CPC6128 boot issue

deano467 · 20:55, 18 July 23

Hi all, I have a boot up issue with this CPC. The white vertical lines flash on and off then the screen locks up at the photo taken. No input from keyboard.

I have some previous experience repairing spectrums and would be keen to have a go at this, any pointers appreciated

Rabs · 21:43, 18 July 23

Interesting, starts to boot, so almost running. Looks like lower ROM starts but does not get to the Ready prompt, so having problems with the upper ROM. And then there are those lines.

Rabs · 21:55, 18 July 23

There is a diagnostics ROM available Diag ROM.

deano467 · 22:13, 18 July 23

Thanks I have used diag ROM on a zx dandanator before. Is there an amstrad version which has this ROM preinstalled?

Rabs · 22:32, 18 July 23

Quote from: deano467 on 22:13, 18 July 23Thanks I have used diag ROM on a zx dandanator before. Is there an amstrad version which has this ROM preinstalled?

Yes there is. Check ebay for a Dandanator mini. Or you can burn your own ROM and replace the onboard ROM.

Rabs · 22:41, 18 July 23

But I am not convinced it will tell you much in this instance other than confirm the RAM is OK.

Rabs · 23:01, 18 July 23

I had a similar issue with a 664 see 664 issue but the PCB was in a bad state in this instance and no vertical white lines.

Rabs · 23:03, 18 July 23

I had a broken track between IC205 and IC212 on the Disk Controller side of the PCB.

IC205 PIN 2 AND gate input was floating high, should have been held low by IC212 PIN 5.
IC205 PIN 1 AND gate input is connected to A15
IC203 PIN 3 AND gate output is connected to ROMDIS

Every time A15 went high, the ROM was disabled.

So lower ROM worked, upper ROM was disabled. Took me a while...

Note these are 664 IC numbers

deano467 · 23:07, 18 July 23

This CPC itself is in excellent condition, looks like it has been stored in a dry location. Think I'll recap then do a finger test for hot chips at the weekend.

Rabs · 23:34, 18 July 23

Quote from: deano467 on 23:07, 18 July 23This CPC itself is in excellent condition, looks like it has been stored in a dry location. Think I'll recap then do a finger test for hot chips at the weekend.

Not had any cap issues with a CPC so far.

eto · 11:36, 19 July 23

I would expect that a RAM IC of the first RAM bank is malfunctioning. Not to a degree that it halts the whole computer but based on the vertical lines and their colour I would say that D5out is always high.

The CPC will be able to boot to a point where a value read from RAM will lead to a crash.

The easiest way to check if it's really RAM, you can also force the CPC to boot from the secondary RAM bank by removing the PAL chip and replacing it with a socket, where a few pins are connected: https://hackaday.io/project/170949-amstrad-cpc-6128-halmem-tester

deano467 · 12:05, 19 July 23

Quote from: Rabs on 23:34, 18 July 23
Quote from: deano467 on 23:07, 18 July 23This CPC itself is in excellent condition, looks like it has been stored in a dry location. Think I'll recap then do a finger test for hot chips at the weekend.
Not had any cap issues with a CPC so far.

Is there a reset capacitor on the 6128? I know from speccys replacing a failing reset capacitor can get them to boot up correctly.

Rabs · 12:12, 19 July 23

Quote from: eto on 11:36, 19 July 23I would expect that a RAM IC of the first RAM bank is malfunctioning. Not to a degree that it halts the whole computer but based on the vertical lines and their colour I would say that D5out is always high.

The CPC will be able to boot to a point where a value read from RAM will lead to a crash.

The easiest way to check if it's really RAM, you can also force the CPC to boot from the secondary RAM bank by removing the PAL chip and replacing it with a socket, where a few pins are connected: https://hackaday.io/project/170949-amstrad-cpc-6128-halmem-tester

The thing that confuses me is the text. No vertical lines.

eto · 12:32, 19 July 23

Quote from: Rabs on 12:12, 19 July 23The thing that confuses me is the text. No vertical lines.

AH! true... that pretty much invalids my assumption.

But on the "v" of the (v3) there are indeed stripes. As it's basically impossible that it's during output due to the other characters having no stripes, it must be during write?!

Quote from: deano467 on 20:55, 18 July 23The white vertical lines flash on and off then the screen locks up at the photo taken.

Is your screenshot the final version? Looks like it's taken while there is still active screen output. If the boot sequence halts it usually stops after the full line is visible. Is it always identical or do you see slightly different output (with more or less vertical lines inside of the text e.g.) after a reset

deano467 · 12:51, 19 July 23

Quote from: eto on 12:32, 19 July 23
Quote from: Rabs on 12:12, 19 July 23The thing that confuses me is the text. No vertical lines.
AH! true... that pretty much invalids my assumption.

But on the "v" of the (v3) there are indeed stripes. As it's basically impossible that it's during output due to the other characters having no stripes, it must be during write?!

Quote from: deano467 on 20:55, 18 July 23The white vertical lines flash on and off then the screen locks up at the photo taken.

Is your screenshot the final version? Looks like it's taken while there is still active screen output. If the boot sequence halts it usually stops after the full line is visible. Is it always identical or do you see slightly different output (with more or less vertical lines inside of the text e.g.) after a reset

The screenshot was final version, the lines flashed on and off every second or so then settled with them showing in the photo. I will do more testing and checks at weekend and report back

Rabs · 15:53, 19 July 23

Quote from: eto on 12:32, 19 July 23But on the "v" of the (v3) there are indeed stripes. As it's basically impossible that it's during output due to the other characters having no stripes, it must be during write?!

Could it be a lower ROM corruption? The Diag ROM would prove this.

deano467 · 20:42, 22 July 23

I've had a chance to have another look at this. The blue vertical lines come on immediately and flash on and off every half a second or so, eventually stop flashing.

This is what I have done today. Cleaned the switch with contact cleaner thoroughly. It's working 100%.
Removed dust from board. Resoldered DC socket joints
Checked voltages at memory chips and getting 5.07V.
The right hand side bank of memory chips feel warmer than the left side.
The z80 and ROM chips are cool. The HAL chip feels warm and the chip under heatsink is hot but I guess that's why it has a heatsink.

Any other voltage checks I can do to narrow cause. Unfortunately I don't have an oscilloscope.

deano467 · 20:49, 22 July 23

Finally a little bit of life... got a message about overflow in 12497, and it responds to keyboard entry with garbled character

Rabs · 08:36, 23 July 23

That's weird. So the lower ROM is running, sort of, the CRTC is being setup, and maybe even the PIO and AY. Don't understand the vertical stripes and why the text is ok.

If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.

deano467 · 09:57, 23 July 23

Quote from: Rabs on 08:36, 23 July 23That's weird. So the lower ROM is running, sort of, the CRTC is being setup, and maybe even the PIO and AY. Don't understand the vertical stripes and why the text is ok.

If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.

Thanks. Was thinking instead of spending £45 on a dandanator, I could replace the 40015 ROM with one of these in photo. I could have it programmed to switch between the Parados ROM and the diag ROM. When switched to the diag ROM at least it can test the ram. Does this sound reasonable?

eto · 12:18, 23 July 23

Quote from: Rabs on 08:36, 23 July 23If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.

When running the Diagnostics from LowerROm can it still verify the original firmware?

Quote from: deano467 on 09:57, 23 July 23I could replace the 40015 ROM with one of these in photo.

Instead of that thing in the photo, you can also just replace it with a self burned EPROM and select between e.g. Amstrad daignostics and the Fimrware/Basic: https://www.cpcwiki.eu/index.php/DualOS

Rabs · 12:38, 23 July 23

Quote from: eto on 12:18, 23 July 23
Quote from: Rabs on 08:36, 23 July 23If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.
When running the Diagnostics from LowerROm can it still verify the original firmware?

I guess you can only check the Lower ROM if you run the Diag ROM from an external ROM board (i.e. using the Dandanator).

You cannot view this attachment.

deano467 · 12:38, 23 July 23

That would be fine if I had an eprom programmer but I don't. The guy selling that is a respected retro spare parts seller, and his switchable device only costs £7.99... much cheaper than any other source I can find.

Rabs · 12:47, 23 July 23

Quote from: deano467 on 09:57, 23 July 23
Quote from: Rabs on 08:36, 23 July 23That's weird. So the lower ROM is running, sort of, the CRTC is being setup, and maybe even the PIO and AY. Don't understand the vertical stripes and why the text is ok.
If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.
Thanks. Was thinking instead of spending £45 on a dandanator, I could replace the 40015 ROM with one of these in photo. I could have it programmed to switch between the Parados ROM and the diag ROM. When switched to the diag ROM at least it can test the ram. Does this sound reasonable?

I have never used the ROM module you listed. I normally just burn my own EPROM. But I also use the Dandanator. There are pros and cons with both. With your own EPROM you would have to de-solder and remove the on board ROM, which is not without risk and then of course you cannot test the Lower ROM as you have just removed it. With the Dandanator, you don't have to do this but it costs a bit more and you have to have a good working edge connector. But you can use the Dandanator for games once you have the CPC working

Rabs · 13:09, 23 July 23

Quote from: deano467 on 09:57, 23 July 23
Quote from: Rabs on 08:36, 23 July 23That's weird. So the lower ROM is running, sort of, the CRTC is being setup, and maybe even the PIO and AY. Don't understand the vertical stripes and why the text is ok.

If this was me, I would run the Diag ROM next and see what the status of RAM and ROM is.
Thanks. Was thinking instead of spending £45 on a dandanator, I could replace the 40015 ROM with one of these in photo. I could have it programmed to switch between the Parados ROM and the diag ROM. When switched to the diag ROM at least it can test the ram. Does this sound reasonable?

I think the BIOS and BASIC ROM is 40025. The AMSDOS ROM is 40015. See Amstrad Part Numbers. Not sure how the Diag ROM would work if you replaced 40015.

News:

CPC6128 boot issue