CircuitBored
Well-known member
Earlier this month I made an impulse purchase of something I'd always wanted: a Macintosh SE/30. I've been trying to find a good deal on one of these machines for half a decade or more but they simply keep getting more and more expensive in working condition. I finally decided that I was going to have to buy a broken one and fix it so that very same week I won an auction for an SE/30 that was described as powering on but not chiming or doing anything. A few days later it showed up in the post, blissfully intact. After a thorough cleaning the case looked pretty great.
I pulled the logic board, fearing a horrible soup of capacitor and battery juice. Luckily this board gave me a rather good first impression!
Eight RAM sticks installed, no visible damage, and an intact battery dated 1989 that I hastily removed. The speaker cable wasn't attached to the board when I removed it, which I thought might have been the explanation for the lack of chime. Sadly when I reconnected it there was still no chime.
On closer inspection there was some evidence of capacitor damage. UE8 in particular was looking especially dodgy.
I went right ahead and started removing the caps from the board. I also removed UE8, which revealed rather a lot of mess.
Things were going smoothly until I cleaned the pads on C3. I was working with my soldering iron and some solder wick when I heard a click and saw a trace had literally jumped off the board. This is possibly the result of thermal stress on the board when removing the caps. I followed where the trace had gone, cut the ruined remains off, applied a bodge, and soldered on. I replaced all the caps with tantalums and took a chance on simply refitting UE8 rather than replacing it.
I tested the board and was delighted to hear a boot chime, indicating that the board was at least slightly alive. Sadly there was still no video. I still suspected the analogue board but also thought the CRT could be dead. I swapped the tube into my Classic and confirmed it was indeed working. After swapping the tube back into my SE/30 I noticed that the brightness adjuster on the analogue board was turned all the way down. I turned it back up and voila!
Still not great, but evidence that the analogue board was at least doing something. At this point I turned to the wisdom of the folks over on the 68KMLA IRC channel and started probing the video circuitry on the logic board under their instruction. We established that something wasn't quite right with the video signal so I replaced UE8. I also decided to replace all the other video muxes to be on the safe side.
I optimistically powered up the board but was still met with no video. At this point I checked the continuity on the legs of UE8 and found that one had a high resistance. I bodge-soldered the leg to the trace with a very tiny bit of wire and gave the board another try.
Slight success! At this point @Bolle pointed out that this looked like a stuck data line, likely meaning there was still a bad trace. I went back to UE8 yet again and found yet another bad pad. After applying another bodge...
Success!
I decided to start out with recalibrating the CRT as I needed to do the Classic's too. After a few hours of fiddling around I had both tubes looking rather fantastic.
The files on the SE/30's hard drive revealed that it was in active use in a small business until 1998. I was also able to find out how much RAM that machine actually had - 20MB. Not bad, not amazing.
At this point I tried to get the SE/30 networked with some of my other Macs, primarily so I could access my local AppleShare server and load some apps and extensions.
The SuperMac S900 was a great help here, as it has both gigabit ethernet and serial ports. Sadly the SE/30 would not connect to the LocalTalk network no matter what I did. This is where the Classic came into play, as it has serial I know to be good and a software configuration that definitely works with my network setup. Using the Classic I was able to establish that there was something wrong with the SE/30's serial ports. A further test with a direct SE/30 to Classic connection (as well as a hail-mary experiment with a serial-MIDI converter) confirmed that serial does not work at all. Note that I had to load all of the test software onto the SE/30 with floppy disks, a process that added hours to this endeavour.
After doing some probing and using the SE/30's schematics I figured out that something was going wrong with UB12 and UC12. The replacements for these chips are in the post so I will have to wait for the slow Christmas postal service to get them to me. I'll be sure to update this thread once the new chips are fitted!
I hope you enjoyed this summary of my experience with this charming little computer. I must thank my friends over on the 68KMLA IRC channel for all their help getting this thing working. I could probably have done this myself but it would have taken me weeks or months rather than a few days. I learned a lot about how the SE/30 works along the way and now have my mind set on finding another "dead" one to revive. This was pretty fun.
Thanks for reading!
I pulled the logic board, fearing a horrible soup of capacitor and battery juice. Luckily this board gave me a rather good first impression!
Eight RAM sticks installed, no visible damage, and an intact battery dated 1989 that I hastily removed. The speaker cable wasn't attached to the board when I removed it, which I thought might have been the explanation for the lack of chime. Sadly when I reconnected it there was still no chime.
On closer inspection there was some evidence of capacitor damage. UE8 in particular was looking especially dodgy.
I went right ahead and started removing the caps from the board. I also removed UE8, which revealed rather a lot of mess.
Things were going smoothly until I cleaned the pads on C3. I was working with my soldering iron and some solder wick when I heard a click and saw a trace had literally jumped off the board. This is possibly the result of thermal stress on the board when removing the caps. I followed where the trace had gone, cut the ruined remains off, applied a bodge, and soldered on. I replaced all the caps with tantalums and took a chance on simply refitting UE8 rather than replacing it.
I tested the board and was delighted to hear a boot chime, indicating that the board was at least slightly alive. Sadly there was still no video. I still suspected the analogue board but also thought the CRT could be dead. I swapped the tube into my Classic and confirmed it was indeed working. After swapping the tube back into my SE/30 I noticed that the brightness adjuster on the analogue board was turned all the way down. I turned it back up and voila!
Still not great, but evidence that the analogue board was at least doing something. At this point I turned to the wisdom of the folks over on the 68KMLA IRC channel and started probing the video circuitry on the logic board under their instruction. We established that something wasn't quite right with the video signal so I replaced UE8. I also decided to replace all the other video muxes to be on the safe side.
I optimistically powered up the board but was still met with no video. At this point I checked the continuity on the legs of UE8 and found that one had a high resistance. I bodge-soldered the leg to the trace with a very tiny bit of wire and gave the board another try.
Slight success! At this point @Bolle pointed out that this looked like a stuck data line, likely meaning there was still a bad trace. I went back to UE8 yet again and found yet another bad pad. After applying another bodge...
Success!
I decided to start out with recalibrating the CRT as I needed to do the Classic's too. After a few hours of fiddling around I had both tubes looking rather fantastic.
The files on the SE/30's hard drive revealed that it was in active use in a small business until 1998. I was also able to find out how much RAM that machine actually had - 20MB. Not bad, not amazing.
At this point I tried to get the SE/30 networked with some of my other Macs, primarily so I could access my local AppleShare server and load some apps and extensions.
The SuperMac S900 was a great help here, as it has both gigabit ethernet and serial ports. Sadly the SE/30 would not connect to the LocalTalk network no matter what I did. This is where the Classic came into play, as it has serial I know to be good and a software configuration that definitely works with my network setup. Using the Classic I was able to establish that there was something wrong with the SE/30's serial ports. A further test with a direct SE/30 to Classic connection (as well as a hail-mary experiment with a serial-MIDI converter) confirmed that serial does not work at all. Note that I had to load all of the test software onto the SE/30 with floppy disks, a process that added hours to this endeavour.
After doing some probing and using the SE/30's schematics I figured out that something was going wrong with UB12 and UC12. The replacements for these chips are in the post so I will have to wait for the slow Christmas postal service to get them to me. I'll be sure to update this thread once the new chips are fitted!
I hope you enjoyed this summary of my experience with this charming little computer. I must thank my friends over on the 68KMLA IRC channel for all their help getting this thing working. I could probably have done this myself but it would have taken me weeks or months rather than a few days. I learned a lot about how the SE/30 works along the way and now have my mind set on finding another "dead" one to revive. This was pretty fun.
Thanks for reading!