I recently received an SE/30 that was sold as working. Without trying the machine first, I broke it down, recapped it, and ultrasonically cleaned the motherboard. Unfortunately, upon powering up for the first time, I was greeted with the Simasima display.
It turned out that there were several problems with the motherboard. I suspect that many traces and solder joints were hanging by a thread, and that cleaning and working on the board finally caused them to disconnect. That is, the cleaning process was not at fault, it just exposed the reality that the board needed repair.
1. Simasima display but no chime
I recapped everything, including the analog board, power supply, and motherboard. I measured the power supply voltage and swapped in a known-working motherboard to verify that the power supply and analog board were fully functioning. I swapped the 'bad' motherboard's CPU, RAM, Video ROM, and ROM SIMM to the known-working motherboard to verify that those parts all worked.
As I was verifying the main oscillator, I twisted the board to get a good angle for the probe. In doing so, the machine chimed and started working! Twisting it back, the machine stopped working. Laying it flat, pressing on the power connector or UE10 caused it to work again.
I removed and cleaned the power connector and UE10. No improvement. Pressing on them causes the machine to work. I then lazily reheated the pins of many chips nearby UE10 (which turns out to have caused the next issue).
I measured the oscillator (Y2) and noticed it was only active when I pressed down. I replaced Y2 without improvement. I carefully measured everything going into Y2, and discovered power (5V) disconnected and reconnected when I pressed down, but power at the main motherboard connector stayed solid. Ah ha! A cracked trace or component leading to Y2.
By measuring the continuity each step of the way until Y2, I found inductor L13 was cracked at one end. I replaced it from a battery bombed SE/30.
2. Still Simasima display but no chime
After repairing L13 and seeing a valid clock, the computer still did not chime. Worse still, pressing down no longer caused the computer to start working.
I turned to the comprehensive Simasima repair guide:
I followed the guide and measured all the clocks and the reset lines. Everything was good.
Then, I learned something wonderful. I never knew the 68030 walked the address space, and in doing so, provides easy to follow frequencies to detect broken or shorted address lines. Thank you Simasima Repair Guide page 46.
Carefully measuring them revealed address lines A12 and A5 were combined. Nothing is going to work with bad address lines. Consulting the Bomarc schematic, I found that A12 and A5 are used on video chip UB8. That's one of the areas I reflowed lazily.
Turning the board upside down revealed a solder bridge. Solder flowed out the bottom of a hole and bridged to the next hole. Oops.
Removing the blob with desoldering braid resulted in a happy Mac chime!
3. Double vision. Multiple images.
What the what?
Multiple copies of the startup icon and mouse cursor appear on the screen. Lots of flicker. However, the horizontal and vertical size and placement of the screen is good. And, the images aren't stretched or condensed.
Fortunately, a post on the other forum by @KhanTyranitar and @Bolle provided a really strong clue.
68kmla.org
Pins 2 and 14 are shared for UA8, UB8, UC8, and UD8. On my motherboard, there was an imperceptible break in the trace for pin 2 of UB8.
Adding a small wire to connect the trace fixed the multiple image problem.
4. No SCSI
At this point, the computer had a perfect video image and happy startup chime. So, I was already pleased with the progress. The computer boots from a Floppy EMU but not from SCSI.
Well, this is obvious. UI12 (the SCSI chip) is near leaky capacitors C8, C9, and C10.
Learning from my earlier mistake, I decided to completely remove the SCSI chip with a hot air gun, to make sure everything was fully cleaned. Boy! I'm sure glad I did. It was gross.
After cleaning and reinstalling the SCSI chip, the SE/30 happily booted from the SCSI drive. It ran and passed all of the Mac Test Pro tests.
5. Long time to chime
There was still one oddity. It took at least five seconds after turning on power until the happy chime played and the screen was initialized. That is, the SE/30 sat for a while with a Simasima screen and then everything was fine. I assumed the ROM was waiting for the disconnected SCSI chip -- but that was now repaired.
Using the Simasima guide, I measured both reset pins (pin 5 on UB10 and pin 5 on UB11) on the Sony chips. They are distinctly separate. UB10 responded almost immediately after power up, but UB11 did not. Finally, UB11 changed state and that's when the Mac chimes.
During this diagnosis, I noticed pressing the reset button (S1) wasn't doing anything. Measuring the resistance of S1 indicated a steady 300ish ohms whether it was pressed or not. Clearly the reset button is damaged. As it was desoldered, a lead just fell off due to corrosion. I replaced it with a new old stock button.
6. Still long time to chime
Replacing the bad reset button did not fix the delay until hearing the startup chime. I check all the traces and nearby capacitors for continuity. Finally, I figured if the reset button was internally corroded, perhaps UB11 was as well. I desoldered it, cleaned and checked underneath, added a DIP socket, and swapped in UB11 from a battery-bombed board.
With that, this SE/30 board is fully functioning.
Perhaps you've noticed a lot of bad spots of corrosion on traces in the above photographs. Also, there are areas that still don't look clean. This is one of the unfortunate realities of some Simasima-afflicted boards. If there is enough corrosion to cause functional damage, then there is likely weaked traces and chips nearby. I chose to fix this board anyway, as I felt the practice would teach me many possible causes and symptoms. So, even if this motherboard is not one that I would rely on, it was worth the repair.
- David
It turned out that there were several problems with the motherboard. I suspect that many traces and solder joints were hanging by a thread, and that cleaning and working on the board finally caused them to disconnect. That is, the cleaning process was not at fault, it just exposed the reality that the board needed repair.
1. Simasima display but no chime
I recapped everything, including the analog board, power supply, and motherboard. I measured the power supply voltage and swapped in a known-working motherboard to verify that the power supply and analog board were fully functioning. I swapped the 'bad' motherboard's CPU, RAM, Video ROM, and ROM SIMM to the known-working motherboard to verify that those parts all worked.
As I was verifying the main oscillator, I twisted the board to get a good angle for the probe. In doing so, the machine chimed and started working! Twisting it back, the machine stopped working. Laying it flat, pressing on the power connector or UE10 caused it to work again.
I removed and cleaned the power connector and UE10. No improvement. Pressing on them causes the machine to work. I then lazily reheated the pins of many chips nearby UE10 (which turns out to have caused the next issue).
I measured the oscillator (Y2) and noticed it was only active when I pressed down. I replaced Y2 without improvement. I carefully measured everything going into Y2, and discovered power (5V) disconnected and reconnected when I pressed down, but power at the main motherboard connector stayed solid. Ah ha! A cracked trace or component leading to Y2.
By measuring the continuity each step of the way until Y2, I found inductor L13 was cracked at one end. I replaced it from a battery bombed SE/30.
2. Still Simasima display but no chime
After repairing L13 and seeing a valid clock, the computer still did not chime. Worse still, pressing down no longer caused the computer to start working.
I turned to the comprehensive Simasima repair guide:
I followed the guide and measured all the clocks and the reset lines. Everything was good.
Then, I learned something wonderful. I never knew the 68030 walked the address space, and in doing so, provides easy to follow frequencies to detect broken or shorted address lines. Thank you Simasima Repair Guide page 46.
Carefully measuring them revealed address lines A12 and A5 were combined. Nothing is going to work with bad address lines. Consulting the Bomarc schematic, I found that A12 and A5 are used on video chip UB8. That's one of the areas I reflowed lazily.
Turning the board upside down revealed a solder bridge. Solder flowed out the bottom of a hole and bridged to the next hole. Oops.
Removing the blob with desoldering braid resulted in a happy Mac chime!
3. Double vision. Multiple images.
What the what?
Multiple copies of the startup icon and mouse cursor appear on the screen. Lots of flicker. However, the horizontal and vertical size and placement of the screen is good. And, the images aren't stretched or condensed.
Fortunately, a post on the other forum by @KhanTyranitar and @Bolle provided a really strong clue.
SE/30 double image
Ok this is on an SE/30 that I’ve been trying to get working for a while. I took it to JJ Dasher and he was able to get it to this point. It does boot to the finder. But at every stage it has this weird double image thing going on. Has anyone encountered this before. Before anyone asks, yes...
68kmla.org
Pins 2 and 14 are shared for UA8, UB8, UC8, and UD8. On my motherboard, there was an imperceptible break in the trace for pin 2 of UB8.
Adding a small wire to connect the trace fixed the multiple image problem.
4. No SCSI
At this point, the computer had a perfect video image and happy startup chime. So, I was already pleased with the progress. The computer boots from a Floppy EMU but not from SCSI.
Well, this is obvious. UI12 (the SCSI chip) is near leaky capacitors C8, C9, and C10.
Learning from my earlier mistake, I decided to completely remove the SCSI chip with a hot air gun, to make sure everything was fully cleaned. Boy! I'm sure glad I did. It was gross.
After cleaning and reinstalling the SCSI chip, the SE/30 happily booted from the SCSI drive. It ran and passed all of the Mac Test Pro tests.
5. Long time to chime
There was still one oddity. It took at least five seconds after turning on power until the happy chime played and the screen was initialized. That is, the SE/30 sat for a while with a Simasima screen and then everything was fine. I assumed the ROM was waiting for the disconnected SCSI chip -- but that was now repaired.
Using the Simasima guide, I measured both reset pins (pin 5 on UB10 and pin 5 on UB11) on the Sony chips. They are distinctly separate. UB10 responded almost immediately after power up, but UB11 did not. Finally, UB11 changed state and that's when the Mac chimes.
During this diagnosis, I noticed pressing the reset button (S1) wasn't doing anything. Measuring the resistance of S1 indicated a steady 300ish ohms whether it was pressed or not. Clearly the reset button is damaged. As it was desoldered, a lead just fell off due to corrosion. I replaced it with a new old stock button.
6. Still long time to chime
Replacing the bad reset button did not fix the delay until hearing the startup chime. I check all the traces and nearby capacitors for continuity. Finally, I figured if the reset button was internally corroded, perhaps UB11 was as well. I desoldered it, cleaned and checked underneath, added a DIP socket, and swapped in UB11 from a battery-bombed board.
With that, this SE/30 board is fully functioning.
Perhaps you've noticed a lot of bad spots of corrosion on traces in the above photographs. Also, there are areas that still don't look clean. This is one of the unfortunate realities of some Simasima-afflicted boards. If there is enough corrosion to cause functional damage, then there is likely weaked traces and chips nearby. I chose to fix this board anyway, as I felt the practice would teach me many possible causes and symptoms. So, even if this motherboard is not one that I would rely on, it was worth the repair.
- David
