News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu

6128 screen has lost vertical sync?

Started by nockieboy, 17:59, 30 April 18

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

nockieboy

Hi everyone,


I'm pretty new here, so go easy on me.  ;D


I've had a CPC6128 in the loft after buying it off eBay many years ago.  After a house move, I decided to get it out and dust it off and see if it worked.  The drive was fine, everything worked just great with the CTM644 monitor I used with the CPC464 I have as well, picture was spot on.


However, a few weeks back I turned the 6128 on and the screen was fuzzy and rolling vertically upwards.  I could just about make out that the Amstrad BASIC message was there and could see text I was typing appearing as a stretched-out line of disintegrated pixels on the rolling screen, so the computer itself seems okay - the problem is restricted to the display circuit somewhere.


I tested the monitor with the 464 and the picture was fine, so it's not the monitor.


I've tried adjusting the vertical hold at the rear of the monitor, but this doesn't help.


I've taken the 6128 apart and checked the top and underside of the PCB for broken and dry solder joints around the monitor connection, but there's nothing of concern there.  I couldn't see any obvious blown caps, either.


So does anyone have any ideas what this could be and what I need to do to fix it?  Any help would be appreciated.  :)

Bryce

Hi and welcome,
           Have you checked (and cleaned) the contacts in the monitor socket of the 6128?

Bryce.

nockieboy

#2
Quote from: Bryce on 10:36, 01 May 18
Hi and welcome,
           Have you checked (and cleaned) the contacts in the monitor socket of the 6128?

Bryce.


I've done a visual inspection of the contacts - they look okay - but I'm not sure how I'd clean them... I guess I'd need a toothpick and some really fine emery paper or something.


Here's a pic:





If I was forced to pick something, I'd say the bottom three contacts don't show up as well as the top three, but taking the monitor apart to get at the other end of the cable and testing for continuity through the cable and connector seems a bit of a long shot..


EDIT: ... but buying a 6-pin DIN plug off eBay will allow me to test the socket connections much more easily.

Bryce

Just spray it full of contact cleaner, no more is needed. Also check that none of the socket pins have a dry joint on the PCB.

Bryce.

nockieboy

#4
Quote from: Bryce on 16:17, 01 May 18
Just spray it full of contact cleaner, no more is needed. Also check that none of the socket pins have a dry joint on the PCB.

Bryce.


Hmm... nope... that hasn't improved anything.  Here's a screen shot of what I'm seeing.  Will see if I can upload a short video clip too.





EDIT:
Video clip here


EDIT 2: I checked the PCB for dry solder joints before posting for help - the PCB looks to be in good order.

arnoldemu

To me that looks like horizontal sync.

are any of the connections around the crtc or the gate-array dodgy?
check the hsync and vsync going to the gate-array are ok.

My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

Yup, that looks like a dodgy connection at the Gate array socket or a failing CTRC.

Bryce.

nockieboy

#7
Quote from: arnoldemu on 20:07, 01 May 18
To me that looks like horizontal sync.

are any of the connections around the crtc or the gate-array dodgy?
check the hsync and vsync going to the gate-array are ok.


Thanks for the suggestions.  I've got the 6128 apart again now and can check the chips on the PCB.  The connections to the CRTC and gate array look okay, but what pins should I be checking on them with my sillyscope?

EDIT: nm - found a pinout here on cpcwiki.  :doh:  Is there anything I should be looking for in particular?  My electronics knowledge is basic and all self-taught...  ::)


Bryce

You should definitely test pins 39 and 40 of the CRTC. If either of these are missing then the CRTC is bad.

Bryce.

gerald

Some thoughts :

- The video sync going to the monitor is a composite one. That is, horizontal and vertical are sync combined.
- The horizontal sync coming from the CRTC is used to generate interruption to the Z80. The FW uses the interruption to do the keyboard scan and the sound queue management. If the keyboard and sound is fine (del key test), then the CRTC is fine and its connection to the GA also.
- The composite video sync is going straight from the gate array to the DIN connector via a  220 Ohm resistor (R137).
If you have a scope, you should have something like this with 5V amplitude on pin 5 of the gate array (40010): https://en.wikipedia.org/wiki/File:Videosignal_vsync.jpg

I have a GT65 monitor that is exhibiting similar issue (no sync) when powering the CPC, but not when the CPC is powered with an external supply. I've never tried to diagnose the exact problem (I've no use for a monochrome monitor), but it could be that your monitor is the problem, not the CPC.

nockieboy

Quote from: gerald on 13:24, 11 May 18
Some thoughts :

- The video sync going to the monitor is a composite one. That is, horizontal and vertical are sync combined.
- The horizontal sync coming from the CRTC is used to generate interruption to the Z80. The FW uses the interruption to do the keyboard scan and the sound queue management. If the keyboard and sound is fine (del key test), then the CRTC is fine and its connection to the GA also.
- The composite video sync is going straight from the gate array to the DIN connector via a  220 Ohm resistor (R137).
If you have a scope, you should have something like this with 5V amplitude on pin 5 of the gate array (40010): https://en.wikipedia.org/wiki/File:Videosignal_vsync.jpg


Ah, interesting... Well, it seems the keyboard was working - I couldn't read what was appearing on the screen thanks to the lack of hsync, but characters were definitely appearing.

Quote from: gerald on 13:24, 11 May 18
I have a GT65 monitor that is exhibiting similar issue (no sync) when powering the CPC, but not when the CPC is powered with an external supply. I've never tried to diagnose the exact problem (I've no use for a monochrome monitor), but it could be that your monitor is the problem, not the CPC.


Not so sure about this - as I mentioned in the OP, I also have a CPC464 which displays just fine with the same monitor, excluding pretty much all monitor failures I can think of.  I did just go back and plug the 6128 into the monitor without the 12V connector, though, to match the power draw of the 464 more accurately (in a rough 'n' ready way), but the display was still corrupted.


I have an oscilloscope, but won't get the chance to test the 6128 until late next week unless I get any spare time before then.  Will post an update when I can.

nockieboy


Disclaimer:  I'm not an electronics expert.  I just about know which way around batteries are supposed to be fitted.

Okay, managed to get the oscilloscope running.  Haven't got pics unfortunately as I only have two hands to hold the probe, ground lead, etc.  But from testing the MC6845P, it seems to be putting out reasonable hsync and vsync signals.  At least the sillyscope is able to lock on to a repeating signal.


I've also taken a look at the gate array - in my 6128, it's an Amstrad 40010.  HSYNC (pin 11) is putting out a nice clean signal, looking something like this:


   _      _     _
  /      /      /
|      |      |
|      |      |
-    ---    ---   ---  etc


(Apologies for the crap ASCII art representation...)


HOWEVER... VSYNC (pin 13) isn't outputting any discernible signal that I can lock on to with my oscilloscope - I'm not expert, as I mentioned at the start, but it sure looks like there's nothing but garbage coming out of pin 13.


I will try again later if I can get the other half to take some pics for me, but it seems the 40010's VSYNC may have died..?

Bryce

What's the make/model of the scope you're using?

Bryce.

nockieboy

Hitachi V-212.  Simple 20MHz 'scope, nothing special.

Bryce

Ok, so it doesn't have an "Auto" button which would try to make sense of the V-Sync. Is the V-sync pulsing at all or can you better describe the garbage? Maybe even take a "screenshot" (photo of the scope screen).

Bryce.

gerald

Quote from: Bryce on 13:51, 17 May 18
Ok, so it doesn't have an "Auto" button which would try to make sense of the V-Sync. Is the V-sync pulsing at all or can you better describe the garbage? Maybe even take a "screenshot" (photo of the scope screen).

Bryce.
But it has a TV-H and TV-V trigger mode and that should be enough  :)

Bryce

So it does (just checked the manual online)! That should work.

Bryce.

nockieboy

Okay, apologies for the delay, have been really busy.  Got a couple of pics for you, but the quality isn't great as I needed four hands really to hold the probes and the phone to take a picture!  I've tried to get the time and volts settings in too, so you have a better idea of context.


This first image is the HSYNC (pin 14) on the 40010:






And this one is the VSYNC (pin 15) on same:





Had to turn the timebase right down to get that last trace and the oscilloscope couldn't lock on to it properly - the broken line was constantly moving to the right.


I'm no electronics buff, but I'm trying to learn - though I don't really know what I'm doing with the scope so feel free to provide explicit instructions if you want further images!  ;)




nockieboy

Quote from: gerald on 16:11, 17 May 18
But it has a TV-H and TV-V trigger mode and that should be enough  :)


Ah darnit, just seen this - will have to give it a go and see if I can get a better pic of the VSYNC..

nockieboy

Erm.. further to the above, still haven't been able to re-test the 40010 with the oscilloscope set up properly yet, but it does seem as though the 40010 ULA could be at fault.  I'm just wondering - if it IS the 40010, where can I get a replacement?  There appears to have been the odd seller on eBay and elsewhere, but they're all old (and now defunct or out of stock) listings.


I read somewhere that the 40010 was reverse-engineered and it was possible to burn your own PLA or ULA as a replacement? Anyone know anything about that?

Bryce

I've never heard of the 40010 being reverse engineered. As far as replacements are concerned, ToTO may still have some for sale.

Bryce.

Duke

Quote from: Bryce on 08:26, 17 July 18
I've never heard of the 40010 being reverse engineered. As far as replacements are concerned, ToTO may still have some for sale.
gerald did http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/gate-array-decapped!/msg133636/#msg133636

If anyone has tried to make a replacement, I don't know.

Bryce

Oh yeah, forgot that one, but it doesn't look like anyone has ever confirmed its function.

Bryce.

gerald

Quote from: Bryce on 09:45, 17 July 18
Oh yeah, forgot that one, but it doesn't look like anyone has ever confirmed its function.

Bryce.
Validation process stared a while ago ;)
But I found an issue on the re-implementation running some demo, and still did not find the time to find the issue  :-[

nockieboy

Quote from: gerald on 16:24, 17 July 18
Validation process stared a while ago ;)
But I found an issue on the re-implementation running some demo, and still did not find the time to find the issue  :-[


Ah hot damn... looks like I'll have to source an original, and the few expired (but cached by Google) sales pages I've found online seem to show a steep rise in prices of the chips.  Think I might as well look for a replacement 6128 entirely...  :-[


In the meantime, I'll try to find the time to pin down exactly where the problem is, but it's looking likely to be the 40010.

Powered by SMFPacks Menu Editor Mod