AMSTRAD CPC6128 power up failure

[rpalmer]

Hello fellow enthusiasts,

I have had a second AMSTRAD 6128 from a very long time ago and decided to see if i could get it to boot up. It was originally for cannibalisation of parts for my original, but decided it deserved a second life.

I have re-capped the bipolar electrolytic capacitors (see attached - these are red and black).
I have also replaced the CPU andx 6845.
I have socket-ted the four 74LS153 chips (ICs 105, 104, 109 and 113) next to the 8255. The 74LS244 and 74LS373  (ICs 115 and 114) were also socket-ted.

The screen displays just a yellow colour with short lines which scroll up the screen.

I have checked with a multi-meter (in diode mode) for any breaks/damage between the various chips that were socket-ted and could not find any.

I swapped the Gate Array from my working CPC6128 and found it made no difference and the working CPC still works as expected with the swapped Gate Array.
The RAM was also swapped from a working CPC and it too still works as expected on the original CPC, so the RAM is not the issue.

I have uploaded a youtube video of what happens when I switch it on and can be seen here:


https://youtu.be/QWnNOstGL1A

Any suggestions as to what to look for next would be much appreciated.
rpalmer

[llopis]

Weird. You've replaced most chips already!


Have you tried replacing the ROM? Or even better, using an external ROM (for example X-Mem with the quick and dirty RAM test program, or the new CPC Dandanator)? If any of those works and gives you a clear picture, then it's just your ROM.


If those don't work either... then I'm somewhat at a loss. I'm assuming you've checked Vcc voltages everywhere (they tend to run a bit low on my systems, around 4.3V).


BTW, I love that the random YouTube URL almost spells "nostalgia" :-)

[rpalmer]

Have you tried replacing the ROM? Or even better, using an external ROM (for example X-Mem with the quick and dirty RAM test program, or the new CPC Dandanator)? If any of those works and gives you a clear picture, then it's just your ROM.

If those don't work either... then I'm somewhat at a loss. I'm assuming you've checked Vcc voltages everywhere (they tend to run a bit low on my systems, around 4.3V).

Yep I have a copy of the ROM and that too was tried and still had no joy. I did notice if I boot up with no rom then the same thing appeared to occur, but could not conclude that neither ROM was the problem.

I did check the important chips for voltages, but will check again to make sure they are all within reason.

I do not have X-MEM/Dandanator or Dirty RAM test program, so would be quite trying to go down that path.

I also have "Bitscope" connected to my laptop and can use it as a Logic analylzer which I have been doing to get a clearer picture, but this will take quite some time.

I even tried connecting a keyboard to see if the speaker could be made to produce sound (by using the backspace and by just holding down any key), but that too was no joy.

rpalmer

[rpalmer]

LLopis,

I checked the voltages and find all chips are getting between 4.8 and 4.9 Volts, so power seems to not be the problem.

The bitscope logic analyzer shows the NRAS/NCAS and NWE signals from the Gate Array are changing, suggesting the memory is being accessed. I cant tell if this is okay, but will investigate further.

I even tried socketting IC 110 (74LS132 between the 8255 and 6845). My thinking was the CRTC could not be selected as the Enable signal (Pin 23) was always low via bitscope suggesting it was faulty. Alas that was not the case as the new one also did not fix it. I even socket-ted IC 112 and it too did not fix the problem.

My next step was thinking the output from the Gate Array was faulty, so I changed the resistors and again no joy.

I even tried swapping the PAL from my working 6128 and it did not fix the problem. The original with the changed PAL was not affected, so the PAL is okay.

rpalmer

[rpalmer]

Finally got the motherboard to show power text, Wooohoo.
I had found by accident that when i pressed down on the OS/BASIC ROM (during voltage measuring of the CPU) there was text on screen which corresponded to the boot screen (although garbled). After replacing the socket and testing there was no joy until I removed the DISC ROM/socket and viola there was the power on screen text as see in attached zip file. I even tested with just the DISC ROM removed and no proper boot was the result, hence the need to remove the socket. I also checked the traces for any breaks and could not see any, so cannot explain why the socket was a problem and may be this is one where the break is to small to see.

Now need to investigate why the text is garbled and why there is no "READY" prompt (any ideas here??)

The youtube link to the power on test video is here:
https://youtu.be/FXVE0TobpPw
I also replaced the original ROM with a new one (see attached photo of it) and found both the original and new ROMs show the same screen on switching the power on. This means the ROM is okay as far as I can tell. I also tested the system reset via the expansion board and the same occurs. See photos attached.

rpalmer

[rpalmer]

here are the pix

[Bryce]

The Ready is missing because you've removed the AMSDOS ROM. It needs to be in place for the boot sequence to get to the Ready prompt.


Bryce.

[rpalmer]

The Ready is missing because you've removed the AMSDOS ROM. It needs to be in place for the boot sequence to get to the Ready prompt.


Bryce.

Sorry to say bryce, but that is very very wrong. I can get my working CPC6128 to boot up with no DISC ROM (See attached file).
rpalmer

[rambler]

I have noticed that you have the key sequence: 31*KJFDA
They are all on the same row of keys: f3 - f1 - - * -  K  J - - D - A


Ref: wiki http://www.cpcwiki.eu/index.php/Programming:Keyboard_scanning bit 5

[Bryce]

Sorry to say bryce, but that is very very wrong. I can get my working CPC6128 to boot up with no DISC ROM (See attached file).
rpalmer

Oh, maybe I'm thinking of the µPD765 then. I remember that one of the main chips for the floppy circuitry stops the computer from booting if it's removed. I thought it was the ROM (but I'm getting old ).


Bryce.

[llopis]

I also checked the traces for any breaks and could not see any, so cannot explain why the socket was a problem and may be this is one where the break is to small to see.

If it was fixed removing a socket, maybe you had a bad soldering job and you were shorting two pins? Pushing on the chip was bending things a bit and breaking the short perhaps.


The garbled screen is usually a symptom of a bad CRTC. Now that you've made progress, try replacing it again with a known working one. Then, if that's still not working, check the soldering on the CRTC socket for shorts and other suspicious joints.

[Bryce]

Never heard of a bad CRTC giving a garbled screen. Usually you get a completely black screen if anything isn't working inside the CRTC.

Bryce.

[llopis]

Never heard of a bad CRTC giving a garbled screen. Usually you get a completely black screen if anything isn't working inside the CRTC.

Really? Maybe I was the GA I was thinking then. Like sometimes it'll show you the startup letters, but in the wrong screen mode, or half way over in the screen.

[Bryce]

Yup, that sounds more like a GA failure.

Bryce.

[rpalmer]

I have test the Gate Array and find it is not faulty when put into my original CPC.

rpalmer

[Audronic]

Hi Ray P


Have look at the link, I had a similar problem look at around  Reply #14


http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/repairs-to-a-cpc6128/msg113304/#msg113304


Ray W

[rpalmer]

Hi Ray P


Have look at the link, I had a similar problem look at around  Reply #14


http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/repairs-to-a-cpc6128/msg113304/#msg113304


Ray W

Ray,
I have tested the PAL with the one from my working CPC6128 and then with the RAM swapped from my working CPC as well. Both tested with the same result.

I tried with the PAL removed and linked up the required pins as per schematic to no avail. This resulted in just a box on the screen and no text.
I will next test with the suggested changes in the link supplied.
rpalmer

[rpalmer]

RayW
I have tried the test connection as per the suggest link and get the exact same result.
I have noted it appears to display the message twice. I have no idea why or how this is done (maybe stack corruption however that would result in endless loop).
I even tried a test with no 765 IC and it made no difference.
I also considered that I socket-ted the 74LS153s (all 4 of them) might be the cause. I concluded that they are okay as if one was bad then the screen would be completely garbled with bad addresses for screen data.

rpalmer

[Audronic]

RayP


Have you swapped the AY-3-8912 (IC 102).
Thats about all thats left ( No Beep )


Ray W

[rpalmer]

Ray W,

I do not have a spare sound chip, so that test is not possible.

rpalmer

[rpalmer]

Hello everyone,
I have made some progress is getting something to work on the 6128 motherboard. Some good and sadly some not so good.

The good:
I can boot up to a point where the cursor is displayed and can use the the keyboard.
The not so good:
- The screen is still garbled (much like before) and now display a "Z" and then the block square for the cursor.- Not all keys are being detected.
I have socket-ted more of the 74xxx chips and looks like I will likely have to do the remaining one at a time til the problem goes away.
The link to the youtube video is:

https://youtu.be/GS41OaCQA-o
rpalmer

[Bryce]

Have you checked the reset circuitry? Possibly the computer is booting too fast for some of the ICs to be ready?

Bryce.

[rpalmer]

Have you checked the reset circuitry? Possibly the computer is booting too fast for some of the ICs to be ready?

Bryce.

To fast.. wow I could never think an AMSTRAD as being too fast...LOL
Bryce, do you mean the 74HCU04 inverters are faulty?
rpalmer

[Bryce]

Maybe... The 74HCU04 is doing two things. One of its gates (Pins 1 and 2) are converting the clock signal to TTL levels. The other function is the reset circuitry (with the help of IC110, 74LS132). The purpose of the reset circuitry is that it hold the reset signal low for a few milliseconds at startup, so that all ICs have time to initialise and all start at the same time. If any part of this isn't working, then the reset may not be bieng held low at all, or for too short. Check Pin 6 of IC110 and see how long after switching on that it stays low.

Of course you can check it even easier: Just short the reset pin to ground for a second, if the CPC boots properly, then the reset circuitry isn't doing its thing at startup.


Bryce.

[rpalmer]

I have tried the reset to ground and it made no difference.
I have replaced IC110 with a socket and new chip. The 74HCU04 is still socket-ted to the board and I do have a new to replace it.
rpalmer