SE/30 gray screen with cursor post recap when SCSI drive attached

[Elemenoh]

I have a board that will boot okay to a floppy, but will be stuck on a gray screen with movable cursor if a SCSI drive is attached.

I removed and cleaned under and later swapped out the SCSI IC and checked continuity for most of the SCSI and sound circuit. The one odd thing I've noticed is that RP2 and RP3 (Bourns filters) have odd readings compared to a working board. Across each resistor (ex pins 2 and 19) they register a consistent 50 Ohm, but measuring resistance to ground is very different from a working board. I tried swapping RP3 but the replacement measures the same way. So it must be something up/down stream from the filters.

I'm still hunting for bridges and broken traces, but curious if anyone here has some knowledge about what might cause this issue assuming the filters themselves are okay.

RP2 resistance to ground (Reference board) Pin 12 14M Pin 13 100K Pin 14 100K Pin 15 22M Pin 16 23M Pin 17 100K Pin 18 6M Pin 19 6M	RP3 resistance to ground (Reference board) Pin 12 6M Pin 13 26M Pin 14 17M Pin 15 24M Pin 16 6M Pin 17 6M Pin 18 6M Pin 19 6M
RP2 resistance to ground (Bad board) Pin 12 OL Pin 13 100K Pin 14 100K Pin 15 OL Pin 16 6M Pin 17 100K Pin 18 6M Pin 19 6M	RP3 resistance to ground (Bad board) Pin 12 OL Pin 13 OL Pin 14 OL Pin 15 15M Pin 16 6M Pin 17 6M Pin 18 OL Pin 19 6M

 

[Kai Robinson]

Sounds like a lot of patching needed, potentially rotten via's from leaky caps in the area?

 

[Elemenoh]

Nothing obvious yet but yeah I figure it’s either a broken nearby trace or some failed component. A lot of testing ahead

 

[Elemenoh]

Update:

• Continuity from J4 > J5 > UI12 is OK
• Continuity from UI12 to everything else is OK to all other points (ROM, RAM, VRAM, GLUE etc)
• I replaced UA8 and UB8 since I had some spares and they share address lines. No change.

Though all connections to the ASC (UE10) are okay, I get a garbled startup chime. I didn't care about that earlier figuring I'd deal with it after getting a full boot from SCSI. But that chip also interacts with the same data and address lines. So I'm going to try swapping it next. Have any of you had a failed ASC interfere with SCSI?

And has that IC been reverse engineered or is the only option getting a salvage part?

 

[caver01]

My SE30 had multiple issues—one of them being distorted sound, and occasional crashing when testing alerts in the Sound control panel. I removed the ASC to clean underneath and glad I did. The crashing stopped, but some distortion remained—not evident in the startup chime, at least that I could hear, but other alert sounds were close, but some distortion.

I ended up having SCSI issues + distorted audio. I found a broken D31 from CPU to SCSI chip. Lo and behold, as soon as I bodged that connection and fixed the SCSI, the audio distortion disappeared! So, bad via, broken trace—whatever it was, the problem was D31 which must have been shared by both the SCSI chip and the audio.

 

[Elemenoh]

Huh that's helpful info. The only IC I didn't check continuity to was the CPU since it's a pain and I figured unlikely that path was broken. I'll check that next. Thanks!

 

[caver01]

Yeah, finding a continuity issue—it is always the last one you check! ha ha ha

 

[Elemenoh]

All the data and address lines from UE10 (ASC) and UI12 (SCSI) to the CPU tested okay. Will try swapping UE10 next. Glad to know the traces are likely fine though.

 

[Branchus]

Just to confirm, if you attach a SCSI device but still attempt to boot from a floppy, you get the grey screen, is that correct? So just the existence of a SCSI device on the bus stops it from booting?

 

[Elemenoh]

Just to confirm, if you attach a SCSI device but still attempt to boot from a floppy, you get the grey screen, is that correct? So just the existence of a SCSI device on the bus stops it from booting?

That's correct.

And if no SCSI device is connected, it boots okay from floppy.

 

[Elemenoh]

So today I quadruple checked continuity for all the address and data lines to ASC were fine. Y3 tested OK. Swapped ASC and the symptom persisted.

I checked all of the ASC address and data lines on the oscope. The waveform shapes themselves seems okay compared to a working board. But the bugged board just shows those waveforms repeating steadily whereas the goodboard shows a flurry of activity.

Here’s an example of D28. Any suggestions for what to try next?

 

[Branchus]

Have you swapped out the SCSI chip? If not, I'd be trying that next.

 

[Elemenoh]

@Branchus yeah that was one of the first things I tried. I've also swapped all of the socketed ICs with known-working chips and had the same result.

It's a long shot but I'm wondering if the CPU might be bad.. freezing when the ROM gets to SCSI. I might try that next.

 

[Elemenoh]

Swapped CPU and the Sony sound ICs (UB10,UB11) but no change.

(╯‵Д′)╯彡 ┻━┻