S/3 Model 15 Project Denmark 2025
In 2011, 2014, 2022 and 2023
Jurgen and I made a start with the restauration of an IBM S/3 model 15D
at the Dansk
Datahistorisk Forening.
This is the blog of my fifth visit. Since my last vist Jurgen powered up the machine every few months to keep the capacitors alive.
After my last visit Jurgen was having problems with the 3277-1 console.
IBM 3277 Repair
All 3277 equipment in DDFH storage was moved to the 5415D location. Two 3278 and two 3277-2.
After inspection, one 3277-2 was already used for spare parts, the other one was dead.
The +34V fuse was blown. This is a 'standard' issue with the 3277's . An capacitor on the analog PCB decided to become a short circuit.
Replaced it, but still no live. The high voltage for the tube also was missing. Here we stopped with resuscitate this 3277-2 unit.
Let see if one of the two 3277-1 can be brought to live.
After replacing a fuse and powered up without an keyboard connected, some green light appeared on the screen.
We had one working 3277-1 which we needed for running the diagnostics.
IBM 5415 CPU start-up
Loading the diagnostics from the 3741 into the CPU ended every time in a processor check (PC).
After pressing "SYSTEM RESET" and "START" the message
+-------------------------+
! D C P L O A D E D !
+-------------------------+
was displayed.
The missing keyboard is not a real problem. All instructions to the diagnostics can be done with the hexadecimal dial switches on the operator panel.
IBM 3340 Disk drive restart
By accident the BUS cable between the CPU and the disk unit was pulled of at the 3340 side.
I replaced the broken connector in the 3340 and, together with Jurgen, all wires in the BUS cable where checked for continuity.
Loaded the diagnostics module C11 to test the internal memory of the 3340 attachment. This ended also in PC.
A further diagnostics of the cause of the processor check was needed.
We checked all PSU voltages, these where all within specs.
Check and reseated all flat cable connections. Still PC's.
Started replacing CPU cards with the spare ones we have.
After the 4th card swap the PC was gone and now we where able to run the 3340 diagnostics.
All diagnostics PASSED (!).
The big moment: IPL the system from an 3348 disk pack.
Mounted a 3348 data module,
put the drive online,
pressed "LOAD" and
fingers crossed...
After some CPU and disk activity a '--' halt was shown which indicates no IPL data available on that pack.
DDFH has several 3348 data modules. Of some modules the internel rubber mounting where broken, due to deterated rubber.
Some other modules made a terible noise due to dryed out grease in the bearings.
Finally, with the last data module (I am not making this up), we where lucky. More CPU and disk activity before the CPU went
into a wait loop. It was trying the load the print image address of the 1403 with a LIO instruction.
But the 1403 unit is not physically connected and the attachement goes into BUSY state preventing loading of the
image address.
We tried to bypass the LIO instructions, but now we where hit an other problem in the CPU: displaying content of memory locations
on the operator panel was malfunctioning.
Jurgen to propossed to bypass this problem by dumping the system pack to 8" diskettes, find the 1403 LIO's, patch them to NOP's and
restore the modified diskettes back to the system pack.
For this software trick Jurgen has to write a standalone dump/restore assembler program on the System/3 emulator and
transfer it to an 8" diskette and IPL it from the 3741.
IBM 3277 Repair
We still had a 3277 console without a connected keyboard. When connecting the keyboard the screen went immediately dark.
A warm smell was comming of the -12V power supply PCB.
Again short circuit, this time in the keyboard.
Replaced, just to be certain, both capacitors and the keyboard works perfectly again.
My fifth visit came to an end with:
- IBM 3277-1 working with keyboard
- Two running IBM 3340 drives
- CPU working (except displaying storage locations)
- IPL from a 3348 works ! (mile stone)
Not bad for 50 years old system.
This is the blog of my fifth visit. Since my last vist Jurgen powered up the machine every few months to keep the capacitors alive.
After my last visit Jurgen was having problems with the 3277-1 console.
IBM 3277 Repair
All 3277 equipment in DDFH storage was moved to the 5415D location. Two 3278 and two 3277-2.
After inspection, one 3277-2 was already used for spare parts, the other one was dead.
The +34V fuse was blown. This is a 'standard' issue with the 3277's . An capacitor on the analog PCB decided to become a short circuit.
Replaced it, but still no live. The high voltage for the tube also was missing. Here we stopped with resuscitate this 3277-2 unit.
Let see if one of the two 3277-1 can be brought to live.
After replacing a fuse and powered up without an keyboard connected, some green light appeared on the screen.
We had one working 3277-1 which we needed for running the diagnostics.
IBM 5415 CPU start-up
Loading the diagnostics from the 3741 into the CPU ended every time in a processor check (PC).
After pressing "SYSTEM RESET" and "START" the message
+-------------------------+
! D C P L O A D E D !
+-------------------------+
was displayed.
The missing keyboard is not a real problem. All instructions to the diagnostics can be done with the hexadecimal dial switches on the operator panel.
IBM 3340 Disk drive restart
By accident the BUS cable between the CPU and the disk unit was pulled of at the 3340 side.
I replaced the broken connector in the 3340 and, together with Jurgen, all wires in the BUS cable where checked for continuity.
Loaded the diagnostics module C11 to test the internal memory of the 3340 attachment. This ended also in PC.
A further diagnostics of the cause of the processor check was needed.
We checked all PSU voltages, these where all within specs.
Check and reseated all flat cable connections. Still PC's.
Started replacing CPU cards with the spare ones we have.
After the 4th card swap the PC was gone and now we where able to run the 3340 diagnostics.
All diagnostics PASSED (!).
The big moment: IPL the system from an 3348 disk pack.
Mounted a 3348 data module,
put the drive online,
pressed "LOAD" and
fingers crossed...
After some CPU and disk activity a '--' halt was shown which indicates no IPL data available on that pack.
DDFH has several 3348 data modules. Of some modules the internel rubber mounting where broken, due to deterated rubber.
Some other modules made a terible noise due to dryed out grease in the bearings.
Finally, with the last data module (I am not making this up), we where lucky. More CPU and disk activity before the CPU went
into a wait loop. It was trying the load the print image address of the 1403 with a LIO instruction.
But the 1403 unit is not physically connected and the attachement goes into BUSY state preventing loading of the
image address.
We tried to bypass the LIO instructions, but now we where hit an other problem in the CPU: displaying content of memory locations
on the operator panel was malfunctioning.
Jurgen to propossed to bypass this problem by dumping the system pack to 8" diskettes, find the 1403 LIO's, patch them to NOP's and
restore the modified diskettes back to the system pack.
For this software trick Jurgen has to write a standalone dump/restore assembler program on the System/3 emulator and
transfer it to an 8" diskette and IPL it from the 3741.
IBM 3277 Repair
We still had a 3277 console without a connected keyboard. When connecting the keyboard the screen went immediately dark.
A warm smell was comming of the -12V power supply PCB.
Again short circuit, this time in the keyboard.
Replaced, just to be certain, both capacitors and the keyboard works perfectly again.
My fifth visit came to an end with:
- IBM 3277-1 working with keyboard
- Two running IBM 3340 drives
- CPU working (except displaying storage locations)
- IPL from a 3348 works ! (mile stone)
Not bad for 50 years old system.