Found this power supply in a pile of spares.
Plugged it in, standby reads 5.06V. Makes a horrible whining sound just in standby, with no load.
Turned it on (PSON jumper), 12V reads a low 11.6V, 5V reads 5.1V, and 3.3V is only 2.85V. If I attach a disk drive to it, the 3.3V rises to about 3.1V, still too low, and the 12V rises to 11.8V (5V stays the same.) I doubt a computer would boot on this but I'm scared to try! When running the whining turns into a harsh buzzing and it sounds awful. (Whining changes pitch with load.) With even more load, the supply works, and the rails rise to the right voltage (12V @ 11.96V, 5V @ 4.99V, 3.3V @ 3.25V), but it really doesn't like load transients: for example, ejecting the disk drive on a CD rom causes the 12V to drop to 11.8V then slowly climb back to 12V.
Opened it up to find bulgers on the 5Vstby and 3.3V rails. Looks like an L&C/Deer.
Most of the AC filters, aside from two Y caps, are missing; and the board says it's a 200W supply (the label claims 400W.)
A rather creative pair of diodes are tied together and installed in place of the 5V diode. Yeah, like that's going to do 20A.
Output filter caps are 1000u + 470u for the 3.3V (yes, that's it), plus 1000u for the 12V, and 1000u for the 5V. The -12V has 470u. At least that's the right size. 5Vsb has a 1000u.
Input caps are 330u/200V each, which at 230V (which it is permanently wired into... no selector switch) means it's got about 165u of bulk capacitance which is far too little.
It's possible to make the controller IC (2003 chip) enter an undefined state if you power it on and off for about 2 seconds... The output voltages drop to about 10% normal voltage, but it continues switching. Seems like some kind of latch up from a transient.
Primary transistors are C5763, standby switcher is a CEF02N6.
Switch only interrupts live.
This thing could probably do 400W for about 1 millisecond before it pops.
My load tester (work in progress) shall have fun with this
.
Plugged it in, standby reads 5.06V. Makes a horrible whining sound just in standby, with no load.
Turned it on (PSON jumper), 12V reads a low 11.6V, 5V reads 5.1V, and 3.3V is only 2.85V. If I attach a disk drive to it, the 3.3V rises to about 3.1V, still too low, and the 12V rises to 11.8V (5V stays the same.) I doubt a computer would boot on this but I'm scared to try! When running the whining turns into a harsh buzzing and it sounds awful. (Whining changes pitch with load.) With even more load, the supply works, and the rails rise to the right voltage (12V @ 11.96V, 5V @ 4.99V, 3.3V @ 3.25V), but it really doesn't like load transients: for example, ejecting the disk drive on a CD rom causes the 12V to drop to 11.8V then slowly climb back to 12V.
Opened it up to find bulgers on the 5Vstby and 3.3V rails. Looks like an L&C/Deer.
Most of the AC filters, aside from two Y caps, are missing; and the board says it's a 200W supply (the label claims 400W.)A rather creative pair of diodes are tied together and installed in place of the 5V diode. Yeah, like that's going to do 20A.
Output filter caps are 1000u + 470u for the 3.3V (yes, that's it), plus 1000u for the 12V, and 1000u for the 5V. The -12V has 470u. At least that's the right size. 5Vsb has a 1000u.
Input caps are 330u/200V each, which at 230V (which it is permanently wired into... no selector switch) means it's got about 165u of bulk capacitance which is far too little.
It's possible to make the controller IC (2003 chip) enter an undefined state if you power it on and off for about 2 seconds... The output voltages drop to about 10% normal voltage, but it continues switching. Seems like some kind of latch up from a transient.
Primary transistors are C5763, standby switcher is a CEF02N6.
Switch only interrupts live.
This thing could probably do 400W for about 1 millisecond before it pops.
My load tester (work in progress) shall have fun with this
.
