Re-capped SE/30 still with bad video

[clc]

Hi all,

I've owned an SE/30 for a number of years now. A few months ago, it started having vertical garbled bars on the screen. This sounded a bit like the Simisai Mac, so I opened it up and replaced all 11 electrolytic capacitors (which were exploded all over) with tantalum SMDs. I also cleaned the board, reseated the video ROM, reseated the main ROM, reseated all SIMMs, and replaced the 3.6V battery.

Powered it up... and... no change. The video is still bad.

The strange thing is that the computer functions fine. It boots up, mouse moves, I can run programs, audio works fine, etc. It's just the vertical bars of garbage. But when I move the mouse around, the garbage is repainted with copies of the good areas (see the image around the mouse pointer, where I wiggled it around).

Any ideas what this might be? Almost sounds like bad video RAM, but where is that?

Thanks,

Chuck

 

[bigmessowires]

Looks like one of the address lines to video may be bad, or shorted to another line. There are 32 vertical bars, each 16 pixels wide, which is 2 bytes. So check line A1.

 

[WhoaMan]

how much memory does your se/30 have? what size and location are each of the SIMMS? it sounds to me like one of your paired SIMMS has died or needs to be reseated.

 

[clc]

So far, I haven't found a short. The board looks clean (any corrosion from the leaked caps appears to be superficial and cleaned up nicely) and I don't know enough about the board to try to track down a short with an ohmmeter.

The board has 4 4 MB SIMMs (in banks 1 and 2) and 4 1 MB SIMMs. I have tried using only the 4MBs, and then only the 1MBs (each time, having the 4 SIMMs in the first two banks). The bad video remained unchanged. I also tried reseating the 5 PALs (UI6, UG6, UE6, and two others whose labels I can't see right now) next to the SIMMs.

The video "badness" seems very digital in nature, so I've been looking on the motherboard. Is there any common problem I should check for with the vertically-oriented power/video PCB?

Sadly, I don't have another SE/30 to swap parts with. :'(

Thanks for the hints! :beige:

 

[clc]

Looks like one of the address lines to video may be bad, or shorted to another line. There are 32 vertical bars, each 16 pixels wide, which is 2 bytes. So check line A1.

bigmessowires, your suggestion to check A1 makes perfect sense. Thanks. But I'm still a noob on this board... this clearly isn't A1 to the system RAM (right?), because otherwise the machine wouldn't run at all. Where would I find the video bus's A1?

 

[mcdermd]

I like this thread. I have one SE/30 with a similar but different issue that recapping had no effect on. Maybe there's hope yet.

 

[bigmessowires]

I realized after I posted about A1 that it doesn't quite make sense, because if your address lines were shorted, the computer wouldn't work at all. But if the A1 output from the video address generator were failing somehow, that might do it. On the Mac Plus, the video address comes from some counter chips and a buffer. I don't know about the SE/30-- probably it comes from one of the custom ASICs? Maybe the A1 trace to that ASIC was partly burned away by capacitor goo, or the chip itself has gone bad? Or if the chip is socketed (doubtful), maybe it needs the pins cleaned and resocketed?

Sorry I don't know exactly where to look. I dimly recall seeing that someone posted SE/30 schematics on another thread not long ago, though.

 

[WhoaMan]

just to make things easier, here is the link. i believe these were the ones you are talking about?

https://docs.google.com/open?id=0B3mQLg8d1xThNzQxMTA5ZGQtNmNlMy00MWUwLTg3YWMtMmE3MTc3ZjQxZTYz

 

[clc]

just to make things easier, here is the link

Great! Thanks... I love having schematics for my machines.

I've freshened the solder on the VRAM (the two chips near the SIMM banks, labeled 41264). Then I've been checking continuity according to the schematics (especially SE30_P5.GIF) for UE6, UE7, UG6, UG7, UE8, and UF8. So far, everything checks out okay. The MESS project has a page -- http://mess.redump.net/mess:driver_info:mac_technical_notes?s%5B%5D=rom -- that has some good information about the purposes of various chips. That and the schematic helped me focus here.

I like this thread. I have one SE/30 with a similar but different issue that recapping had no effect on. Maybe there's hope yet.

My project-finishing-latency is legendary (real life intervenes) but if I do find the culprit I'll be sure to post. In the mean time, I'm on the look-out for an SE/30 that I could use for parts and testing....

 

[WhoaMan]

After taking a look at the schematics for a few minutes, I would suggest checking out the 4 74253 mux's. for problems. I assume UF8, UG6, and UG7 are working correctly as the system is getting its vblank and hblank signals. I also assume UE8, UC6 and UC7 are working correctly as UE8 (shift register) is shared between the two RAM chips and each RAM chip is outputting 4 bits....

Chips of interest:

PALs:

UE6

UE7

Mux:

UD8

UB8

UC8

UA8

bigmessowires' A1 assumption for the 16 pix of garbage does sound correct when you consider that the VRAM and RAM are seperate.... so i would check out pins 9,10,11,12,13, and 15 on UD8

 

[techknight]

I agree. A Failed 74F253. which one? not sure. but thats my feeling. one of the address lines as indicated earlier appear to be inop. the multiplexers would be the first thing that would cause this. I guess an easy way to check is to flip around the mux ICs and see if the positioning of the video corruption moves...