VOGONS


First post, by MMaximus

User metadata
Rank Oldbie
Rank
Oldbie

A few days ago I pulled my XT clone out of storage - it's a computer I bought new old stock a few years ago and it has seen only occasional use since then, the last time being around 6 months ago.

I turn it on and start noticing a burn smell moments after. Then I hear a small pop, and there's no more electricity in the house - it looks like the PSU tripped the circuit breaker 😕

Upon inspection it looks like a resistor has blown but I'm not sure if that's what it is:

hEaab82l.jpg

What do you think might have happened and how would I go about repairing it? Sorry if this sounds dumb - I know very little about electronics and soldering but I want to learn more about it 😀

Hard Disk Sounds

Reply 1 of 6, by Deksor

User metadata
Rank l33t
Rank
l33t

I don't think a resistor can burn on its own, it's more like a symptom rather than the cause.

Since it was never used for at least three decades, I suspect capacitors (especially tantalum) are dead. Tantalum caps often fail short. Check for continuity with a multimeter between GND and +5v, +12v, etc. If you get 0ohm resistance, one of the tantalum capacitor is probably dead.

Trying to identify old hardware ? Visit The retro web - Project's thread The Retro Web project - a stason.org/TH99 alternative

Reply 2 of 6, by keenmaster486

User metadata
Rank l33t
Rank
l33t

Agreed. Blown resistors are usually a sign of more serious problems; it means that something else failed, and sent way too much current through the poor resistor.

Also, your electrolytics are bulging. They should be replaced.

World's foremost 486 enjoyer.

Reply 3 of 6, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie

It looks like something spilled near that resistor? Inspect all the electrolytic caps ASAP.

That resistor is on the secondary side and should not cause a breaker to trip. What probably happened is a catastrophic cascade of failures:
- the resistor blew open
- quite likely it was a part of the voltage or current sensing
- that caused the PWM controller to max the cycle
- transformer saturated since there wasn't that much power draw
- collector current on the switching transistor went throught the roof
- that fried the transistor short, and so the recitifier shorted as well
- short on primary side tripped the breaker
- PSU is toast now, hopefully only that

Failure modes like these are _very_ common in switching PSUs. If it seems like the worst case scenario, it's not, the worst case scenario would be also overvoltage spike on secondary that damaged the PC parts. In any case, it can be repaired if you clean it up and replace the parts but it's not exactly a trivial task if you're not into electronic repair. I prefer to replace such PSUs unless it's so uncommon that there is no easy way to swap it.

Reply 4 of 6, by MMaximus

User metadata
Rank Oldbie
Rank
Oldbie

Thanks for all the useful replies. I have to admit didn't realize the problem could be that serious - I hope the fault didn't destroy any other component like the motherboard or the peripherals. 😳 I guess I need to try the parts with a different PSU to check they still work.

Even though I have many other AT PSUs, the problem with this one is its specific form-factor with the power switch on the side - I'm not sure how I could fit a different PSU into this desktop case, which is a shame as it's NOS. Ideally I'd like to rebuild this PSU - but I'd probably need to find someone who can do it for me!

Hard Disk Sounds

Reply 5 of 6, by retardware

User metadata
Rank Oldbie
Rank
Oldbie

For some reason most PSU boards are the same size, with mounting holes at quasistandard places.

Back then I didn't bother to repair the PSU boards, just transplanted the insides of another PSU, if case form factor did not match. This can often been done even without soldering.

Reply 6 of 6, by DAVE86

User metadata
Rank Newbie
Rank
Newbie
MMaximus wrote:

What do you think might have happened and how would I go about repairing it? Sorry if this sounds dumb - I know very little about electronics and soldering but I want to learn more about it 😀

In the circuit the blown resistor at R50 position creates a so called "snubber" with capacitor on C24 for the 12V rectifier diode that is on the small heatsink next to the blue regulation coil (D25?)
Same like R40 and C19 to D24 (5V rail rectifier).
The primary fuse looks intact. This might mean that the 12V diode got shorted and the sudden high current draw tripped your circuit breaker.
As long as the short on the secondray is present the psu will remain in fault mode and not start. This will prevent any deadly ic killing AC voltage to appear on the secondary.
The psu is okay-ish btw. Seems to be lacking a decent EMI filtering but otherwise the circuitry is good. lm339 + tl494 combo and that current sensing transformer on the primary (T4?) usually prevens motherborad killing events.