CPC 6128 - Basic does not start and there is garbage on the screen

[Nigel]

Good day.
Amstrad CPC 6128 - when turned on, the "Amstrad 128K Microcomputer (v3)" start screen is visible for a split second, and then the screen is covered with artifacts. At the same time, if you look closely, the inscription in the background is still visible, but the computer does not load the Basic. There is no reaction from the keyboard. Previously, before artifacts began to appear steadily, the picture was normal, but the computer also did not load into Basic.

Is it possible to determine the maintainability of this machine based on the existing symptoms? What kind of malfunction could it be?

[Skunkfish]

Good morning Nigel.

I think the most likely cause of the issue is that one or more of the RAM chips inside the CPC is faulty.

Luckily, these are an easy to find part but since they are soldered directly to the mainboard, you would need to be comfortable with using a soldering iron to replace - in fact, you might be best installing sockets instead so that the memory can be replaced more easily in future if required and to reduce the risk of damaging the chips when installing.

A common sign of a faulty RAM chip is that it will heat up more rapidly than the other chips. Open the case to the CPC after it's been on for a period of time and see if one chip is hotter than the others.

The RAM chips that you're after are 4164 15ns chips, you should be able to find these quite easily on eBay for not too much money.

Hope that helps?

James.

[Nigel]

Good morning Nigel.

I think the most likely cause of the issue is that one or more of the RAM chips inside the CPC is faulty.

Luckily, these are an easy to find part but since they are soldered directly to the mainboard, you would need to be comfortable with using a soldering iron to replace - in fact, you might be best installing sockets instead so that the memory can be replaced more easily in future if required and to reduce the risk of damaging the chips when installing.

A common sign of a faulty RAM chip is that it will heat up more rapidly than the other chips. Open the case to the CPC after it's been on for a period of time and see if one chip is hotter than the others.

The RAM chips that you're after are 4164 15ns chips, you should be able to find these quite easily on eBay for not too much money.

Hope that helps?

James.

James, good morning and thanks for reply!
If it's the memory chips, then that's probably the least of the problems. I thought about the memory chips when garbage appears on the screen, but in the case of problems with BASIC I assumed something more serious. Can I use faster 4164 memory - for example, 12 or 10 ns?

[Bryce]

Hi Nigel,
     first a confirmation of James' diagnosis - Yes it's definitely a RAM issue. And to your other question - Yes, you can use faster RAMs.

Bryce.

[Nigel]

Hi Nigel,
    first a confirmation of James' diagnosis - Yes it's definitely a RAM issue. And to your other question - Yes, you can use faster RAMs.
Bryce.

Bryce, thanks for reply! I really hope it's RAM and I'll be able to repair this CPC 6128.

[Bryce]

Hi Nigel,
    first a confirmation of James' diagnosis - Yes it's definitely a RAM issue. And to your other question - Yes, you can use faster RAMs.
Bryce.

Bryce, thanks for reply! I really hope it's RAM and I'll be able to repair this CPC 6128.

It's definitely a RAM issue, but your soldering skills will determine whether you get it working again or not.

Bryce.

[McArti0]

Try keep READY line to GND when not all characters will be write. yet.

Other test:
Try one touch RESET line to GND after crash.

Other test:
After crash, Try keep RAMDIS line to +5V, next one touch RESET to GND,

[Nigel]

Try keep READY line to GND when not all characters will be write. yet.
Other test:
Try one touch RESET line to GND after crash.
Other test:
After crash, Try keep RAMDIS line to +5V, next one touch RESET to GND,

Thank you. If possible, I will check all your instructions.

[Bryce]

Try keep READY line to GND when not all characters will be write. yet.

Other test:
Try one touch RESET line to GND after crash.

Other test:
After crash, Try keep RAMDIS line to +5V, next one touch RESET to GND,

Hello McArti0,
      can you please stop doing this! This is not how to diagnose a hardware problem and I believe you are well aware of this. The symptoms shown are textbook symptoms for failed RAM. You should know this too. But you suggest questionable experiments without providing anything further information.
So if I " Try keep READY line to GND when not all characters will be write. yet." (or either of the other two experiments), what result should Nigel expect and what would it actually tell him? What result of any of these tests would confirm or rule out the most obvious cause? This forum is full of similar enquiries and 99.9% of them were confirmed to be a RAM failure. What value does your suggestion add? If there really was any doubt, then you could suggest testing the temperature of the RAM chips or even piggybacking chips! But telling someone to bridge points on their CPC is irresponsible and completely unhelpful. Why do you do this?

Bryce.

[McArti0]

Try keep READY line to GND when not all characters will be write. yet.

Cpc should hang. It is certain that the processor will do nothing further. If the memory does not break, this will suggest that only the rom may be damaged.

[McArti0]

Other test:
After crash, Try keep RAMDIS line to +5V, next one touch RESET to GND,

After enabling it, the colors in GA will be set, then the RAM will be cut off from the processor, you will be able to observe whether the content changes over time. 

[McArti0]

Other test:
Try one touch RESET line to GND after crash.

If the memory is cleared after the reset as it was immediately after switching on, this may mean that your memory is fine.

If the reset does not do restart, it may mean RAM is broken.

[GUNHED]

imho a broken BASIC ROM.

[moxjemi]

Good day.
I have the same problem as Nigel.
What I did:
Replaced the entire RAM and installed it on the panels. I checked the new chips on the tester.
Installed a panel for ROM, and replaced it with 27С256, programmed a file compiled from the OS + BASIC into it.
For the experiment, I tried to change the РAL16V8 to GAL16V8, programmed into firmware from WIKI.

This did not give any effect, it does not reach the text BASIC 1.1.
What else can be checked?

[McArti0]

Why do you do this?

Thats why...

What I did:
Replaced the entire RAM and installed it on the panels
(...)
This did not give any effect

[McArti0]

This did not give any effect

What exactly do you see on the screen?

Try set ROMDIS to Hi. Are You seeing vertical bars?

[moxjemi]

Good day.

Video:

[moxjemi]

No, when installing ROMDIS in HI, the stripes do not appear.
If i set ROMDIS to Hi before start, i just see a clean grey rectangle with a black border.

[McArti0]

Is the contact on ROMDIS very stable?
These artifacts on the straps are disturbing.

Now add RAMDIS Hi and wait 15 minutes to see if the stripes become blurred with artifacts.

[eto]

Yes it's definitely a RAM issue.

Just for curiosity? What would cause this? The CPC properly boots and even shows the boot message, so I would have expected that then the RAM should be fine and would have out my focus towards (BASIC) ROM execution. I could imagine damaged cells in the RAM where it only behaves incorrectly for certain rows/columns but I have never seen a CPC where the RAM was only partially defunct.

[moxjemi]

Контакт на ROMDIS очень стабильный?
Эти артефакты на ремнях беспокоят.
Теперь добавьте RAMDIS Hi и подождите 15 минут, чтобы посмотреть, не размываются ли полосы артефактами.

The contact ROMDIS is stable.
The photo shows the moment when the RAMDIS line was in Hi after turning on the computer.

If I put it in Hi before turning on, I also get a gray screen without artifacts with a black border.
I waited for 15 minutes here and there, and there was no blurring.
Caught an interesting moment (video).
I periodically put RAMDIS in Hi, then released it. When I put it, the dots on the screen stopped moving.

[McArti0]

The photo shows the moment when the RAMDIS line was in Hi after turning on the computer.

Here you have not yellow bar?

Video:

This pattern 00:02 and dots are very bad.

[moxjemi]

The problem was where it was not expected. The resistor R112 was broken. Probably the previous owner broke it when he was inserting the FDD connector. And it is not immediately visible because of the internal cable.

McArti0, Thanks! Now I will start installing GOTEK there.

[McArti0]

CPC turn on normally?

[moxjemi]

CPC turn on normally?

Yes.
Shows me BASIC 1.1 and READY.