5196106273

expect. Each one had a regular square-wave signal, with the frequency on each of the higher lines being half that of the one before. Obviously the Z-80 was cy* cling through all of its 65,536 addresses over and over again, trying valiantly to find an address that didn*t contain a NOP instruction codę. It was performing perfectly normally, at least in that respect.

So it was a matter of turni ng off the power again, removing the wire ground-ing all of the Z-80 data lines, and recon-necting them all back to the data bus. Then time for some morę thinking.

If the video RAM was OK, and the CPU was also OK, that seemed to sug-gest that the CPU wasn’t doing the right things because there was something wrong with the initialising routines in the monitor program. As this program was in the ROMs, along with the com-

Perhaps DSE would still have a technician who remembered the System 80 and could offer some help?

puter’s built-in BASIC language interpreter, this suggested that the fault • might lie in one of these ROMs.

Hmmm ... I didn’t have replace-ment ROMs to do any substitution, ei-ther. What to do?

At this stage I decided to leave it until first thing on the following Mon-day, and ring up the service people at Dick Smith Electronics. Their company hadn’t sold the System 80 for a couple of years, but perhaps they’d still have a technician who remembered the model and could offer some assistance. They might evcn be able to help with a loan of some replacement ROMs.

When I did ring on the Monday morning, they were indeed able to help. One of the technicians who had worked on the System 80 was still on the Staff, and as a special favour I was able to . havc a quick word with him. After lis-tening to the symptoms and the results of my sleuthing so far, he suggested that it could possibly be one of the ROMs as I thought, but in his opinion it was morę likely to be one of the main RAM chips. Apparently the monitor program used the main RAM to storę some of its working counters, and in the techni-cian’s expcrience the RAM chips were somewhat less reliable than the ROMs.

In cases like this, he’d gcnerally found the cause was a faulty RAM chip.

Just in case it might be the ROMs, he did have a set of them still tucked away 44 ELECTRONICS Australia, September

in the DSE service dcpartment, and was happy to make them available on loan for me to trv substitution. All I had to do was airange for them to be picked up. In the meantime, he suggested, why not try substituting for the RAMs?

That sounded like a good idea, so as soon as I had arranged for Mrs Service-man to cali in later in the day to pick up the ROMs, I decided to give it a try. Only to discover that .1 didn’t actually have any spare RAM chips — they were type 4116s (16k x 1 bit dynamie RAMs). Murphy’s Law strikes yet again!

Then I had one of those all-too-rare strokes of inspiration. The computer’s own expansion unit had two morę rows of the very same RAM chips in it, and they were almost certainly OK! The logical thing to do was open up the ex-pansion unit and use one of these to find the faulty chip in the Computer it-self.

I soon had the expansion box open, and discovered to my relief that its RAM chips were plugged into sockets, like those in the Computer. Whew! This was going to be easier than I thought — no messy soldering and unsoldering.

So I set about substituting one of the expansion unit chips for each one of the computer’s own chips, one by one, turn-ing on the Computer each time to check if the monitor screen would elear and give the normal sign-on message. (I had temporarily disconnected the expansion unit, because the fault was obviously in the Computer itself.)

The first chip substitution produced no improvement, so that one was prob-ably OK. I replaced the original chip, and tried substituting for the next one. Still nothing, so I proceeded to the next. Nothing again, so I tried the third — with the same result again. It didn’t look too promising.

But then, with the fourth chip, there was a dramatic change. When I turned the Computer on this time, most of the garbage cleared off the screen and a rough approximation of the normal sign-on message appeared. It was still a bit garbled though, and there were a few weird characters still sprinkled over the screen in places where they shouldn’t have been.

Hmmm — the RAM chip Pd just substituted for was obviously faulty, but could there be a second gone faulty as well? It didn’t seem too likely, but there was nothing to be lost by pulling out a second chip from the expansion unit, and using it to continue the substitution (with the first one in the fourth posi-tion, replaćing the known faulty one).

1987

Surę enough, it turned out that the sixth RAM chip was also faulty. Substituting for it suddenly brought every-thing back to normal, with a cleared screen and the normal sign-on message.

So I had found the source of the trou-ble, with a little help from the friendly DSE serviceman. The only thing was that by now, Mrs Serviceman was prob-ably at DSE, picking up the set of ROMs which were now not required. Hopefully if I could get a message through in time, she would be able to pick up a couple of 4116 RAM chips in-stead.

I madę quick cali to DSE, only to find — yes, you guessed it — she had already been there and gone! Some-times onc’s better half is a little too effi-cient. . .

Needless to say I had a certain amount of explaining to do when she dropped in at the shop, but she took it calmly and even volunteered to drop them back next morning and pick up the RAMs. So by lunchtime the next day, our trusty System 80 was back on the air again, looking after our spare parts and spitting out invoices.

All in all I didn’t feel too badlv about the time Pd taken to find the trouble, or the help Pd needed from the guy at DSE. After all, a normal “GP” service-man like myself doesn’t get much of a chance to get really familiar with per-sonal computers. Nor do we do enough of this work to justify investment in the specialised test gear needed to trouble-shoot them efficiently.

So considering my sketchy knowledge of these gadgets and the limited tools available, I don’t think I did too badly. Hopefully my experience may give you cncouragement to “have a go”, if you find yourself in a similar position. $

TETIA Fault of the Month

Hitachi CRP-141 (NPC6C-2 Chassis)

Symptom: Won’t start up, or starts but runs poorly with Iow B+ raił Cure: C919 (3.3/xF 25V electro) dropped in value. This capacitor forms a bypass on the supply to the chopper pre-drive transistor. Excess ripple at the emitter tends to cancel drive pulses at the collector.

This Information is supplied by courtesy of the Tasmanian branch of The Electronic Technicians' Insti-tute of Australia. Contributions should be sent to J. Lawler, 16 Adina St, Geilston Hay, Tas. 7015.