Items Covered This Month IBM 600Z notebook computer. NEC 2650CDT notebook computer. NEC N4853 (48cm) TV set. Metz Kreta VT Stereo 7949 Ch 687G TV set.

I don't get involved in computer repairs all that often. That's because, by the time a computer gets to me, it simply isn't worth repairing. However, that rule doesn't always apply, especially if the computers aren't your standard desktop models.

Which brings me to the notebook saga. In the last few weeks, I had not one but two notebook computers lob in with exactly the same fault: dead screens.

Well, to be truthful, they weren't entirely dead. I could see that a small portion of the left side of each of them was working but that was all. And the reason was clear - both had full-length cracks running top to bottom on the LCD panel. Once that happens to the LCD, it's history.

As with (I believe) pretty-well all notebooks, it's easy to plug in a standard VGA monitor and operate them that way. But while this might prove that everything else is still kosher, lugging around a monitor somewhat defeats the purpose of owning a notebook!

Both machines were of the early Pentium (well, early compared to today's speedhogs) genus. One was an IBM 600Z, a 300MHz machine, while the other was an NEC 2650CDT, a 166MHz model. The question was, with the price of notebooks haven fallen so far in the last couple of years, would these be an economic repair?

The IBM belonged to Mrs Brien. Or, to be more exact, it belonged to the organisation for whom she was the voluntary secretary. It's an organisation I'm also involved with, so this repair was to be basically a love job.

I think she was at least a little embarrassed at having broken the organisation's toy (by leaning on the screen) and wanted to fix it herself, rather than going through channels. She had already been to a repair agent and nearly died at the quote for the repair.

Swallowing her pride, she came to me almost in desperation; could I do it cheaper, on the quiet? She would pay for it herself, rather than facing the wrath of the President and the committee. She was willing to spend a modest amount, say $300 or so, rather than fork out the almost $1000 she'd been quoted.

The second machine, the NEC, came to me as "payment" for another job. Of course, I would have preferred coin of the realm but without wanting to go into the circumstances, I figured I was lucky to get the notebook, cracked screen notwithstanding.

It was fairly well tricked-up for a machine of its type - extra memory, a selection of PC cards (once known as PCMCIA cards) and even some original software, with manuals.

I reckon the NEC would have set the original owner back at least $5000, if not more. If I could get it working for a modest sum, I would have a nice little machine to play with. Maybe a bit slow by today's standards, but nice.

As I mentioned, Mrs Brien had already had a quote for repairing the IBM. I thought I might as well try the same thing with the NEC.

No go, from the genuine sources. They didn't keep new screens for a notebook this old (1998, mind you!). What about secondhand? "We have them from time to time but haven't seen a 2650 screen in many months." "Any ideas," I asked? "They're not worth repairing. Why don't you buy a new machine . . ."

It wasn't exactly the news I wanted to hear.

My next thought was to try the 'net. There are several places, especially in the USA, that specialise in notebook screens, mainly secondhand, through dismantling. They at least came with a guarantee even if they were on the other side of the planet.

Late one night, when Mrs Serviceman had finished playing Solitaire, I logged on to the 'net and typed "notebook screens" into Google (what a great search engine!).

Sure enough, at least a dozen possible sources emerged - as expected, all in the USA. But the prices they were asking were something else again: $US300, $US400 and more - and remember, these had to be doubled for our little Aussie bleeder. And then I had to get them here and in one piece.

Scratch that idea. Neither Mrs Brien nor I were willing to outlay that amount of money.

What next, then? Did we both now own basically fairly useless (ie, old, slow) pseudo-desktop computers? And did I have to give Mrs Brien the news that she'd have to front El President'e and 'fess up?

What about Ebay?

And then I had another brainwave. What about Ebay? For the uninitiated, Ebay is the world's largest on-line auction and has an Australian site. Maybe I could pick up a couple of cheap, preferably dud, notebooks which still had intact screens?

So I logged onto Ebay and . . . nothing! Searching for the particular model numbers yielded not a skerrick, even though there were notebooks-a-plenty. I called Mrs Brien and gave her the two lots of bad news. She took it very well and said that she appreciated my efforts - and to hang on to it, if I was willing to persevere . . . something might come up.

Well, something did come up. A couple of days later I was on Ebay again and there was an organisation offering refurbished IBM notebooks - not Mrs Brien's model number but close to it. And the bids were reasonably low (around $300), even with just a few hours to go.

This was one auction where the seller was happy to identify themselves - and even give a phone number. So I rang the company, (Cost Plus Computers in Melbourne) and asked for their technical department.

The lady asked me what I needed and I told her I was willing to buy one of the IBM computers on Ebay if the screen was compatible with the 600Z. "No," she said, "they aren't".

As they say in the Toyota advert, "bugger!"

But then she said "hang on a sec, I'll check with our tech - we might have screens to suit." Just a few seconds later she was back. "Yes, we have some of those screens. $270 including postage. Did I want one?".

It arrived the very next day, very securely packaged and protected. Fitting the screen was almost an anticlimax - about 10 minutes very simple work. I had read a number of warnings on the US websites about the potential (pardon the pun) for a nasty, possibly lethal, belt from the fluorescent tube inverter used in these machines (even when turned off) so I was very wary to use my best HT practices when handling this area (one hand only, rubber mat, insulated tools, etc, etc).

But it was basically just a matter of undoing a few screws, unplugging the old screen and removing it, putting in the new and doing up the screws again.

Fingers crossed, I turned the computer on and. . . Bewdy! I rang Mrs Brien and she was around in a flash!

Now for the NEC

OK, that was the IBM. How about the NEC. A few days later, I again logged onto Ebay. And sure enough, no 2650's - but there was a "case and LCD screen, no mobo (motherboard)" being sold from, wait for it, Darwin! I checked NEC's website: the one being offered was a newer, faster machine but otherwise had fairly similar specs to my 2650, same screen type (TFT), resolution (800 x 600) and size . . . dare I risk it? Hey, what did I have to lose?

I rationalised to myself that if it didn't work, I'd put the whole lot back on Ebay and hopefully recover most, if not all, of my costs. So I bid for the lot and, after a short bidding war with someone else who obviously wanted it almost as much as me, eventually bought it for the princely sum of $108.00. If this worked, I was laughing!

After a day or so, the lady in Darwin who was selling the lot emailed me with her details. I rang her and arranged a direct deposit, along with instructions to package it up very carefully. I also arranged through a mate who works for a freight company for it to be picked up.

It arrived less than 24 hours later -not bad from Darwin, I thought. Hopefully, everything else would go as smoothly.

Rule No.1: don't tempt Murphy.

Of course, it didn't go smoothly. Getting the NEC apart proved to be a lot more difficult than the IBM. There is obviously a right way and a wrong way to do it and I chose what must have been the wrong way. There are hidden screws, there are keyways, and there are various bits to remove in the right order. I made a mental note of these as I disassembled it.

Finally, I was at the stage where I could unscrew and unplug the cracked display. I compared this and the new one and my initial observations were positive. The plugs and sockets on both were identical, as was the overall size. That's a good sign. The bad sign was that the mounting screw locations were different - not much mind you, but different enough.

Even so, I managed to get a couple of screws into the mounting holes to hold it (just enough!) in place. There was little point in reassembling the case if it didn't work so I rechecked all the connections and gingerly turned it on.

Oh, happy day! I have never been so glad to see the Windows logon, complete and in beautiful living colour! I checked the basic operation of the machine and then turned it off, ready for final reassembly.

Of course, my mental notes were now nowhere to be found (perhaps I had powered down my brain without hitting <control-s> first?) but I did finally work out which bit went where.

One problem was that the screen was in a slightly different position and the left edge was hidden by the front of the case. So I had to move the display over a few millimetres (not easy!). Another problem was that I must have damaged some of the retaining lugs getting the front off the display because it no longer stayed locked to the back.

I hate to admit to such a blue-and-white striped apron approach but a tiny dollop of super-glue cured that problem (I hope I don't need to get it apart again!).

In use, about the only difference I've noticed between the old and new screens is a lot less brightness control. But where it's set is more than adequate for me. And just to prove the point, I've written this on it!

So with a bit of perseverance, a bit of a gamble and finally (for me!) some good luck, I was able to get both notebooks working again for not too much money.

Mrs Lawrence's NEC TV set

Mrs Lawrence was having problems with her 8-year old NEC TV set. This was a 1994 model N4853 (48cm) set which used a Thai-built MM-1 chassis and PWC-3850A PC board. She complained initially that the colour went all pink, then the set intermittently wouldn't start and now it was finally dead.

I advised her that it was better - not to mention cheaper - to bring the set to the workshop rather than vice versa. Alternatively, I could pick it up for a modest fee but she opted to get her son to drop it in.

When I removed the back on the workshop bench, I really couldn't imagine that it was going to be very difficult to fix. The set was, to all intents and purposes, dead but the switchmode power supply was working and supplying healthy voltages on both the high and low output rails. I didn't have a service manual but I did have a poorly photocopied circuit, with no voltages marked on it. Based on similar circuits, I guessed that the two voltage rails should be at about 115V and 20V respectively and this turned out to be pretty well spot on.

There was 115V on the collector of the horizontal output transistor (Q902) and also on the horizontal driver Q901. And that meant that there was no horizontal drive from pin 37 of the jungle IC201 (TDA8362).

I checked the crystal for faulty joints before measuring the supply voltage to pin 34 of IC201. This showed that there was no worthwhile voltage to power up the oscillator. Normally, the supply voltage is derived from Q106's emitter and this transistor has its collector connected to the 20V rail via R194. Q104's base is controlled by Q105 and pin 31 of IC101 (an M37210M3 microprocessor), to power the set up from stand-by. The problem was that Q104 wasn't being controlled.

I checked and replaced Q2001 and Q2002 as a matter of course. They are part of the EHT/x-ray protection circuit from pin 2 of the horizontal output transformer and they can give trouble. Microprocessor faults can be ticklish and expensive to fix, so the trick is to concentrate on the easy and cheap solutions first.

I started by measuring the +5V rail (Vcc) to pin 27, which was correct. I then measured the reset voltage at pin 30, which was very low. This is derived from the collector of Q104 via R192. There was 8V applied to the emitter of Q104 but only 4V or so at the collector. Similarly, the base of Q104 was low.

This looked straightforward. It could really only be the transistor, R191, zener diode ZD101 (7.5V) or a load that was causing these voltages to be so low. Sometimes it's just quicker to replace parts rather than measure them so I changed the transistor and R191 (47kΩ) but neither made any difference.

I didn't have a 7.5V zener diode in stock for ZD101, so I thought I'd give it the best test I could devise. I removed it and connected it to my power supply, which is current and voltage controlled. With the cathode connected to the positive and the current set to nearly minimum (this is only a 0.5W device), I turned the voltage up. When the current meter was just on the verge of reading, I checked the voltage - it was spot on at 7.5V. In my book, that meant that the zener was working properly. The only problem was that back in the real world (ie, in the circuit), it wasn't.

I resolved the problem a day later when my stock of 7.5V zeners was replenished. I popped in a new one and this part of the job was fixed.

Now I had a picture but it was indeed pink, which meant there was no green. I examined the CRT socket board and could see it was full of doubtful joints. Resoldering R908 restored the missing primary colours.

I put it aside to soak test and noticed that, after a while, the purity was wrong. I degaussed the screen and put it down to someone carrying a magnet near the set. The next day I switched the set on and it went "bang" in an embarrassingly loud way. Back it went onto the bench.

One problem with this set is when the back is removed, the PC board sticks to the rear half of the shell but the leads to the speaker and degaussing coils, etc are still connected to the front half of the shell and the tube, causing stress on the connections.

In this case, though, the set was dead because the dual posistor (PTC) had died and taken the fuse with it. A new one soon fixed this but after I had put the back on again and tried it, there was even more trouble. There was now no sound, so it all had to come out again. The ridiculously short loudspeaker leads had finally been pulled too hard and broken in two places, one at the loudspeaker and one at the PC board socket connection.

Change in attitudes

Readers who have been reading this column for a long time will have noticed how customer's attitudes have changed over the years.

Technology has become so complex and yet, at the same time, so cheap. This double whammy really means we no longer have the luxury of fitting an exact cause to an effect; one just has to accept that this is just the way it is and move on. Customers are now more demanding than ever before, expecting more for their money and being fully aware of their rights.

In the black and white TV days, we would have been happy if a TV set had only broken down three times a year. Now we are bitterly disappointed if a colour TV set breaks down once in three years and this is despite it being at least a fourfold leap in complexity, technology and performance - all at a fraction of the cost of black and white TV sets.

Repair methods are changing too, with some of the emphasis now on software monitoring - either that, or immediate disposal. Very few new VCRs are worth repairing now, the only exceptions being hifi and Super VHS models. Many DVD players are also uneconomical to repair and small 34cm TV sets have become "throw-away" items.

More importantly, installation of new equipment is becoming a growth area. New equipment generally is now so much more complex to install and the setting up methods can vary widely from one brand to another. Indeed, some equipment can take an enormous amount of time to set up, often due to its complexity but sometimes because the setting-up method is cumbersome.

For example, to skip all the unused program sites on some Panasonic VCRs, it can take up to four separate commands to navigate through 100 programs. Many people just opt for the auto installation, which invariably means the channels don't correspond with the numbers on the remote control.

Another drama is digital TV co-channel interference with the VCR and working out how to turn off the RF output (if available) and use only AV leads.

I heard of a case recently involving the installation a European TV set. To do this, it is necessary to get into a menu, and the first question is, naturally, "What language do you want to read?" After selecting "English", it asks which country you reside in. Easy, you say? The problem was that Australia (CCIR system B/G) wasn't offered as an option and selecting either Germany or Britain didn't help much.

In the end, the technician used a variety of different countries to tune in the Band I, Band III and UHF channels. He did this on a channel-by-channel basis until he had them all in. Madagascar was used for Australia's channel 2!

I suspect there was probably an easier way but he reckoned he couldn't find it in either the instruction book or the on-screen help. And have you noticed how thick some of these instruction books are now? There just isn't time to read through one on a home service call - but the customer expects me to know everything about it!

Metz stereo TV set

Getting back to the bench, I had an interesting set brought in recently. It was a Metz, made in Germany. More exactly, it was a Metz Kreta VT Stereo 7949 Ch 687G with a 66cm tube - about 10 years old, I think. It was one of the first of this brand I had seen and it came in with the complaint of no picture.

The fault wasn't difficult to diagnose. The tripler had been arcing and had died. The replacement took a while to get from overseas and it wasn't cheap but I thought that would be that. As usual, it wasn't - I still had no picture.

First, I checked the EHT. This was OK at 29.5kV and I could see that the CRT filaments were alight (I measured 6.3V across them). The screen voltage (VG2) varied from 400-1000V, the HT was correct at +163V and there was 210V applied to the CRT base. In addition, there was +12V on the first grid of the tube.

I then tried my dangerous and totally inadvisable trick of momentarily shorting a CRT cathode (any colour) to chassis and observing the screen. (This is particularly dangerous on a set like this with complex push-pull automatic greyscale tracking video amplifiers.) Despite this, the result on the screen brought good and bad news. It was good news because the screen was full and the vertical deflection was working. On the other hand, it was bad news, because it was a dull greyish raster no matter what colour I shorted.

I was expecting an intensely coloured raster at least.

The only thing I hadn't checked was the CRT focus circuit where I would expect the focus potential to be about 4.5kV. Checking this was difficult because the 4.5kV cable from the tripler is very effectively insulated right up to and over the focus pin base on the socket. I overcame this by using a single strand of very fine gauge wire which I managed to wrap around the focus pin on the CRT, before replacing the CRT socket. I attached the other end to my EHT meter.

"Dodgy" is the best word to describe all this. What's more, my home-made EHT meter is 10μA fsd (1kΩ resistance) and is fed via 2 x 130MΩ resistors, so it would still create some loading on this supply. Other problems involved insulation, sparking and residual charges. Anyway, my makeshift "focus meter" read zero volts.

I disconnected the tripler wire from the CRT focus pin and connected the EHT meter to this lead. This time I was cooking with gas - now I did have 4.5kV. So how was it being pulled down when it was connected to the tube? With the socket off the tube neck, I checked the resistance between the focus pin and chassis. It measured completely open circuit with my best ohmmeter on the x100kΩ/volt range. But this was a low voltage test and did not necessarily tell the true story.

Normally, I would have simply fitted another CRT socket but this one was unusual and quite small; smaller than anything likely to be available locally. And the customer was getting tetchy about the time since the job started, so ordering a new one from Germany wasn't really an option.

Using a pair of sidecutters, I hacked my way into the focus spark gap housing to have a sticky beak. Fortunately, this revealed that all the inside was a mass of black carbon - it was like a bomb had exploded inside. Obviously, it had been sparking severely. I cleaned out the carbon as best I could, removed the chassis half of the spark gap, filled it all with silicone rubber and let it set.

Late the next day, I switched the set on and the picture, after slight adjustments, was now perfect. My theory, for what it is worth, is that an insect probably crawled into the spark gap and caused the sparking. Gradually, as it became carbonised, the current drawn became excessive, causing the tripler to overheat, crack its insulation and finally fail.