try to repair my cpc 464

[rt1k]

Hi Guys,
recently bought 2 cpc 464 in good esthetic condition but unfortunately with not working hardware. They have the same problem, probably caused by overvoltage or wrong use of power supply unit.
So in my adventure to bring back to life these my old friends i have tried to :
1) reading the forum, some experienced user, suggest to identify the damaged ic by measuring their temperature; using this method 1 ram IC is really hot, IC113, desoldered and replaced with new one; before the cure the video out is not stable, borderless like untuned tv; after the screen is stable with blue border but with garbage on the middle;
2) replaced Z80 with some one tested on speccy 2+ : same result
3) all other ic are cold, also after some times of operation, only the Z80 become slightly warm;
Now what should i do to investigate other problems ?
I could use an oscilloscope to measure signals.
Thanks in advice for all your support and experience.

Regards.

[Bryce]

I'd try swapping the ROM IC.

Bryce.

[Munchausen]

From my experience (putting 12V instead of 5V into a 6128 but the unusual one with a 40007 gate array) I had to swap all the RAM ICs and everything else was still basically working. The gate array was a bit dodgy (sometimes worked, sometimes didn't), and I still haven't got everything working completely reliably.

[Bryce]

What rt1k has described doesn't sound like an over-voltage to me. At least not a 12V over-voltage.

Bryce.

[rt1k]

Well,
as Bryce suggested i desoldered the amstrad 40009 ic and tried to read it with eeprom programmer.
My programmer read it correctly as 27c256 eetype.
CPC 464 Board is mc0009a; what can i do now ? any suggestion ?


Regards,
Rt1k

[Bryce]

You may still have another dodgy RAM, that's not causing a short, but also not working. Can you post a picture of the screen you're getting?

Bryce.

[rt1k]

Hi Bryce,
here the snaps; results change on every power up.


I replaced ram chips : IC112 IC113 IC114 IC116 as suggested but without changes.


Regards.
Rt1k

[Bryce]

Oh, it's a costdown version. If it's the ASIC there's not much you can do to save it. Are you using a regulated 5V supply?

Bryce.

[rt1k]

Yes,
+5v 2A.
The other board with the same problem is a v3 - MC0044A.


Do you think i have greater chance to restore that board ?

Many thanks
Rt1k

[Bryce]

Yes, definitely. Almost all the parts on that board can be found/replaced if needed.

Bryce.

[TFM]

Looks like a Gate Array problem

[rt1k]

Well,
switching to the second board that is an MC0044D model.
- Z80 replaced and tested on speccy.
- Removed RAM IC120, that become really hot and replaced it with new  as on the other board.
- Desoldered rom ic, read & tested with eeprom programmer.
Now almost on any power up the screen has black borders and white area, no sound on key press; sometimes the screen stays completly black for approx. 2 seconds, before going into the condition above.


what do you suggest now ?


Regards.
rt1k

[Bryce]

What's unusual is that both CPCs, although completely different models, have very similar failures. Did the previous owner give any hints to how they broke? It definitely wasn't an over voltage, reverse voltage is also unlikely. I would suggest socketing the other RAMs, if you have spares to try this.

Bryce.

[rt1k]

Well guys,
i'll update the status of recovering the boards:


working on MC0044d as bryce suggested :
1 - desoldered all ram IC;
2 - desoldered all relative caps  to ram ic;
3 - desoldered 74ls244 and relative capacitor;


Once removed i've tested ram ic one by one on a socketed commodore 64 machine; and i have found 1 faulty ram chip. (as Bryce expected. Great!)
With new capacitors and replacing damaged ram ic the cpc now boot;though during the operation of desoldering some tracks needs to be fixed.


Now the worst news:
the keyboard is not working, sometimes the cursor moves on the screen randomly.
Reading the forum and the 464 schemes the culprits could be the 8255, the 74ls145 or the 8912 ?
What now ?


Regards
Rt1k

[TFM]

Well, I assume you cleaned the keyboard. So maybe some tracks are broken, that could explain it.

[Bryce]

The AY-3-8912 is the most likely to have failed.

Bryce.

[rt1k]

Hi Bryce,
found and fixed a broken track on 74ls32, now the keyboard is partially working.
On every boot i get always this string  : 654;lhgs->X (also with keyboard disconnected from the board)
Swapping AY-3-8912 with another one (from speccy+2) makes no difference.
Keyboard is clean and working.
Seems like data bus is unstable on keyboard scan cycle.
Regards.
rt1k

[gerald]

654;lhgs->X (also with keyboard disconnected from the board)

This is bit 4 of the keyboard matrix low for all lines.
Since you already swapped the AY, you should check that this bit is not grounded on the board itself (from connector to AY).
If you also get this without the AY on board, then it could be the 8255 port A bit 4 that is dead.

[rt1k]

Hi guys,
well good news, as suggested by gerald inspecting on bit 4 of the ay and i've found the problem. (ground signal on pin 10 of ay)
I can be satisfied of the result achieved :
- 1 board fully recovered - msc0044d; 
- 1 board missed in action (swapped all the ram ic, an Z80 but unfortunately i can't replace the asic). 
Many thanks to Bryce, Gerald and all the other cpc guys for your valuable help.


Kind Regards.
rt1k

[Bryce]

Congratulations, happy gaming.

Bryce.

[Gryzor]

Yay Happy news!

[TFM]

Hi Bryce,
found and fixed a broken track on 74ls32, now the keyboard is partially working.

Ha! I did know it.    Glad your system works now.