Can someone please help me with repairing a PC power supply. My supply quit and I have been only partially successful. I found a burned out resistor on the board I replaced > a 5 watt 20 ohm and a bulging cap that I also replaced. When I try to power up now with the supply hooked up to a working pc the motherboard lights come on and I get a momentary turning of the fans and then it stops. The board lights still stay on and about every 15 to 30 seconds the fans again do a momentary spin and then stop. I tried to do some power test with a volt meter and did not have any power form the 12 volt rail and from the 5 volt rail I get about .02 volts. I did not test for the 3 volt rail though. I do not have a working knowledge of electronics but can do some minor test with a mulit meter and replace diods capacitors and resistors and such. Any help would be appreciated

Thanks Sparkie54

Dani AI

Generated

As reported by , the supply briefly tries to start (board standby lights present, fans twitch) then shuts down — a pattern that almost always means the SMPS is sensing a short or overcurrent on the secondary and opening. ’s “inverter” note most likely points at the switching/standby regulator rather than a separate CRT-style inverter, and is right that the secondary filter parts and fast rectifiers are prime suspects.

Safety note: this work involves lethal mains voltages and charged capacitors. Unplug the unit, wait, and discharge large caps before touching anything. If unsure, stop and replace the PSU.

Practical checks (in order, simple multimeter methods where possible):

  • Visual first: darkened PCB, cracked solder, burnt varnish on transformer windings or scorched components are strong clues.
  • With PSU unplugged, measure DC resistance from each +rail pin to ground (pins on the main ATX connector). A near-zero reading (<~1 ohm) indicates a hard short; a low single-digit ohm value is suspicious.
  • Check +5VSB (purple wire) with the PSU plugged in but not running — it should be ~5V. If absent, the standby regulator or its input fuse/diode is bad.
  • Bench-start carefully (short PS_ON green to any black ground with insulated tool) and measure rails briefly. Do not probe with hands inside.
  • Diode test the rectifier pack on the heatsink and check switching transistors for shorts (drain-source shorts on MOSFETs).
  • If caps are suspect and ESR meter is available, test ESR; otherwise replacing the secondary electrolytics with low-ESR, 105°C rated parts is a common, effective repair.
  • Some older ATX units need a minimum load on the 5V rail — a 10 ohm, 10W resistor across 5V to ground (~0.5A) is a safe test load.

If the short is in MOSFETs or the PWM controller, repair becomes advanced; replacing the PSU is usually the safer, cheaper option. Rebuild only if comfortable with high-voltage SMPS work and use correct low-ESR capacitors and proper soldering practices.

Recommended Answers

All 2 Replies

Have you checked the inverter?

If you have some sign of power up, it has hope!!

You still have a short which the regulator ic is sensing replace all filter capacitors on the 5v,12v and 3.3v line even if they don´t have signs of leaks or bulging, cause they might just have an ESR problem, Check fast rectifier diodes ( transistor like components mounted on secondary side heatsink ) are not shorted, open or leaking, also is the large toroid transformer looking kind of burnt or dark, enamel insulation on the wires might have failed due to heat and are shorting , swap it from another broken psu or if you feel brave disasemble taking notes of wire size and turns and rewind it again...

also you might want to consider regulator ic is defective ???

Hope this helps !!

Good luck

Abel

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.