[rescue] SUN IPC woes

[Patrik Schindler]

Hello Hauke,

some thoughts from my electronics knowledge…


Am 04.08.2025 um 15:33 schrieb Hauke Fath via rescue <rescue at sunhelp.org>:

> The "exothermal" one had blown its fuse, and suffered burns on the transformer's primary side, unsoldering one pin and flat out losing a second one. I am kind of reluctant to hook it up to the grid.

Broken soldering spots are relatively common in PSUs. Mechanical vibrations from the transformer itself due to the magnetic field stress the soldering spots over time. Sometimes they crack. Then flashovers in the loose tin grain generate heat and lead to the situation you've seen.

If you manage to fix the soldering spots, the main issue is gone. It might happen that there are more defects due to the cracked spots, but that's rare.

Use a multimeter in ohms setting, and measure across the bridge rectifier + and - pins. You'll see or hear a beep signal indicating almost a short for some seconds until the primary caps charge from the multimeter batteries. If the meter reading doesn't change from the indicated short after 10 seconds, main causes are:
- switching transistor shorted
- bridge rectifier shorted
- other circuitry shorted

Desolder the components (or single pins thereof) and measure until the short is gone. Then you have the culprit.

If you can't measure a short, there's no way around: Plug in and test. :-)

All in all, it's always a good idea to resolder each and every grainy looking solder spot, especially when they're supporting large and heavy components, or those which become hot. You can tell this by the PCB being browner around/beneath those.

> The second one shows a short blip when switching on, then nothing - a frequent and mysterious error pattern, it looks. I might unsolder the transformer, since I practiced on the "exothermal" already.

A tiny working period ("blip") might indicate a short on the secondary side. Use a known good PSU to check the machine. If it's not the machine, there might be a short within the faulty PSU on a secondary. Use your multimeter in ohms setting as directed above, to measure the individual power rails (5V, 12V, etc.) for shorts, and try to isolate the fault.

If it's not a short on the secondary, there might be problems with other components. A good start would be the optical coupler. This is a device which provides electrical insulation between primary and secondary sides of the PSU, and feeds back voltage level information to the regulating circuit on the primary side of the transformer. Details are highly individual per PSU manufacturer and an are where I also lack experience and skill.

> Anyway - the restored PSU lead to more problems... Powering on the IPC gave me
> 
> PROM Checksum test
> 
> on the serial line, then silence. Status, or error? When I pulled the NVRAM, I got
> 
> PROM Checksum test
> Context Reg Test
> Setting Segment Map
> Sizing Memory
> Mapping ROM
> Setting up RAM for monitor.
> Setting up memory used in decompress.
> Decompressing code to RAM ...
> Remapping monitor's virtual addresses to RAM.
> 
> and letting it sit over night didn't give me any more.

Does not sound PSU related.

:wq! PoC