I chiselled out the hard drive and found that it was still operational, and I was able to recover my colleague's data for him. He was overjoyed and promised to tell Dell how impressed he was with the durability of their product, and he took away the bag of ashes for a decent burial.

Actually, it was a bit sad. This sturdy little machine was nobly doing its job and didn't ask, or deserve, to be baked to a crisp. I thought it was an effort worthy of high commendation that, with its last breath, it mustered enough life energy to let us know that it wasn't quite beaten. If it had not managed that, I wouldn't have attempted to remove the hard drive and try recovering the data, and all my colleague's work would have been lost.

Rest in peace, little champion.

Simply Sony [Sony TC-FX25]

Today's special is a Sony TC-FX25 stereo cassette deck with no tape movement. The belts were all intact but had sat unused for so long that they had set into oval crispy things, so that was the first port of call.

This is a very straightforward deck - single playing direction, logic controls, manual selection of tape type and Dolby, LED bar graph recording level indicators and those horrid slide pots which were so popular at the time. You need to take off the top case (just a few obvious screws) and the bottom plate, which is held by three screws along the back edge and hooks under the front panel. The tape carrier door slides up and off with the door open. The front panel is clipped into place and just comes off forwards, and you have to disconnect a couple of plugs on the main circuit board to free it by threading the wiring loom out through the chassis. You can re-connect them directly to the PCB and the machine can still be operated, but note that on this old model, the mains transformer is fully exposed with tags carrying mains voltage on the top - I made up a cardboard box to cover the terminals while I was working on it.

To remove the transport mechanism there are four front-facing screws around its edge - coloured dark to distinguish them from any others. I needed to undo a few cable ties before the deck could be taken out, then disconnected some plugs on the main PCB - there are several similar but they are all different numbers of ways, so although I did make a map first, it's difficult to go wrong anyway. I guess it might help any viewer if I were to put in a scan of things like scribble maps, but I'm pushed for time at the moment. If anyone wants more info just stick a comment in.

The back plate with motor comes off easily allowing access to the first two of the four belts in this model. Again, without access to the exact replacement parts you have to do a bit of substitution. I found that the capstan belt was flat section, and CPC Farnell had a suitable one in their AVBELT5 (71mm dia. by 2.8mm wide, 0.5mm thick). The fast-forward belt driving the spools is replaced by an AVBELT76 (37mm dia. by 1.2mm square section). Then the back plate can be re-fitted.

The other two belts are accessed by removing the front plate and tape carrier assembly. This involves taking out a couple of brass screws each side which pivot the stays and I seem to recall there were a couple more at the top or side fixing the plate to the deck frame - but it was all pretty obvious. Once the screws are out, the assembly falls out forwards. It's probably worth keeping it in about the same position as it comes out, to save getting in a 'folding deckchair' pickle when re-fitting. At this point I was able to fit the other two belts - AVBELT69 (32mm dia. by 1.2mm square section) driving the take-up spool carrier, and the tape counter belt AVBELT100 (64mm dia. by 1.2mm square section) from the take-up spool spindle to the counter. As the tape counter is mechanical and still mounted on the main chassis, I left the drive belt loose having threaded it through the metalwork and then I re-fitted the front plate and tape carrier assembly. Throughout all this I made sure that I cleaned and lubricated the heads, pinch wheels and bearings as usual.

After putting it all back together, the tape mechanism worked perfectly with no slipping or wobble. Unfortunately there was only one channel playing and touching the PCB or wiring caused much crackling. Eventually I traced the problem to the tape head connectors - there seemed to be no dry solder joints so I used contact cleaner (Electrolube) on all the connectors, making sure to work them off and on to scrape off any tarnish. This sorted it out and the whole machine is now performing well.

This machine was hi-fi for its day but certainly not anything special, and even fully working it can't outperform any decent bit of kit - it's very basic. Added to that, it's not pristine - one or two rusty scratches - and uses cassettes which are hardly the medium of the future any more. But it has two standard (quarter inch) microphone inputs on the front and is still good for basic recording or transcribing old tapes, so maybe it will find a use somewhere. Anyway, it's satisfying to see it back in working order.

Archaeology revisited [Aiwa AD-R450]

As I now have some of the drive belts I need, I went back to my favourite cheapo cassette deck, the Aiwa AD-R450 which I was looking at last Sunday. It only took a couple of minutes to take off the backplate of the tape transport - a couple of screws and the plate was free to move out enough to thread in the flat drive belt. I had obtained service info and I was right about the belt path - from the top of the motor pulley, horizontally over to the RH capstan flywheel and down and around, up between the two flywheels and back to the bottom of the motor pulley. This drives the RH flywheel anticlockwise and the LH flywheel clockwise for the forward and reverse capstan drive.

 At this point I refitted the backplate and tested the drive - all was working now except the mechanical tape counter. To replace the counter belt, I had only to remove the four screws securing the whole transport mechanism  - two into plastic pillars right at the top of the front panel, and one each side into the base of the bottom tray (from inside) and pull back the machinery. Another two minutes to loop the new belt around the back of the take-up hub onto its integral pulley, then put back the mechanism and secure it and stretch the counter belt over the tape counter pulley. While the assembly was out, I took advantage of the easy access to clean up the heads, capstans and pinch wheels with some methylated spirit (wood alcohol to our colonial cousins).

This restored everything to full working order, and I had my old hi-fi deck back again.

The belts I used were from CPC Farnell, AVBELT9 (flat, 81mm dia., 3.5mm wide by 0.6mm) and AVBELT98 (61mm dia., 1.2mm square section). I had intended to try AVBELT10 (88mm dia., 4mm wide by 0.3mm) but this was out of stock. The original capstan belt was about 86mm dia, 5mm wide and 0.5mm thick - it's difficult to accurately measure a length of soft liquorice.....

The only comment I'd make was that the AVBELT98 was slightly too long and although it drives OK, will probably loosen fairly quickly. A better choice would have been the AVBELT96 (55.6mm dia.) or AVBELT95 (54mm dia.).

CPC's website is one of the worst for finding items. They have an astonishing range of stuff, and usually good prices, but the organisation and search is rubbish. For example, to get a drive belt you can eventually find your way through the menus in the sidebar (> Components, Kits & Spares > Spares (Brown & White Goods) > Audio Spares > Belts > Square Section) but then you're presented with a filter matrix. If you select just one wrong box, you will get wrong answers. By selecting the minimum of filters, you can get a matching list of bits (belts, in this case), but they are in apparent random order. They show part numbers, prices and the Length - by which in this case they mean DIAMETER - and the depth - by which they mean WIDTH. If you click 'More details' the actual item page shows the price and hopefully an accurate picture (but sometimes generic) but no dimensions, If you select the obscure "Further Information" link from there, you get a list of EVERY belt with dimensions and a BUY NOW link, but no prices! When searching for a number of possible alternatives, as you do with belts, it's a matter of having a very good memory and the patience of a saint. Many times I have tried to find something in CPC stock and given up, only to find out later that the item was available but classified as something else. And the filter options rarely filter the things you want to filter.

Another point to note with CPC is that they very kindly send you sales flyers - floods of them! - listing the current special deals. The catch here is that the last two digits of the part number are specific to the offer catalogue only, and if you try to get something from that catalogue after the offer period has expired it will show a 'Not Found' error. Disregard the last two digits and search again, and you will probably find the item still in stock but at a different price. You may be happy to buy at that price as it may be better than thinking you aren't going to get it at all. Once I tried to be an arrogant smartarse and thought I would remove the last digits because I didn't like my buying habits being logged (from a given catalogue). Hoist by my own petard, I was charged the full off-sale price for everything. And serves me right.

Not too hard [Aiwa AD-M100K]

Yesterday I turned to resurrecting the Aiwa AD-M100K stereo cassette deck. This is quite a nice straightforward machine but the design is very much of its day (1980) and showing its age. It has mechanical tape loading via piano keys, single direction transport, analogue VU meters and no frills. But it was typical of the kind of stuff I was using at the time.

This one had toffee drive belts and although the motor hummed, that was all. I couldn't find any service info but the nice people at Vintage Cassette have a spec sheet here.

After lifting the lid I did some measuring using the tested method of wrapping ribbon round the belt path and overlapping it, then cutting both ends together. This gives the length but be careful, some suppliers quote "length" meaning the belt laid out in a line but still intact - so their "length" is simply half of the measurement. Other suppliers use the diameter of the belt in a circle as "length", including CPC Farnell who complicate matters by referring to the height and width of the belt, meaning the width and thickness as I would call it. But hey ho, it keeps the brain from coagulating I suppose.

This deck needs four belts, which I got from CPC as follows:

Capstan belt: AVBELT4 - flat, 69mm dia, 3.5mm wide, 0.5mm thick
Auto-stop drive: AVBELT78 - 38mm dia, 1.2mm square section
Tape counter primary: AVBELT92 - 52mm dia, 1.2mm square section
Tape counter secondary: AVBELT66 - 28mm dia, 1.2mm square section

These are just what I could find which worked - there may be others. You can pay a lot of money for a dedicated set of belts specifically for this machine, which might just be these ones re-packaged with a fiendish markup. I notice the 'set' suppliers never list the belt sizes in their kits!

The tape transport mechanism can stay in place - I just removed two screws (mid left and bottom right, out of direct sight) which allowed the backplate to hinge away and give space to thread the capstan belt through, followed by the auto-stop belt which runs off a coaxial pulley on the flywheel. I then re-fixed the backplate and fiddled for ages to get the two belts around the flywheel rim/motor pulley and the flywheel pulley/autostop pulley respectively. Then check that it all worked - success! Pretty well everything on these is done mechanically so once the drive is back, only the tape counter didn't work. This could have been tricky but there is enough space around the take-up spool hub, where it passes through the front plate, to thread the belt in. You then catch it from behind and stretch it up to the idler pulley top right, and then add the last belt from the idler to the tape counter.

That was all there was to it this time - I spent the next hour or more listening to Pavarotti's Greatest Hits (Decca K236K 22, 1980) which was very appropriate, just the kind of thing I might have played on it in 1980. It works beautifully. While I listened I put together a full description for the blog, peppered with snappy witticisms and useful tips. Then when I went to publish it, there was a 'Save error' and Firefox crashed, losing the lot. That's why this account is a bit boring, coz I can't be bothered to do it all again.

(Postscript: I found that using copy & paste doesn't work in the blog for inserting images. The image appears but the uploading during Save or Publish gets a bit sniffy with the Javascript...)

Archaeology [Aiwa AD-R450]

Archaeology. The process of digging up old things and trying to understand them. In this case, my old Aiwa AD-R450 cassette deck, which was an excellent machine until one day it stopped responding to some of the panel controls - apparently a logic fault in the soft-touch circuits. I got another deck but, as is my way, I archived the old one until I could find the time to look at it.




The big feature of this one is the "Quick Reverse" function. The erase and rec/play heads are mounted on a tiny turntable and there are two capstans and pinch wheels. Reversing the tape direction involved engaging the other capstan and flipping the turntable round, which could all be done in 0.2 seconds thus enabling playback or recording both sides of the tape without an audible gap.

The picture above is courtesy of www.radiomuseum.org, a mine of information.

Unfortunately this unit has been in storage for around 20 years, so the belts are pretty well liquid. In fact I found only a small portion of the tape counter drive belt and the main capstan drive belt was all wound around the motor pulley, like self-amalgamating tape, and it had melted into an impenetrable tyre of rubber. I had to cut it off so the only info I have is that it's 5mm wide and is made of the same stuff they used for the belts in the AD-F770 and AD-F660, which also have liquefied (both of these being repaired currently).

The logic seems to be doing its stuff, but it's clear that the deck function depends on mechanical drive to move the heads and drive the tape. The reel motor is separate so the tape will fast wind OK but it needs a new main belt. I have no service info on this as yet, but I think the belt probably wrapped around the > flywheel and drove the < flywheel off its back, so that they revolved in opposite directions. (The motor only goes one way no matter which direction is selected). So the next step is to confirm this, if possible, then measure the path length of the belt and find a replacement. I also have to work out where the other end of the tape counter belt was driven from and find a new one of those; then I can think about dismantling the mechanism and fitting them. After all that I may have a working cassette deck (which nobody wants!). But I like seeing things working again.

 I suspect that when I first encountered the fault the belt was loose, and may have driven one flywheel but not the other. This may have given me the impression that the logic was wrong, as the deck would have run in one direction but not the other, and the heads wouldn't flip. If I could see the motor running and the flywheel turning I would have thought it was OK and the fault must be in the logic circuits; I knew less then about how much functionality still depended on the mechanical motor power in the designs of the day.

And I've got a horrid cold and cough today.