SE/30 CRT / Display / Video Issue

[pgreenland]

Hi All,

I've recently joined the classic mac owners club, with three fixer upper SE/30's in various states of decay.

Having selected the best logic board of the three, with no leaking caps or exploded battery. I've cleaned and recapped it. Not yet recapped the analog board or PSU but have checked the rail voltages, which all seem pretty close.

It boots happily from floppy to MacTest SE/30 and seemingly passes all the tests which don't require additional hardware (serial loopback cable for example).

There's something funky going on with the video hardware or CRT that I was hoping someone might be able to give me some pointers on.

Here's the CRT at the MacTest desktop.



The desktop is visible under whatever's going on. Of the three systems I have two CRTs that work and display largely the same thing.

I dropped a logic analyser on the video, hsync and vsync signals and wrote a bit of software to decode them.

Leading to the following:



It could well be my questionable quality python, but there are similarities between what's shown on the CRT and the image extracted from the raw video signal present on J12.

Thought there might be an issue further upstream, so moved to the video ram data lines, shift register load line and vsync. Some minor python adjustments and I ended up with:



To my amazement the video ram does indeed contain a valid image.

So I'm thinking the problem lies somewhere between the video ram and the signals present on J12 for the CRT. As well as possible other problems on the analog board that I've got even less knowledge on than the video generation hardware.

Any pointers / advice would be much appreciated.

Thanks,

Phil

 

[Johnnya101]

I -think- someone else had a very similar issue and I thought it was UE8. They had the same style lines that showed in a screenshot. Let me see if I can find it.

Edit: can't find it. Maybe someone knows what I'm talking about. If it wasn't UE8 it was something around there.

 

[mg.man]

IIRC, UE8 is one of the buffer(?) chips, and is notorious for collecting cap goo underneath, and sometimes hiding bad traces. Sadly, I don't (yet?) have surface mount hot air equipment or skills, so I've had to leave that effort to other MLArs.

If you have the skill/equipment, I'd try removing UE8 and inspecting / cleaning.

Good luck and please report back.

 

[pgreenland]

@Johnnya101 and @mg.man thanks both for your input.

It seems UE8 is a parallel to serial shift register....another couple of pins on the logic analyser started to make it look pretty suspicious. The parallel loads on the input didn't match the serial stream on the output.

Took it off as you suggested, tracks under it all look ok after a bit of a clean up:



Swapped it for the one on my rusty / crusty donor board and we're in business.



First time I've ever seen that adorable little CRT displaying anything meaningful.....now I just need to learn French

@mg.man I'm in the UK too over near Bath. I'm far from an expert at hot air rework, but I can pretty much get a SOIC off and back on without doing too much damage. If I can help with any of your boards in the future please let me know. I'm very much a newbie when it comes to hardware that was released around the time I was born . It's been fun coaxing it back into life so far. I was amazed by how familiar it feels, compared to the MacBook that the logic analyser was hooked up to. Aside from a tinker in the emulator, this is the first time I've got my hands on a classic mac.

Seems to pass all the tests that don't involve SCSI hardware or serial loopback cables now.

Installer still won't boot, complaining about a bus error....kinda hoping that's related to my lack of a hard drive. Considering the MacTest disk hasn't choked while I was poking around all over it.

Ordered a ZuluSCSI, see if it'll get me to the next stage. Then its recap the analog board, PSU and try to clean / grease up the floppy a bit. Managed to get it reading disks, but it needs a helping hand loading or ejecting disks.

Again thanks for the advice, another issue solved.....not the worst way to spend a Sunday

Phil

 

[cheesestraws]

I'm in the UK too over near Bath

Hello from Salisbury!

And congratulations on the fix. Good tricks with the logic analyser there—what do you use?

 

[jajan547]

Did you try different RAM Simms, had a similar issue about 2 years ago.
Edit: this was resolved while typing lol

 

[pgreenland]

Hello from Salisbury!

And congratulations on the fix. Good tricks with the logic analyser there—what do you use?

Howdy neighbour

It's a Saleae Logic. They're a bit on the pricey side for hobby use unfortunately, but they're very good. That one's generally doing mundane snooping on SPI and I2C busses at work. Here's yesterday's operating theatre.

 

[pgreenland]

Did you try different RAM Simms, had a similar issue about 2 years ago.
Edit: this was resolved while typing lol

I didn't, although saw it suggested somewhere online in a "so the video stopped working after you changed the RAM" sort of situation. If the video RAM didn't have the correct image in I'd have tried swapping it out for some spares. Thanks for the input though! Its really encouraging to see that there's still a community of folks helping with these old machines .

 

[mg.man]

Swapped it for the one on my rusty / crusty donor board and we're in business.

Cool! Glad you got it working, and, that you've added some pretty clear issue / cause data. I guess I need to drag that SE/30 'board back onto my bench. Thanks for the offer of assistance - I may take you up on that!

And greetings from Oxfordshire.

 

[mg.man]

I guess I need to drag that SE/30 'board back onto my bench.

Well... I finally had a chance to do just that today. But before I took the drastic action of removing and replacing UE8, I decided, spurred on by @Bolle's post over in another thread :

Broken address line somewhere around the ASC. A0, A1 and A2 pass along under the ASC close to that one leaking cap that also kills UE8 ever so often.
If an address line breaks there you will get faulty audio and video.

...to do some more buzzing around. This was also prompted by the fact that I'd buzzed all the connections from UE8 a week or so back and they were fine. Well... whaddayaknow... @Bolle was right - I did have issues with A0 and A1 connectivity - to various places including the Video ROM, ASC and GLUE. Here's my solution :



NO, it's not pretty - I *will* tidy this up. But... just to be clear... TA DA :



NO MORE video corruption!!! Big thanks to @Bolle and the rest of the MLA community!

 

[pgreenland]

Well... I finally had a chance to do just that today. But before I took the drastic action of removing and replacing UE8, I decided, spurred on by @Bolle's post over in another thread :

...to do some more buzzing around. This was also prompted by the fact that I'd buzzed all the connections from UE8 a week or so back and they were fine. Well... whaddayaknow... @Bolle was right - I did have issues with A0 and A1 connectivity - to various places including the Video ROM, ASC and GLUE. Here's my solution :

View attachment 43401

NO, it's not pretty - I *will* tidy this up. But... just to be clear... TA DA :

View attachment 43402

NO MORE video corruption!!! Big thanks to @Bolle and the rest of the MLA community!

Nice one @mg.man !

Did you happen to take a before picture showing the corruption? Could be useful for others to see what failed address lines look like compared to other problems

 

[mg.man]

Did you happen to take a before picture showing the corruption?

I did, but it was buried over in another thread (which I've also updated) :

Troubleshooting SE/30 garbled screen, no chime before and after recap

Any ideas? I didn't replace the axials because they still look good. I have a working SE, can I put the SE/30 board in the SE to see if it works there? Are they interchangeable?

68kmla.org