[weblacky]

Probably more affective that “killing it with kindness” would have been.

PSUs are a serious thing that several members are getting started on, however there are so many stations in need that we have to prioritize our efforts. I fear unless a dedicated collector of big iron SGIs makes the effort that those units will be reprioritized in favor of workstations first. I think eventually it will get there but if a collector can push these to the front for their own benefit, that would be best.

Also those PSUs are unique, in that someone had gotten partial schematics in the past. If those ever surfaced, it would obviously force those to the head of the line.

[Geoman]

Probably more affective that “killing it with kindness” would have been. 

PSUs are a serious thing that several members are getting started on, however there are so many stations in need that we have to prioritize our efforts. I fear unless a dedicated collector of big iron SGIs makes the effort that those units will be reprioritized in favor of workstations first.  I think eventually it will get there but if a collector can push these to the front for their own benefit, that would be best. 

Also those PSUs are unique, in that someone had gotten partial schematics in the past. If those ever surfaced, it would obviously force those to the head of the line.

Yes that's true. Well I wanted a Crimson for years now, never thought one would surface near my location, and even from whom you could consider the first owner. Then only now I have got the means to place such a big computer - that I consider to be already big iron (like deskside Onyx2 e.g.).

Concerning the PSU I have got an lets say academic in the family, who even could build me a new PSU with modern parts. The required voltages and so on it's all there. And I bet you even could rebuild lets say the Cherokee on hand: doulbe sided PCB with all traces clearly visible and all the ICs I saw on it are still buyable.

But I have the same thought: Big machines like Onyx and Origin, as tall as a man, they will be scrapped in the end. Who can storage such devices? Onyx and Onyx2 (Deskside) and Crimson etc. on the other hand - they have a fair chance of survival.

---

So I had some time to work on the Crimson today again and replaced all the caps on the boards. What I saw then was a leaked capacitor that ate away a bit of the - what I hope - silkscreen-traces and not real traces. These orange lines. Cleaned with vinegar and afterwards destilled water and IPA in the end.

[robespierre]

I can see that cap is visibly bulged, often the first sign of leakage before anything else is found.

The silkscreen is the letters and lines in white, which aren't affected by acidic liquid from capacitors. The orange lines are copper, which in this case, have several traces eaten.
If you have the means to do so, what needs to be done here is to desolder the DIP sockets and other components in the area, and cut back the damaged traces and lay down repair tracks.

Unlike with spillage from NiCad or alkaline-manganese batteries, acetic acid will not neutralize the leaked liquid. Electrolytic capacitors use an acidic solution for electrolyte, containing boric acid. To neutralize that, you can use a solution of sodium bicarbonate. Best would be ultrasonic cleaning using a proper electronic cleaner, with socketed chips removed, but this requires a very large tank for these size boards. The big danger if this can't be done, is that corrosives can move through the internal layers by wicking through holes.

[Geoman]

I can see that cap is visibly bulged, often the first sign of leakage before anything else is found.

The silkscreen is the letters and lines in white, which aren't affected by acidic liquid from capacitors. The orange lines are copper, which in this case, have several traces eaten.
If you have the means to do so, what needs to be done here is to desolder the DIP sockets and other components in the area, and cut back the damaged traces and lay down repair tracks.

Unlike with spillage from NiCad or alkaline-manganese batteries, acetic acid will not neutralize the leaked liquid. Electrolytic capacitors use an acidic solution for electrolyte, containing boric acid. To neutralize that, you can use a solution of sodium bicarbonate. Best would be ultrasonic cleaning using a proper electronic cleaner, with socketed chips removed, but this requires a very large tank for these size boards. The big danger if this can't be done, is that corrosives can move through the internal layers by wicking through holes.

Yes you're right- under the orange silk screen run traces - fortunately only 3 traces are impaired and there is still contiunity.

[Geoman]

I gave the PSU another chance by exchanging the optocoupler. That gets me 4 seconds of runtime until "UVL" on all rails.

[Raion]

I suggest breaking this off into a Hardware Triage/Repair thread, but I'm not sure what posts you all want me to take into such a thread...

[Geoman]

I suggest breaking this off into a Hardware Triage/Repair thread, but I'm not sure what posts you all want me to take into such a thread...

I guess moving the whole thread would't do any harm

[Raion]

Split it for you guys.

[Geoman]

Split it for you guys.

Thank you

[Geoman]

Cosmetics matter, too 

So the aged glue has been removed, the badge cleaned with IPA and reglued with 2 layers of fresh adhesive transparent foil. Not as new, because time took its toll, but stable for years to come. And a revision is always possible (non permanent glue).