CPC 6128 broken: strange screens

[locked]

Hello,

I've been trying to troubleshoot a CPC 6128 for some time now and I'm a bit at a loss.

- The computer's voltage are OK: 5v and 12v, no short circuit
- Monitor appears to be OK (normal sound, no horizontal line)

- I can see 1Mhz clock signal on Z80, 40010, ..
- I can see what looks like normal activity on address lines
- Swapped the Z80
- Swapped the CRTC (HD46505SP)

After all this, the symptoms remain the same but can be different on every startup: sometime it's black screen, sometime I can see some patterns (cf first 2 pictures)

The one thing very strange I found is the data lines on the CPU seem very wrong (cf picture of scope)

Does anyone have any idea of the next step to try/test ? I'm thinking now that it's either bad RAM or more probably bat ROM or worse bad Gate Array ?

Thank you

(First post here, please let me know if I'm missing something :p)

[robcfg]

Welcome to the CPCWiki!

I take you have no other 6128, or no one nearby with one to swap the Gate Array and check.

Do you have an eprom programmer? That would allow you to check the ROM.

Do you have any expansion that would allow you to use a diagnostics ROM?

And, where did you measure the signal shown on the oscilloscope picture?

Cheers,
Rob

[locked]

Hi ! Thank you for you reply !

Yes unfortunately I don't have other 6128 or other components to swap and check.. I don't have any expansion too :/

I do have an eeprom programmer, I could try to desolder the 40015 and 40025 ROMs and check their contents, yes, I'll do that.

I measured the signal on the DATA pins of the Z80: pins 7 to 15 (except 11 of course) were all showing this pattern.. Whereas ADDRESS pins were showing more "normal" activity (I could take a screenshot of them too).

(also I didn't say it but I have already "deoxit"ed all the sockets)

[locked]

I've managed to dump both ROMs 40015 and 40025, data looks good (40025 matches OS_6128.ROM and BASIC_1.1.ROM and 40015 matches AMSDOS_0.5.ROM), so it's not them..

Is there a way to test the 40010 without replacing it ?

[d_kef]

I'd put my money on bad RAM.
You can try to boot your 6128 from the secondary RAM bank following the instructions here.

d_kef

[Bryce]

That definitely looks like dodgy RAM. You can remove the PAL and use jump leads to swap the upper and lower RAM banks. If it boots, then you just need to replace the lower RAM chips and put the PAL back in.

Bryce.

[Rabs]

- I can see 1Mhz clock signal on Z80, 40010, ..

Just checking, are you sure should be more like 4MHz.

[locked]

Thank you

I did the PAL tweak (cool one), but it didn't help, same issues.

I'm starting to think it may be the cpu that I thought was good that maybe isn't , or that its socket is bad, I need to dig deeper into this (I've ordered a new Z80A). I've just found out that pushing on the cpu sometime changes the sound the computer makes.. (from the speaker and also from the chips).

Either that, the gate array or the 74LS373/244..

What really boggles my mind is that the symptoms are not exactly the same every time I start the computer..

Etienne

[locked]

Concerning the frequency, it is indeed 4MHz as it should be on pin 6 of the CPU, I was looking at the wrong pin

[Rabs]

I've just found out that pushing on the cpu sometime changes the sound the computer makes.. (from the speaker and also from the chips).

You get sound? Is this when pressing DEL?

[locked]

No it's not by pushing del, it's either clic-clic-clic or high frequency (but faint) noise from the speaker

[Rabs]

I am guessing you have checked for Z80 activity (M1, WR and IORQ)?

Have you checked HSYNC and VSYNC on the CRTC? Correct frequencies here would imply the Z80 is initialising the CRTC.


PIN 39 HYSNC 15.62KHz


PIN 40 VSYNC 50Hz



PIN 23 EN (you will only see activity if the z80 is running and issuing IN/OUT instructions).

[locked]

Thanks, I'll have a look

[locked]

Hello,

So it appears the CPU is not initializing the CRTC (or not correctly), here's what I have on the pins:

  pin 39/HSYNC: ~5Khz
  pin 40/VSYNC: nothing
  pin 23/EN: nothing

And on the Z80:

  pin 27/M1: 250Hz square signal
  pin 22/WR: 500KHz signal
  pin 20/IORQ: 5V flat

I can take screenshot of the oscillo screen if it can help

I think I'll change the socket and wait for the new CPU unless someone have a better idea

[Rabs]

I think I'll change the socket and wait for the new CPU unless someone have a better idea

The single leaf sockets can create problems but I have not experience any issues myself.

Worth continuity checking all pins after de-soldering an IC, just in case.

[locked]

It worked !!! 

The socket was indeed not looking great beneath

[Bryce]

Oops, looks like water got in.

Bryce.

[Rabs]

It worked !!!

The socket was indeed not looking great beneath

Well Done, is that it all fixed and running now?

[locked]

I only tested basic, all keys were working fine.

Now the disk drive needs to be fixed, I need to replace the belt.. Do you have any resources to share to do this correctly ?

Thanks again a lot for all the help !

[Rabs]

I only tested basic, all keys were working fine.

Now the disk drive needs to be fixed, I need to replace the belt.. Do you have any resources to share to do this correctly ?

Thanks again a lot for all the help !

Excellent,

Try Noel's Retro Lab Servicing a 3" Floppy Disk Drive

[locked]

Thanks 🙏

I'll buy a belt and hopefully some disks will still be readable ^^