CPC464 no display :(

[NorthDevonRetroArchive]

Hi All,


I have just been gifted a CPC464 with green screen monitor and a box of sofrware but it's not working.


I purchased a Scart cabe for it with the 5V powr feed and connecting this up to my OSSC I get no dsiplay and no sync.


Connected to the green screen monitor I just get a bright green screen when I turn on the CPC. I tried pressing the dlete key but I get no beep.


Opening up the CPC I have a long board version with part number Z70200 written on the board. One of the RM chips was getting hot so I purchased 8 new chips and socketed them on the board, they no longer get hot but I still get no display.


The Z80 has a stable 4 Mz clock and the Reset and M1 pins show activity. I don't see any activity on the Data pins though. I do see activity on the Address lines. I swapped this out with a known working Z80 but I am going to try this again because I can't remember if I did that before or after replaing the memory.


I checked the clock on the CRTC chip and that is stable at 1Mhz. I see activity on the Horizontal Sync but the Vsyc appears to be stuck high. Could this signify a dead CRTC chip? or would there be another cause for there to not be a vsync.


What should be my next course of action to try and bring this back from the dead? Because I am impatient and hae waiting for deliveries I have already ordered a replacement Z0, CRTC chip and a blank EPROM to replace the ROM chip if needed.


I have an oscilliscope, logic probe and multimeter so is there anything else I should be testing now?


Thanks in advance for any help. Would love to get this working as I have never used a CPC464 before 


Thanks, Matt

[Bryce]

I think there should always be some sort of signal on the vsync, even if everything else is not working, so I'd guess it has an issue. However, that wouldn't explain the lack of activity on the data pins. Try the other Z80 first.

Bryce.

[NorthDevonRetroArchive]

So I have just powered up my spectrum +2 and measure the Data lines to make sure I could see a signal, which I could. I then transferred this Z80 over into the CPC and still nothing. The scope shows a flat line at 1.28V across all the Data pins.


Clock signal is fine. Address lines show activity on all apart from A14 and A15 (Is this right?).


Memory request shows activity, Read pin shows activity but write pin does not (Is this right?)


Would be such much eaier to have a working CPC alongside so I can compare stuff


Thanks, Matt.


EDIT*


Just done some more testing comparing the Z80 on the +2 to the 464 and also doing some reading up. I document I have read says a number of pins need to be high for the Z80 to work, one of these is the Wait pin (24). On my +2 this pin is high but on the 464 I am getting a 1Mhz clock signal. This doesn't seem correct. Where should the wait signal come from?

[llopis]

The Z80 has a stable 4 Mz clock and the Reset and M1 pins show activity. I don't see any activity on the Data pins though. I do see activity on the Address lines. I swapped this out with a known working Z80 but I am going to try this again because I can't remember if I did that before or after replaing the memory.

There is your problem. The /RESET pin should be constantly high. If not, check out the series of logic gates leading up to it. It's possible that they're faulty and preventing the Z80 from starting to work correctly.
Although it's really weird that M1 shows activity, that indicates the Z80 is fetching instructions from memory.

[NorthDevonRetroArchive]

Looking at the circuit diagram the Ready/Wait signal comes from Pin 2 on the Gate array (40010). I have mesured this pin and can see the 1Mhz signal still. Could this still be caused by one of the logic chips? I am worried that the gate array is at fault. But if one of the logic chips (IC114 if I am reading correctly?) is pulling the line low then I guess that could still be the fault?


Is there any sure way to tell if a gate array is working properly without replacing it?


I am new to troubleshooting at this level so please correct me if I am speaking nonsense lol.


Thanks, Matt.

[Bryce]

Check the reset circuitry.

Bryce.

[NorthDevonRetroArchive]

Could you give me a bit of detail on how to check that? I can see the pin is high on the Z80, what else should I check?

[Bryce]

the Reset and M1 pins show activity

I can see the pin is high on the Z80, what else should I check?

These two comments contradict each other. Is there activity on the RESET pin or does it stay high? If it stays high it's fine.


Bryce.

[NorthDevonRetroArchive]

Sorry the Reset pin stays high.


The Refresh and M1 pins show a 1 mhz signal


*EDIT


For the moment ignore the fact I said I didn't see anything on the data bus. I have just done a capture with my logic analyser and there seems to be something there. I see a signal on the address pins on my osicilliscope but not and the data pins. I am learning how to use this equipment at the same time so possibly have it set up wrong for reading these signals.


I have attched an image from Pulseview, does this look in anyway normal?





*EDIT 2


Saying that I checked the data lines on the Z80 in the spectrum and they showed up fine my my scope, will do some more checking after work to confirm.

[NorthDevonRetroArchive]

Ok,


So I plugged the speccy back in and went through each pin on the Z80 and made a note of what I see on the scope, then did the same with the CPC.


My results are attached below. There seems to be a whole bunch of 1Mhz signals on the bottom right of the Z80 on the CPC where as on the spectrum they are random data signals not a non changing clock.


Also noticed the data lines are all stuck at 1.2v solid on the CPC.


Any ideas where to go next? Can't find any article or videos online with this sort of issue


Thanks, Matt.


*EDIT 1


Looking at the signals all around the CPC (like the address lines on the Z80) loads of them dont look like "data" signals, they all seem to be regular pulses. It like the computer is just turning on an off 100 of times per second. The Z80 gets pretty warm to touch after being on for a little while

[Bryce]

It is quite likely that the ROM has failed and the CPU is stuck in a continuous loop.

Bryce.

[NorthDevonRetroArchive]

I already have a blank chip on it's way (Is this the right one? https://www.ebay.co.uk/itm/273533780048)


Will burn the ROM when I get it and try that next.


Thanks, Matt.

[Animalgril987]

Yes, that should be fine.

[NorthDevonRetroArchive]

I have just removed the ROM chip from the board and have stuck it in my EPROM pr0grammer. Using XGPro I selected the chip as M27C256B and tried reading the chip, all I get is blank.


Is there anything else I am supposed to do to read the ROM? Or is the rom chip just dead?


Thanks, Matt.

[llopis]

I have just removed the ROM chip from the board and have stuck it in my EPROM pr0grammer. Using XGPro I selected the chip as M27C256B and tried reading the chip, all I get is blank.

Is there anything else I am supposed to do to read the ROM? Or is the rom chip just dead?

You should be able to read it just fine. But the fact that comes out as totally blank is really weird. Usually a bad ROM chip will report some bad data, but not all blanks (unless something catastrophic happened).


Whenever you get an EPROM chip (or some kind of ROM board), I would recommend you try the diagnostics ROM see if that helps any: https://github.com/llopis/amstrad-diagnostics/releases

[NorthDevonRetroArchive]

Hopefully, the chips will be here in the next couple of days.


I have some AT28C256 chips here which I just tested reading and writing with the programmer so it seems that is all working properly, but when I read the Amstrad ROM it says it has read ok but its just filled with FF.


Will try the diagnostic ROM first. With a bit of luck it's this dead ROM chip that is causing the fault.

[Animalgril987]

If the ROM is only reading as all &FFs then this will be your problem as FF is the code for RST38 ie, a call to &0038 which is of course going to be read as FF ( another call to exactly the same location). So, as Bryce said, your CPU is stuck in a loop.

[Bryce]

I have just removed the ROM chip from the board and have stuck it in my EPROM pr0grammer. Using XGPro I selected the chip as M27C256B and tried reading the chip, all I get is blank.


Is there anything else I am supposed to do to read the ROM? Or is the rom chip just dead?


Thanks, Matt.

I've no idea of your soldering skills, but it's extremely easy to fry an EPROM while removing it. Is this something you've done before? Although it's also highly plausible that it was dead to start with.

Bryce.

[NorthDevonRetroArchive]

I've not specifically unsoldered an Amstrad ROM chip before but have been soldering for 25 years and not killed anything yet . Can't say I've ever removed an EPROM with the intention of re-using it before though, things I have done previously was just removing the old chip and replacing it with a new one so who knows!


I prefer the thought that the ROM was dead before and replacing it will spring this back into life, but we will see

[Audronic]

@ NorthDevonRetroArchive
Is the rom marked as 40009 ?


Attached is a 40009-rom.bin
Keep Safe


Ray

[NorthDevonRetroArchive]

Yes that is the correct version, thanks


Matt.

[NorthDevonRetroArchive]

Thanks everyone for all the pointers. EPROM turned up today, flashed it with the ROM supplied above and soldered socket to board.


And Bingo!! We have life! Just need to put it all back together now and see if anything else needs attention


Matt.

[Animalgril987]

Excellent news! Another CPC brought back to life.

[Gryzor]

Heart-warming photo