Fortron FSP-ATX-400PNF

[Darko442]

Capacitor is solved,found rest of dead elco, 2.2uF 50 V(TEAPO),found it in PC case,what a explosion!!good idea for new year crackers...
Resistor is for now mysteries.

[momaka]

If I'm not mistaken, it looks like the burned resistor is connected to the Base of Q1. This PSU uses half-bridge design, so the circuits that drive the Bases of the two D209L transistors should be exactly the same. Start at the Base of Q2 and follow the circuit, noting the components connected to it. There should be a resistor connected in the same way as the burned resistor.
From what I was able to trace, I think the Base-drive resistor for Q2 is on the other side of the heat sink, circled in yellow in the attached picture. The Base-drive resistor for Q1 should have the same value.
Even if the resistor for the Base-drive of Q2 looks okay, check it with a multimeter to make sure it hasn't gone open-circuit or out of spec.

** EDIT **
I looked up some pictures on Google and from what I was able to gather, the Base-drive resistor of Q2 might be R12. Check to see if the resistor I circled matches the location of R12 - if yes, then that's resistor value you should use for the burned resistor.

[Darko442]

Yes you are right,and R9,R12 is open,and now is time to reconsidering cost of repair or just to throw in junkbox.
Thank to all people for help
Darko442

[Darko442]

Both transistors is short-circuit betwen base and collector,IC have very big chances to be dead,and how now what else.Yes PSU is for junkbox,in Croatia is very dificult to obtain adequate parts for repair(transistors and IC).One more time
thanks to all
Darko442

[trez0r]

Hi, all!

I had two of these PSUs to convert for usage with 200W Lipo charging device. The plan was to leave just 3Y and 3B wires coming out from the enclosure. I clipped and isolated the unneeded wires and grounded the green (PS_ON) wire. Upon assembly I found out that the psu will not start anymore - it will try to spin the fan for a split second and then shut off. If I manually apply G wire to ground it will emit a "tsk" sound from somewhere (best heard when it is done several times in succession like "tsk-tsk-tsk") - is that normal?
When working with second PSU and getting the same results I figured this is not a coincidence. What went wrong? There does not seem to be any visible damage anywhere. Is there a possibility to reanimate them?
Thank you!

[Scenic]

You might have a short somewhere. It tries to start up, detects a short and shuts down.
Other possibility is that your LiPo charging thingy uses so much current on startup that it appears as a short to the PSU.

Other than that, the startup cap(s) might be bad, but if both PSUs worked fine before you did the mods, that's kind of unlikely..

[trez0r]

On the first attempt I tried to solder out the unneeded wires so forming a short was a possibility. On the second attempt on second psu I did not even do that - I just clipped the wires and isolated them then I attached the green wire to one of the screws in the corner of PCB that has ground and tried to start the psu with the result already known.

The lipo charging thingy is not even attached yet. Though I used it previously with great success before I decided to r*pe off the bunch of wires that I don't need and that clutter up the space. I regret that already.

Btw which are the startup caps? The teapo ones in the middle? If caps are to suspect I could rather easily exchange them with new ones but at this point it would seem as a shot in the dark. I would rather come to some conclusion before attempting repairs. All of the caps seem fine and I tried to "diagnose" them (while one leg is soldered out) with dmm and saw nothing unusual (yes I know I should have used ESR meter but I don't have one).

[selldoor]

You should probably not be trying it with no load

http://www.pcguide.com/ref/power/sup...Loading-c.html

[trez0r]

I know that some older psu's had to have load applied to start up. Thing is that these psus were starting up (before I messed with them) by connecting the green to ground with no load whatsoever.

[selldoor]

Perhaps post some pictures top and bottom might help with spotting the problem.

[rogfanther]

The case screws are not always connected to the secondary GND. Have you tried connecting the green wire to some of the black wires from the psu ? Also, remove the connection between it and the case to do this test.

[trez0r]

Yes I've tried connecting green to black. Also that screw hole has a direct gnd connection beneath the pcb.

I'll try to put up some pictures in coming days, though there is not much to be seen, looks just like the pictures in OP except that there are 2 ost caps instead of 2 teapo on the side.

[momaka]

Check if there is a return line on the 5V or 3.3V rails. Pretty much any half-decent PSU will have it. It's usually a wire of the same color as the rail it is intended to monitor. If you still have the original motherboard connector for these PSUs, you'll easily be able to spot on which rail it is. Basically, one of the pins (again, usually 3.3V rail or 5V rail) will have 2 wires going to it (and one may be slightly skinnier than the other).

When you find on which rail it is, connect that skinny wire in the PSU to the rail it is supposed to monitor. So for example, let's say the return line is on the 3.3V rail. Then you connect a wire from the 3.3V rail to that skinny wire inside the PSU.

Originally Posted by rogfanther

The case screws are not always connected to the secondary GND.

Oh they better be. If they weren't, you *would* have noticed just by touching the case of the PSU.

[trez0r]

Oh my god, momaka!
You are the man, I completely overlooked this. It was a brown wire on 3.3V line I just connected it to 3.3V and psu started right up!
Gazillion thanks, buddy!

[momaka]

No problem, I'm glad I could help .

[trez0r]

Hey, I'm back with another troubleshooting question!
This time it's an ATX-400PNR psu from fortron (note the PNR part - a bit different model). This psu was junked by my employer as it was working unreliably and made a godawful noise when switched on.
Opened it up to find it built a lot less impressive than the PNF model of this topic (also the poorer v. lines spec support this). Has less caps in it (just 1 long 3300uf jamicon on 12V line) and also fewer voltage regs on heatsinks. It seems to be a newer iteration of the line as it even sports a 6pin pci-e power connector.
The teapo caps in the middle were bulged and I replaced them right away (I didn't have 6.3V caps on hand so I replaced them with two 10v 1000uf ones). That cured the startup problems as well as did away with the unpleasant buzz it made. All would be well except the line voltages are somewhat out of spec - 3.3 sits around 3.38, 5 - around 4.87 and 12 is way off at 13.2V. By shooting in the dark i replaced the long jamicon on 12V line with two 16V 2200uf but to my disappointment it didn't make any change. Could the volts be higher because of the voltage mismatch with replaced teapo caps? Unfortunately I didn't measure the voltages before I did the replacement.

[momaka]

Probably needs a load to regulate properly. Try hooking an (unimportant) hard drive and a few PC fans to it.

[pdavid]

I fixed a few these forton pnf units and after a proper cleanup and retouch I always added a 500Ohm potmeter in place of the 2 resistors (it's close to the brown 3.3 sense wire). Also there is another pot on the secondary daugtherboard for the 12V and 5V adjustments. The pnr series has these too. It's going to be easier later to tune all the voltages.

[trez0r]

Thanks for suggestions.
I just tried to put a 6 amp load onto 12V line and it dropped down to 12.2V, at 1 amp it fluctuated around 12.8V.
Also at 6 amps the psu became rather noisy (noise not coming from fan). I guess this one will have to go to parts bin..

[Th3_uN1Qu3]

Originally Posted by trez0r

I guess this one will have to go to parts bin..

Good idea. Fixable, but more trouble than it's worth, so just take it apart.