CPC 6128 shows only huge bright rectangle with a darker frame

[Hurrican]

Hi to all CPC experts here!


I am trying to repair a defective (German) CPC 6128 and I am at a point where I need support.


The problem:
When I start the computer, there is no startup message or cursor.
On the screen you can see  a huge bright rectangle with a darker frame.
The computer shows no reaction.
Presumably the previous owner supplied the computer with incorrect voltage or polarity.


What have I already done / checked:
1. I checked the RAM and found one chip to be faulty (got very hot, also the 5V had dropped to 4,4 V). I swapped it with another chip and the voltage went back to 5V and the new chip did not get hot anymore. No change in behavior.
2. I checked voltages and frequency: 5V and 12V are OK, CPU clock is at 1 MHz.
3. I cleaned the board and the sockets.
4. I scanned through the forum to get more insights.
5. I checked the CPU, the Gate Array and the PA_/HAL (IC118) in another board - they worked fine.
6. I replaced all electrolytic capacitors with new ones.
7. I checked temparatures: In the defective board some chips get warm (CPU 53 deg. celsius, IC118 43 deg. c., VDU 46 deg. c.), in a working board the chips do not get as warm.
8. I checked VSYNC and HSYNC at the Gate Array - they are OK.
9. I checked activity on the data and address bus of the CPU. There is activity, but I did not have a working logic analyser to check further.
10. I checked the interrupt - it's missing both at the CPU and at the Gate Array. Hence, it does not seem to get created.


I have the following assumption:
I reckon that the 40025 ROM is fried, hence the startup fails and the Gate Array never starts to generate interrupts.
My solution would be to replace the 40025 with an eprom. I would need to get an eprom burner and the correct ROM file though.


Before I do that, I would like to get some input, if this is likely to fix it.


From the information above, can anyone confirm or disprove my assumption?
Do you have suggestions for next steps?


Thanks a lot for any input. If you would like to see the broken computer in action (and my failed attempts), feel free to have a look at videos 2 to 7 from this playlist: https://www.youtube.com/watch?v=RH3qXnjJGZI&index=13&list=PLnK1tlcRPaW3ZBnwtLAjNcEwVKvNHr63y&t=0s


Kind regards,
Dennis

[llopis]

A damaged ROM could definitely be causing that, but so could a bad RAM (even if they don't get hot). Since it sounds like you have another board, could you test the ROM from that one? I know it's not socketed, but it's probably worth it at this point (before buying an EPROM programmer). Also, if you narrow it down to a bad ROM, I'd be happy to send you one burned with the correct data if you don't get anyone local to help.


Which interrupt were you checking? Just curious because that's not one of the things I check.


Good luck.

[Hurrican]

Thanks for the quick reply!
The problem with testing the ROM ist that the working model is only a loan. I felt comfortable swapping socketed chips, but I would need to check with the owner if I may unsolder something.


As for the interrupts: I checked pins 16 (IRQ) and 20 (IORQ) on the CPU - oscilloscope pictures are attached to this post. seems I cannot attach them. Please find them here instead:
http://www.virtualdimension.de/temp/IRQ_CPU_16.jpg
http://www.virtualdimension.de/temp/IORQ_CPU_20.jpg

[Bryce]

It's more likely that a further RAM chip is dead. Check the RAM output pins of each IC and see if one of them isn't doing what it should.

Bryce.

[Hurrican]

Hi Bryce!


Can you give me a hint to what I should be looking for specifically?

[Bryce]

Pin 14 of each IC is Data out. If you have a scope or a logic probe you can check that each RAM chip is at least sending something out. If any chip is always 1 or always 0 on this chip, then it's most likely bad.

Bryce.

[Hurrican]

Thanks a lot, I will check that.

[Hurrican]

OK, I checked all RAM chips with a logic probe. For all chips I get both high and low on pin 14.


However, I also checked pin 2 (data in?) and for two chips I only get a high signal (while I get high and low on all other chips on pin 2).
Is that expected or could that hint to defective chips?

[Hurrican]

Oh, BTW: On the board there are 4264 RAM chips.
I only got 4164 chips as replacement for the defective one. From my understanding they should be compatible.


Are they? Can you mix them?

[Bryce]

To answer question one first: Yes, mixing is usually ok. However, if Din pins are stuck high, then some other IC is most likely causing the issue, not the RAM. You'll need to isolate the problem and find out what's holding these signals high.

Bryce.

[Hurrican]

OK, I'll investigate further on the weekend. Thanks!

[Hurrican]

The weekend turned out to be weeks, but being in vacation right now I finally got the time to do more work on the CPC 6128.

I did a lot of measuring together with a fellow computer fan from the local area, but the result were inconclusive and hinted to ROM and RAM.
I have exchanged the ROM (thanks a lot to llopis for sending a replacement!), however, this did not fix it.
I then went on to remove all RAM chips, solder in sockets and putting in new RAM chips.

Now it at least shows part of the startup message!

However, when I turn it on, I get one of three behaviors:
1. It show the startup message up until "Locomotive Software Ltd." but doesn't show "BASIS 1.1" and the Ready prompt. Then it freezes.
2. Same as 1, but the machine resets each time it reaches "Ltd.".
3. Same as 1 or 2, but it show additional graphic bugs (some pixels or rows of pixels on the screen).

Do you have any more insights into what to try next?

Kind regards,
Dennis

[Hurrican]

And it finally works! :-)


FYI: In the end the problem was "only" two broken RAM chips. Unfortunatly the replacement ROM seemed to have an issue, so it sent me off track. It does work without a problem with the original ROM chip, however. :-)


Thanks a lot for the help!


Kind regards,
Dennis