Video problem on STE

[Strider]

Hello,

Yesterday I've recapped the PSUs of my 2 STEs.
I attached a photo, the PSUs have been controlled before starting the STEs.

However I have some video problems with these demos showing "true colour" pictures.
Here is an example:



My 2 STEs have the same problem. I also try swapping the PSUs, no difference.
What could be the issue?
Voltage?
I measured 5.2v and 12.16v on the output. Is it too high?
Thanks.

[exxos]

Can you post the program you are using and I can test it on one of my machines.

[Strider]

Yes, this one: https://demozoo.org/productions/180989/
Here is the direct link if you have a hard disk: http://files.dhs.nu/files_demo/wonworld11_hd.zip
or a floppy disk: http://files.dhs.nu/files_demo/wonworld11_dd.zip (needs a MSA to disk conversion though)

I tried to lower the 5V, now it is 5.01V instead of 5.20V, but it makes no difference, video is still buggy.

[exxos]

OK I can confirm same problem.. I get corruption down the right side as well as "random" corruption in the image...

IMG_2350.JPG (5.85 MiB) Viewed 5992 times

My first thought of a "fix" and I got this...

IMG_2351.JPG (6.11 MiB) Viewed 5992 times

The image still seems to have some corrupt pixels.. but its not flicking corruption (might just be bad image data?)

.. and for the next step.. what value resistor pack do you have for P100 (just in front of the blitter)..?

[Strider]

The image still seems to have some corrupt pixels.. but its not flicking corruption (might just be bad image data?)

.. and for the next step.. what value resistor pack do you have for P100 (just in front of the blitter)..?

Before modifying the PSUs I had those corrupt pixels. But never had the flicking corruption.

The P100 shows "M9-1-122J130" on one of my STE.
On the second STE the right part of the P100 is hidden by the blitter, all I can read is "MP-1-122J0.."
Some pictures:

[exxos]

The P100 shows "M9-1-122J130" on one of my STE.
On the second STE the right part of the P100 is hidden by the blitter, all I can read is "MP-1-122J0.."
Some pictures:

Odd.. The macine I tested has corruption, but 10K pull ups, when UDS has 1K on it, the problem goes away.. so 1.2K you have should work..

I also found swapping the CPU for a 68HC00 also solved the problem , so you could try that ( cue bad DMA thread again )

[tzok]

I don't have any corruption (no flickering, no corrupted pixels) on this demo, I have an MC68HC000FN16, and supposedly "bad" DMA (-38) in my STE with an early version (44th week of 1989) of C4003290 main-board. It has an additional silkscreen saying PCB C300779-001 REV 5.1 and this is the one with a factory "piggy-backed" 3rd 74LS164 chip.

[exxos]

ahhh, this was the thread I was talking about somewhere else the other day.. I don't think anyone with the corrupt pixels investigated it