• Please review our updated Terms and Rules here

Another poorly Epson PX-8

Crashedfiesta

Crashedfiesta

Experienced Member
Joined
Apr 16, 2022
Messages
139
I got my PX-8 out yesterday for a quick play and check to make sure it was still working.

The display was not. It appeared to have a blank screen but after about 10-15 minutes of being switched on, the screen suddenly appeared.

Then I checked the tape drive and that worked fine. But I did notice that operations caused the screen to fade and then darken depending on what the motor was doing.

And today, after I gathered all the tapes I have to have a look through, the screen went blank and now stays blank.

I've managed to replace C5 and C6 on the screen PCB but it hasn't really made any difference. If I angle the screen and shine a bright light on it I can just about see text but even that seems to be fading in and out randomly.

Is this more likely to be power rail related? Last time I had it apart I did a partial recap but I wonder if I should do a full recap now....
 
Hi Crashedfiesta,

My first guess based on the earlier screen behaviour is a problematic battery pack or battery connector. You could try a power supply ~ 5V DC connected directly to the main PCB instead of the battery pack and see if it improves things.

Greetings,
Fred Jan

P.S. I do have one more spare screen, but at least one track of the flexible cable is damaged...
 
Tried the bench supply connected to the board and it made no difference. It's actually slightly worse now as I can't see anything on the screen. The caps lock light does still light up if I press it but I'm not sure if that would happen anyway with power applied..

I'm gutted. I hadn't used it for a while and it's just dead now. The only thing I haven't done is tried letting it boot up, press escape and then mount H: which would trigger the tape drive. If I did that and the tape then rewound it would at least show the main unit is running...

And then if I documented my UD80 correctly in my blog and follow those instructions carefully I might be able to get an output on my TV (if I can find all the cables 🙄).

The joys of vintage computing...
 
So I've tried all the usual resets and still nothing from the screen. I have a theory, however...

In the box of stuff I had with this machine there were several power supplies, all labelled 'Epson' and all at 6v. The one I selected on Sunday was a bigger unit than the one I'd used previously (which I couldn't find). It worked fine after a while - the screen was initially blank but then came back after a few minutes - and, a possible clue, the tape drive worked perfectly which it had not done before.

I've just checked the voltage and it came out as 9.6v but as the supply is unregulated I assumed that it would drop under load to 5v. But measuring the voltage across the board and, more importantly, to the LCD it was actually giving 6.4v. I'm assuming 6.4v into a 30 year old LCD screen when it's expecting 5v is not going to do it any good. But what really puzzles me is, if I use the smaller PSU which I used before, the cassette doesn't work. If I'm lucky, it will just rewind a little bit. But if I plug the bigger supply in I can wait for it to boot, press ESC to get to a command prompt and if I type 'H:' then the cassette rewinds, stops and finds the index of the tape exactly as expected. So it seems like the cassette drive is more than happy to take that extra voltage.

In short, I think I've zapped the screen with too much voltage. :(
 
Crashedfiesta said:
So I've tried all the usual resets and still nothing from the screen. I have a theory, however...

In the box of stuff I had with this machine there were several power supplies, all labelled 'Epson' and all at 6v. The one I selected on Sunday was a bigger unit than the one I'd used previously (which I couldn't find). It worked fine after a while - the screen was initially blank but then came back after a few minutes - and, a possible clue, the tape drive worked perfectly which it had not done before.

I've just checked the voltage and it came out as 9.6v but as the supply is unregulated I assumed that it would drop under load to 5v. But measuring the voltage across the board and, more importantly, to the LCD it was actually giving 6.4v. I'm assuming 6.4v into a 30 year old LCD screen when it's expecting 5v is not going to do it any good. But what really puzzles me is, if I use the smaller PSU which I used before, the cassette doesn't work. If I'm lucky, it will just rewind a little bit. But if I plug the bigger supply in I can wait for it to boot, press ESC to get to a command prompt and if I type 'H:' then the cassette rewinds, stops and finds the index of the tape exactly as expected. So it seems like the cassette drive is more than happy to take that extra voltage.

In short, I think I've zapped the screen with too much voltage. :(
Click to expand...
The screen should get about 20v with +15 and -5, making the 20v difference. The overvoltage protection kicks in around 7.5v so I'm sure that the power supply voltage wouldn't be causing an issue. I was seeing 6.75v from my power supply when it's connected under load and same as you about 9v unloaded.

Also the serial port gets its 8v by deriving it from the LCD voltage circuit, so if that's not getting enough, your serial port won't get enough either - but if you can at least get TTL voltages out of it, you'll probably be fine.
 
So it looks like the screen is not faulty. And I'm fairly certain there's an issue in the voltage circuit that provides the VL on pins 15 and 16 of the LCD connector.

I spent quite a bit of time reflowing and cleaning around the circuit that provides the VL signal and, when I turned it on, the screen briefly lit up again and the VL was around -7v. This seems a bit low (high?) but the screen was definitely on. After a few seconds the voltage rose to about +200mv and the screen is basically off again.

But I'm hopeful I haven't messed it up! I'm not an electronics expert - could anyone explain what that circuit is actually doing?
 
The original mains adapter for the PX-8 (HOOAAx, where is is a country specific variant) is NOT al power supply, but a battery charger. Power should come from the battery pack.
You can operate the PX-8 without the batteries, but voltages get way out of spec. The battery is an integral part of the power supply. And it won't work with cassette or printer. Without batteries, a 5V DC power supply to the battery terminals could be a replacement.
Some more info: https://electrickery.nl/comp/hx20/adapter.html

Greetings,
Fred Jan
 
Yep. You've been telling me this for a couple of years and I think it has finally got through to my brain! The penny has finally dropped...

Have you had any issues using Nimh instead of Nicad as replacement cells? I know it's a really slow charging cycle so I assume there's no issue.

Thanks for all the advice!
 
Crashedfiesta said:
Nimh instead of Nicad
Click to expand...
I tried it, but it didn't charge it right - need to change out the resistor for charging -- the big 2w green one, and then you also have the timing wrong, but it's said that you can leave the battery connected for at least two months to the charger without any battery degradation (check the technical reference manual).
 
For me it did work, but charging the pack with a reasonable current (10% of capacity, 14 hours) before putting them in the PX-8 might have helped. Usually I keep the charger connected during prolonged usage. Occasionally I try to develop code for the PX-4/PX8.

Greetings,
Fred Jan
 
Once again, I should RTFM. Or more specifically, RTFTM (read the flippin' technical manual).

There's a handy flowchart for repairs which immediately tells me:

1) There should be -15v at pin 15 of CN5 (the LCD connector). I get about 0.5v which explains why I'm not seeing anything. It also explains why, when I briefly saw -7v the screen seemed to be slightly working.
2) IC 14D pin 2 should have regular pulses on it (frequency not specified). It does, so that's OK.
3) IC 14D pin 7 should have regular pulses on it too. It does, so that's OK too so I assume IC 14D is OK.

With those three conditions it goes on to suggest that I replace transistor Q29. A pack of 10 is now on order from eBay. If that doesn't work it suggests replacing inductor L3. This is a 0.8mH inductor which seem suspiciously hard to find. I assume that modern equivalents are something like this:

www.ebay.co.uk

Peaking Inductor Coil - 8x10 - 10uh-10Mh - Multiple Values Available - Free P&P | eBay

10uh - 10Mh. Peaking Inductors - 8 10. Resistance 0.05 - 110. United Kingdom.
www.ebay.co.uk

But there's no 800uH or 0.8mH values. Am I looking at the right thing? :unsure:
 
Hi Crashedfiesta,

0.8 is indeed not a standard E-series value. If capacitors are a guide, these come in +/- 20% precision. So 1 mH might work too. It is usally only critically for oscillators. If you order a pack of ten and you have a LC-meter, you might select the one with the least induction. Or remove the plastic cover and remove some windings :-). Induction value is related to the square of the number of windings, very weird.

Greetings,

Fred Jan
 
You must log in or register to reply here.
Back
Top