SE/30 video issue that has ME stumped!

[apm]

When you take a screenshot (Cmd-Shift-3), do you see more on the right side of the picture than actually shows up on the screen? That would point to something similar to the other thread. It's normal for the cursor to go almost totally off the screen -- I think you should still see one column when it's all the way over.

 

[bibilit]

Good idea, i will for sure take a screenshot.

 

[asapreta]

Hi,

some questions.

Made some progress, the board is booting from the hard drive once again and the lines have gone.

I have an SE/30 that is not booting from the internal HDD, what did you do to fix the issue?

 

[Bolle]

that... is weird. But its right up the same alley. 

I am wondering if it isnt an address or data wire that has "resistance" instead of a full break, or full short (like it should). 

It may be the clock line that is the issue here.

When testing my multiple card setup I got noise on the screen that looks similar to what bibilit posted. It also comes and goes when CPU load changes.

I narrowed the issue down to the clock signal bugging down when three extension cards are installed. So the issue might as well show up in combination with PDS cards. If the clock output is weak or there is high resistance in the line somewhere the video signal will start to act out just the way as seen here.

 

[techknight]

Actually now that you mention it, I need to watch that signal. Could be noisy. It comes from the PAL thats in the center of the board, next to the main oscillator. Those things only have a gate fan-out of a certain amount anyways on a good day. 

 

[apm]

Interesting. What would be the best way to artificially load it to see what happens? Add capacitance?

 

[techknight]

and resistance. looking at rise times and fall times, plus ringing. 

And comparing those against the datasheet charts under those known conditions. 

 

[trag]

Techknight, did you say that you had tried swapping out the VRAM on that mystery problem board?   I ask because the weird similar-to-AND-gate behavior you're seeing could be caused by certain RAM failure modes.   If you have a leaky structure in a RAM the behavior can be affected by neighboring cells which would look alike AND gate behavior.

 

[techknight]

Yep, I have swapped the VRAM, the UE8, All of the PALs. and I ringed out all the traces, all ok. 

 

[bonsaichop]

Sorry to necrothread but I just had the exact same problem. Turned out that pins 4 &5 on UE8 which go to 22 and 23 of UC7 were shorted on the underside of the logic board directly where the vias came through by a tiny solder blob that made it look like the vias were actually connected. Everything passed my trace tests because I was testing 4 to 22 and 5 to 23 of UE8 and UC27, only by chance did I trace UE8 pin 4 to UC7 23 and realised that there was a short. 

 

[bibilit]

Hi,

I solved this issue on one board i had at home.

The exact same problem, lines coming and going or fixed, and no lines on the floppy.

after spending most of the day removing and replacing chips (UE8, UD8) that looked bad, checking continuities and heating the board (the board worked fine when heating UD8 at first, but not as soon as it cooled down)

In the end, using a spare board i had (destroyed by a Maxell bomb) i piggybacked the PALs, until the picture came perfect.

UI6 was bad, i replaced the PAL for good.

 

[techknight]

I think I changed the PAL set already. But I cannot honestly remember. 

 

[bibilit]

I think I changed the PAL set already. But I cannot honestly remember. 

 

You changed the PAL set, was said on one of your earlier post.

As i am lazy, tried to piggyback the PALs first... hoping it will help troubleshooting the board, and worked.
 

 

[techknight]

Yea its been to long. I bought some new ICs for it and I sat it all on the shelf and I havent looked back yet. been way too busy with my standard day job. I cant wait until we get enough capital in where we an hire dev teams and customer service people so I aint gotta do it. ;-)

 

[bibilit]

@techknight

The SE/30 has two different sets (revisions) of PALs... i was lucky as apparently the one i changed is common in the two revisions.

For troubleshooting purposes, can i use one set instead of the other ?

 

[techknight]

Either set will work. as long as they are all a matched set. 

 

[Bolle]

Even the ones from the early board revision that has the patch wire installed?

Or does that wire make the routing the same as on the later boards?

Never had to work in depth on one of the early boards yet.

 

[techknight]

I havent tried a patchwire board. I usually dont see those. 

I know there is a PAL set thats 6XX and a pal set thats 7XX and I have swapped the two before. 

 

[NE555]

@techknight Did you ever get the issue of this thread resolved?

 

[techknight]

Nope. I have been busy with work so its been on the backburner. Way too long, I know.