News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu

CPC 6128 RAM Bank 0 failing - Need help

Started by vdbjimmy, 09:17, 14 January 20

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

vdbjimmy


Hello,


I bought a CPC 6128 on which RAM bank 0 is failing. Grey/green screen with black border when I power up.


By trying the RAM swapper (I found an example from Bryce on this message http://www.cpcwiki.eu/forum/technical-support/which-is-bank-0-and-which-is-bank-1-on-6128/msg169671/#msg169671), the computer can boot.


I replaced all chips from bank 0 and added sockets and new RAM chips but still the same. Previously, in bank 0, I had a mix of two type of chips (4164 and 4264). That's why I decided first to change evrything. I replaced bank 0 with 4164-10 and I have two 4161-15. I did not receive the same speed of chips. Could it be a problem? Note that on bank 1, all chips are 4264B-10.


But still the same. I'm still stuck in the grey/green screen. I do the continuity tests as suggest Gerald in this thread (http://www.cpcwiki.eu/forum/technical-support/which-is-bank-0-and-which-is-bank-1-on-6128/msg169773/#msg169773) and everything is fine.


I also tested the HAL chip on another working CPC 6128 computer and it is working correctly. It is hard to say but the RAM chips do not seem hotter by comparing each other. Maybe the HAL chip is hot but hard to say really.


So I'm wondering now, what can be tested? If I'm correct, when I do the RAM swapper jumpers thing, when boot up the computer is using bank 1 instead of bank 0. But by using this trick, bank 0 is disabled, correct? So no way to test it directly with Basics at least?


I'm thinking also if the problem could be related to IC114 and IC115 (74LS373 and 74LS244). Is there a way to test traces or even the chips themselves? I only have a multimeter but I think to buy a logical probe if you think it is a good idea/needed.


The problem could be also related to my new RAM chips where some can be also defective... I bought them on aliexpress. But I fear I'm stuck if I can't diagnose the chips themselves


Thanks for your help !

Bryce

IC114 and 115 are used by both banks. If there was anything wrong with them, the CPC wouldn't boot from bank 1 either. The problem can only be with the 8 chips in bank 0. Either one of them is bad or there is a broken track somewhere under the chips.

Bryce.

vdbjimmy


I checked for broken tracks and everything seems fine. I did carrefully the continuity tests suggest by Gerald in a post in mention earlier. So yes, maybe one of my "new" RAM chip is also defective.

I read something about a diagnotic board ROM which might help to find a defective RAM chip. Any idea how can get/buy one of those? Thank you.


Jimmy.

Bryce

Speak with @TotO , he's in France too and can supply you with what you need.

Bryce.

TotO

Exactly. Please, PM me and we will see what we can do.
"You make one mistake in your life and the internet will never let you live it down" (Keith Goodyer)

vdbjimmy


Thanks to @TotO, he sent me the diagnostic board and I received it today !


I did some tests, I'v got a lot of noise on the screen but sometimes I've got the picture attached. The border seems red everywhere. Does it means my new chips are all damaged? It seems a bit strange for me.


Any other suggestions on what to test?


Thank you.




TotO

#6
Is-it possible to see a picture of your mainboard with the RAM IC on it?
"You make one mistake in your life and the internet will never let you live it down" (Keith Goodyer)

vdbjimmy


Now, I've something interesting. See the attached screen picture. I can see correctly the damaged RAM chips and if I switch one damaged chip with a working one, the red bar is also moving.


This is what I've done. As I had a strange behaviour with the diagnotic board, lot of noise, not every time the same thing on screen, sometimes the border sometimes not, I decided to do a more complete continuity test. I did the checks on the post Re: Memory problem? with a 6128


Quote from: gerald on 19:13, 28 July 18
By the way, can you see something on the screen during the test phase, or it's always black ?


By removing the previous RAM device you may have damaged some track or made short.
If you only have access to a continuity tester, check that the whole memory is properly connected.
That mean :
- Z80 (IC)data bit are going to the DRAM DIN (pin 2), for both banks.
- DRAM data out (pin 14) are going to the data latch (IC114)
- adress are going from the address mux (IC104, IC105, IC109, IC113) to the DRAM
- control signals (RAS, WE) from the GA/PAL, CAS from the PAL


6128 schematic can be found here : http://www.cpcwiki.eu/imgs/4/4a/CPC6128_Schematic.png


But the continuity check was ok.


In addition, I tested the connection between the PAL pin 19 and IC109 pin 6 and PAL pin 12->IC109 pin 10. I don't have continuity when testing on the PAL legs directly. But I have continuity with the  soldering point. It seems that I did not have a good contact of the PAL chip into the socket. I removed it, cleaned it, bent a bit the legs and now the computer is almost working. At least we can see the damaged RAM chips and I no longer have strange things happening on screen.


So it seems that I need two new RAM chips. I bought ten RAM chips and 4 seems damaged...


@totO, I attached also the mainboard picture you asked

TotO

OK. I think that you can ask the seller to resent to you 4x working IC or a partial refund.
"You make one mistake in your life and the internet will never let you live it down" (Keith Goodyer)

vdbjimmy


Good news ! The mainboard is now working correctly !


I installed 2 new RAM chips (thanks @totO for your great help), and the diagnostic board show me "green bars" everywhere which means that the RAM test is done successufy. And now the computer boots correctly.


For the records, this is what I've done to fix this mainboard:


- When I turned on the computer, I get a fixed green/grey screen. It seems to be a RAM 0 bank failing (after reading a few posts here)
- To ensure, it's only RAM 0, I do the RAM swap test (http://www.cpcwiki.eu/forum/technical-support/which-is-bank-0-and-which-is-bank-1-on-6128/msg169773/#msg169773) to use RAM bank 1 instead of bank 0. It disables bank 0 but the computer boots by using bank 1 instead. And YES, the mainboard boots by using bank 1.
- So I unsolder all bank 0 RAM chips, check the traces around the bank 0 area, solder sockets and put new RAM chips.
- But the mainboard still did not boot
- In the meantime, I received a diagnostic board sent by @totO. I really have a strange behaviour with the board. Dirty picture on screen, wrong pixels color everywhere... Sometime I had something on the screen from the diagnostic tool but with a lot of noise. 
- So I checked traces again more deeply (very very import actually) and I found a bad connection with the PAL and his socket, sometimes I had connection betwwen the chip leg and the solder plot and sometimes not. So I removed the chip, bent the leg to ensure the connection with the socket is more accurate, cleaned socket+chip legs with alcohol. Put the chip back into the socket
- And then the diagnostic board works really better. I had a clean picture which shows me 2 bad chips... No luck, I ordered ten new chips though...
- Once I get two new chips installed (thanks again @totO), you can see the attached picture, everything is working fine !


I hope it could helps someone else !


Jimmy

Powered by SMFPacks Menu Editor Mod