News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu

Amstrad CPC 464 - Screen with Garbage

Started by jarcher1980, 13:34, 12 March 22

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

jarcher1980

Hoping someone can answer my above queries.

Suspecting the voltage may be an issue, I decided to test what happens to the voltage when removing certain ICs one by one. I'm getting 4.75v at the CPU, RAM, etc usually. When removing memory ICs, the ROM and the CPU the voltage stayed the same. When I removed the Gate Array the voltage jumped up to 4.92v—which I gather is healthy.

Does this suggest the Gate Array is faulty—or is this normal behaviour?

I do have a CPC 6128. Are the Gate Arrays backwards compatible? Could I swap it over to check if the Gate is the issue?

Thanks!


Bryce

#51
I've never measured the difference, but removing the Gate Array pretty much stops all activities on the entire board which would lower the current. However, that's quite a difference. Are you sure your PSU is able to supply the required current?

The 40007 and 40010 are not pin compatible, so you will need the correct part number for your 6128. If your 464 has the same GA you can swap them to check.

Bryce.

Nworc

Out of curiosity I wanted to find out how low you can go on the 464, using a laboratory power supply, measured using a Voltcraft VC404 at the connector (voltage drops by 0.03V when turning the machine one, voltage measured under load):

The 464 boots stable at a voltage of 4.02V and above.
Between 3.90V and 4.02V is the range of free demos, a different one per different voltage, boot message still visible.
Below 3.90V I could not get a boot message.

These figures will be different from machine to machine, but I think it's obvious to see how stable the 464 runs even if the PSU is not ideal.

In the demo voltage area once I experienced the clicking of the cassette motor relais, another time I could observe the vertical stripes shown in this thread, so it is possible to get the component to fail at a too low voltage, I mean that component which fails regularly in the board mentioned here.

jarcher1980

#53
Quote from: Bryce on 10:24, 25 April 22I've never measured the difference, but removing the Gate Array pretty much stops all activities on the entire board which would lower the current. However, that's quite a difference. Are you sure your PSU is able to supply the required current?

The 40007 and 40010 are not pin compatible, so you will need the correct part number for your 6128. If your 464 has the same GA you can swap them to check.

Bryce.
Thanks for getting back to me about the Gate Arrays. I'm pretty sure the 6128 has a 40010, but I'll double check.
I decided to take out the GA, give it a good clean and remove the original metal heat sink. It was getting in the way of checking memory ICs, so I replaced it with a modern heat sink. Looks good and seems to work fine. The problem still exists though.

I'm starting to think the voltage may not be the issue. I have acquired an Amstrad Diagnostic rom and it's produced some interesting results. https://photos.google.com/share/AF1QipO5iI8JGaMO9iX7d_FN6-ZSRCcRjpGg9ddKLWcdA-2QAOBw_U1Qr9N4luPSxpFjDQ?key=eVhVb3VldG1NLTQ1bzlFeFIzVHpNZEUwb0pvTlNR

The first image was the first test. It showed 3 good ICs at bit 2, 4 and 6.

I have experience of Spectrums, and I know these tests can show RAM faults when there's actually other problems. So, I deliberately moved the green (working) ICs to bit 0, 1 and 2 and they still showed green (see image 2). Does this suggest that these green memory ICs are actually good and the others aren't? I'm guessing so.

The thing is, I have a DRAM tester and I've checked all the original memory ICs and some new ones I bought, and they're all reported as fine.

I'm more confused now. :-)

Is it worth buying some more 4164 memory ICs, or could there be an another problem I'm missing?

Thanks!

jarcher1980

Quote from: Nworc on 13:10, 25 April 22Out of curiosity I wanted to find out how low you can go on the 464, using a laboratory power supply, measured using a Voltcraft VC404 at the connector (voltage drops by 0.03V when turning the machine one, voltage measured under load):

The 464 boots stable at a voltage of 4.02V and above.
Between 3.90V and 4.02V is the range of free demos, a different one per different voltage, boot message still visible.
Below 3.90V I could not get a boot message.

These figures will be different from machine to machine, but I think it's obvious to see how stable the 464 runs even if the PSU is not ideal.

In the demo voltage area once I experienced the clicking of the cassette motor relais, another time I could observe the vertical stripes shown in this thread, so it is possible to get the component to fail at a too low voltage, I mean that component which fails regularly in the board mentioned here.

That's really interesting. Thanks. So it could still be the voltage that's causing one component to fail. Or, it's some other problem—such as the memory ICs (see my post above).

I have tested the machine with two different modern power supplies (built for the 464) and I also used an original working Amstrad colour monitor and the issue persists with all. 

Thanks.

eto

Quote from: jarcher1980 on 21:57, 25 April 22The thing is, I have a DRAM tester and I've checked all the original memory ICs and some new ones I bought, and they're all reported as fine.

Noel's retro lab just recently covered an issue where a ram chip was reported as working, but didn't work in a computer (also not in the CPC):



Bryce

Quote from: Nworc on 13:10, 25 April 22Out of curiosity I wanted to find out how low you can go on the 464, using a laboratory power supply, measured using a Voltcraft VC404 at the connector (voltage drops by 0.03V when turning the machine one, voltage measured under load):

The 464 boots stable at a voltage of 4.02V and above.
Between 3.90V and 4.02V is the range of free demos, a different one per different voltage, boot message still visible.
Below 3.90V I could not get a boot message.

These figures will be different from machine to machine, but I think it's obvious to see how stable the 464 runs even if the PSU is not ideal.

In the demo voltage area once I experienced the clicking of the cassette motor relais, another time I could observe the vertical stripes shown in this thread, so it is possible to get the component to fail at a too low voltage, I mean that component which fails regularly in the board mentioned here.

I don't think I'd trust a Voltcraft VC404 to measure voltages accurately. Your stated results are way too low to be correct.

Bryce.

jarcher1980

Quote from: Bryce on 08:08, 26 April 22
Quote from: Nworc on 13:10, 25 April 22Out of curiosity I wanted to find out how low you can go on the 464, using a laboratory power supply, measured using a Voltcraft VC404 at the connector (voltage drops by 0.03V when turning the machine one, voltage measured under load):

The 464 boots stable at a voltage of 4.02V and above.
Between 3.90V and 4.02V is the range of free demos, a different one per different voltage, boot message still visible.
Below 3.90V I could not get a boot message.

These figures will be different from machine to machine, but I think it's obvious to see how stable the 464 runs even if the PSU is not ideal.

In the demo voltage area once I experienced the clicking of the cassette motor relais, another time I could observe the vertical stripes shown in this thread, so it is possible to get the component to fail at a too low voltage, I mean that component which fails regularly in the board mentioned here.

I don't think I'd trust a Voltcraft VC404 to measure voltages accurately. Your stated results are way too low to be correct.

Bryce.

Any thoughts on the diagnostic rom results? I'd be keen to hear your thoughts before I buy some more 4164 ram. Thanks.

Nworc

Quote from: BryceI don't think I'd trust a Voltcraft VC404 to measure voltages accurately. Your stated results are way too low to be correct.

Bryce.

I know, the VC404 is cheap, but I like the simple design. However, the Hioki validates the result:


jarcher1980

Quote from: jarcher1980 on 21:57, 25 April 22I'm starting to think the voltage may not be the issue. I have acquired an Amstrad Diagnostic rom and it's produced some interesting results. https://photos.google.com/share/AF1QipO5iI8JGaMO9iX7d_FN6-ZSRCcRjpGg9ddKLWcdA-2QAOBw_U1Qr9N4luPSxpFjDQ?key=eVhVb3VldG1NLTQ1bzlFeFIzVHpNZEUwb0pvTlNR

The first image was the first test. It showed 3 good ICs at bit 2, 4 and 6.

I have experience of Spectrums, and I know these tests can show RAM faults when there's actually other problems. So, I deliberately moved the green (working) ICs to bit 0, 1 and 2 and they still showed green (see image 2). Does this suggest that these green memory ICs are actually good and the others aren't? I'm guessing so.

The thing is, I have a DRAM tester and I've checked all the original memory ICs and some new ones I bought, and they're all reported as fine.

I'm more confused now. :-)

Is it worth buying some more 4164 memory ICs, or could there be an another problem I'm missing?

Thanks!
I have acquired 10 new 4164 ram chips (KM4164B-10). I have a tester and they're all reported as good.

When testing the system via the Amstrad Diagnostics rom - none of these new chips are showing as green. Some others chips I've had for a while (which are exactly the same make/number/timings) are showing as green. I even moved these 'green' ones to different IC positions and they show green in those positions too.

Is the CPC very fussy about ram chips, or could something else be going on? I can't keep buying ram chips.

Nworc

Quote from: jarcher1980Is the CPC very fussy about ram chips, or could something else be going on? I can't keep buying ram chips.

Hm, I think you already received a couple of ideas on which chips to test that deal with the RAM logic, but what I see for now is that you have checked the RAM chips, so how did the test results for IC112/114/115 and the address line multiplexers turn out?

jarcher1980

Quote from: Nworc on 19:50, 30 April 22
Quote from: jarcher1980Is the CPC very fussy about ram chips, or could something else be going on? I can't keep buying ram chips.

Hm, I think you already received a couple of ideas on which chips to test that deal with the RAM logic, but what I see for now is that you have checked the RAM chips, so how did the test results for IC112/114/115 and the address line multiplexers turn out?

I've checked for continuity between the RAM chips and IC115 / IC114 and IC112 and it's all good. I haven't sourced replacement ICs yet as I'm not sure that's the problem; although that will be my next step.

I don't understand why Amstrad Diagnostics (by llopis) would report green for certain 4164 chips in any ram position (IC117 to IC1124) but red for others—despite all of the ram chips passing a dedicated 'ram tester' check. Only 1 ram chip failed the test.

Audronic

@jarcher1980 

Can you post a Photo of your Ram tester Please

Thanks

Keep Safe

Ray
Procrastinators Unite,
If it Ain't Broke PLEASE Don't Fix it.
I keep telling you I am Not Pedantic.
As I Live " Down Under " I Take my Gravity Tablets and Wear my Magnetic Boots to Keep me from Falling off.

jarcher1980


Audronic

@jarcher1980 

Ok Thats the one I use and it seems to be a good choice.

I just found 14 (Fourteen) Faulty out of a purchase of 20 - 4116 Memory chips with that Dram tester

Thanks for the photo

Good luck with the Project

Keep Safe

Ray
Procrastinators Unite,
If it Ain't Broke PLEASE Don't Fix it.
I keep telling you I am Not Pedantic.
As I Live " Down Under " I Take my Gravity Tablets and Wear my Magnetic Boots to Keep me from Falling off.

jarcher1980

#65
Just checking a few things periodically. I found this site which suggests that the Gate Array (40007) VCC2 should be reading 5v—just like VCC1. Can anyone confirm if this is definitely the case, as I'm pretty sure the last time I checked it was reading around 3.3v on my system

https://www.grimware.org/doku.php/documentations/devices/gatearray

Thanks.

panagiotis1974

did you finally find the fault?
I have exactly the same symptoms.



Powered by SMFPacks Menu Editor Mod