1300W PSU - Help with replacement mosfet

[sdinelli]

Hey guys,
had this 1300W gigabit PSU with a blown fuse ( i actually have two with the same problem.)
I found a shorted power mosfet and diode and some low Ohm resistors that measure 0Ohm.
I cant find the data sheet on this mosfet at all. need help finding replacement.
any help would be great.
I was able to find this on the mosfet but thats all
https://www.aliexpress.com/i/32967081176.html

[R_J]

Here is the datasheet

[sdinelli]

Thank you so much. hopefully replacing them with the other things brings these back to life. Got em for free and theyre fully modular so would be sweet.

[R_J]

Remember to check the gate drive circuit, when fets short they can damage the gate drive circuit and ic's

[sdinelli]

Thank you for saying that.
I found a 10k smd off the gate measuring 4k and also a smd transistor labeled "2F" testing unlike the 2F on the other FET.
Only one FET blew i compared it to the other side. getting 10k ohm on the other smd Transistor. and voltage drop of .64V and .6V.
On the faulty one im getting .615V one way and 1.4V
also found another smd transistor Labeled "1P" that doesnt match the one on the other side

Im a little confused though, theres also a 47ohm smd on the gate that tests fine though, how is the 10K off but the 47 alright?

[sam_sam_sam]

Do not forget to test with an incandescent light bulb instead of the fuse until you make sure that you do have any other issues

[R_J]

Originally Posted by sdinelli

Thank you for saying that.
I found a 10k smd off the gate measuring 4k and also a smd transistor labeled "2F" testing unlike the 2F on the other FET.
Only one FET blew i compared it to the other side. getting 10k ohm on the other smd Transistor. and voltage drop of .64V and .6V.
On the faulty one im getting .615V one way and 1.4V
also found another smd transistor Labeled "1P" that doesnt match the one on the other side

Im a little confused though, theres also a 47ohm smd on the gate that tests fine though, how is the 10K off but the 47 alright?

I suspect the 47Ω is in series between the gate and the driver, The 10kΩ is likely from the gate drive to ground and something else that is across it is causing the lower resistance. either bad transistor or drive ic output is damaged
the 2F could be a MMBT2907A
the 1P could be a MMBT2222L

[sdinelli]

ok ill have to look that up. not sure how to do that. solder some wires to the bulb and solder them where the fuse goes?
what does this do? limit the current?

[R_J]

It is for limiting the current, You can install the rated fuse in its holder and then you can build a test jig like this to plug the unit into instead.

[sdinelli]

Ive attached an image of pcb.
red are bad componants.(so far)
green is the good FET circuit

yellow is a questionable power transistor labeled GP4S60S. Cant fine a datasheet on that.
The reason why i question them is because there only 20ohms between the 1st and 3rd leads but also a 10K smd resistor between them as well.... Is it right to suspect them to be bad?
I didnt remove them because they're hard to get to.

[sdinelli]

Originally Posted by R_J

It is for limiting the current, You can install the rated fuse in its holder and then you can build a test jig like this to plug the unit into instead.

Thats awesome thanks!

[R_J]

Are you sure about that GP4S60S number?

[sdinelli]

Originally Posted by R_J

Are you sure about that GP4S60S number?

Im sorry its GP47S60X
https://hobbydocbox.com/Radio/757171...ransistor.html
Closest thing i could find

[R_J]

Is this it:

[sdinelli]

Originally Posted by R_J

Is this it:

Thanks again

[momaka]

Check the primary input electrolytic caps too. Usually blown APFC components is a result of the primary caps going either short-circuit or open-circuit / low capacitance. Not uncommon for these to go bad either. You will need an ESR + capacitance meter to test them, though (i.e. can't do it with just a regular multimeter.) The cheap GM328 component testers online will do this... or just spend the $8-12 on new input (and quality Japanese 105C too) caps, whichever you prefer. If buying new, make sure the caps are rated for at least 420V... or better yet 450V. APFC does not last long with 400V caps from what I've been seeing so far.

[sdinelli]

Originally Posted by momaka

Check the primary input electrolytic caps too. Usually blown APFC components is a result of the primary caps going either short-circuit or open-circuit / low capacitance. Not uncommon for these to go bad either. You will need an ESR + capacitance meter to test them, though (i.e. can't do it with just a regular multimeter.) The cheap GM328 component testers online will do this... or just spend the $8-12 on new input (and quality Japanese 105C too) caps, whichever you prefer. If buying new, make sure the caps are rated for at least 420V... or better yet 450V. APFC does not last long with 400V caps from what I've been seeing so far.

I do have both an ESR meter and a capacitance meter.
the two big caps ( i think these are what youre referring two)
are rated 470uf 400V CapXon
1st: 427uf .167 ESR
2nd: 422uf .177 ESR

both measure OL in resistence mode

[momaka]

Originally Posted by sdinelli

I do have both an ESR meter and a capacitance meter.
the two big caps ( i think these are what youre referring two)
are rated 470uf 400V CapXon
1st: 427uf .167 ESR
2nd: 422uf .177 ESR

both measure OL in resistence mode

That's good that both caps measure good. Otherwise, replacing these tends to add a fair amount to the cost of repairing these PSUs.

Since the caps aren't back, I went back and re-read the results you gave in post #5 and #10, along with the picture you posted with tested parts, to which I will be referring to on my comments below.

Originally Posted by sdinelli

I found a 10k smd off the gate measuring 4k and also a smd transistor labeled "2F" testing unlike the 2F on the other FET.

So referring to the abovementioned picture, I think that 4k resistance you're getting may be normal.
If you look at the left-most APFC TO-247 power MOSFETr spot (that you circled in red), notice there are holes/a spot on the PCB for another of the same MOSFET right directly behind it, but that MOSFET wasn't installed from the factory. Nonetheless, you can see that they did however install the small Gate pull-down resistor in spot R5b, along with the 47-Ohm Gate series resistor for that MOSFET. Therefore, when you measure the resistance across the 10k resistor in spot R5a, there could also be a parallel path through the other 47-Ohm resistor for the non-installed MOSFET and its 10k resistor (R5b). Thus, two parallel 10k resistors will read about 5 kOhms. The 4k resistance you got is lower... but maybe double-check that. Or maybe there are other shorted parts still.

I'm also not sure why they would install the Gate-driving transistors for the MOSFET that wasn't installed... but bad comes to worse, those should still technically be good... so maybe you can use them to replace the ones for the blown FET without having to buy new parts?
.
.
Next, look all the way to the right at the blown power resistor you circled in red - this one is likely the current sense resistor for the main switching PS. If this is bad, there's a good chance one or both power transistors next to it (circled in yellow) could also be bad. As such, you may have to remove them to check them (or at least one of them), along with their Gate driver circuits. On that note, I'm not able to determine yet if this PSU uses double-forward or resonant LLC topology, based on the pictures you provided. Thus, I can't really comment how the Gate drive components for those MOSFETs you circled in yellow should measure.

[sdinelli]

Originally Posted by momaka

That's good that both caps measure good. Otherwise, replacing these tends to add a fair amount to the cost of repairing these PSUs.

Since the caps aren't back, I went back and re-read the results you gave in post #5 and #10, along with the picture you posted with tested parts, to which I will be referring to on my comments below.


So referring to the abovementioned picture, I think that 4k resistance you're getting may be normal.
If you look at the left-most APFC TO-247 power MOSFETr spot (that you circled in red), notice there are holes/a spot on the PCB for another of the same MOSFET right directly behind it, but that MOSFET wasn't installed from the factory. Nonetheless, you can see that they did however install the small Gate pull-down resistor in spot R5b, along with the 47-Ohm Gate series resistor for that MOSFET. Therefore, when you measure the resistance across the 10k resistor in spot R5a, there could also be a parallel path through the other 47-Ohm resistor for the non-installed MOSFET and its 10k resistor (R5b). Thus, two parallel 10k resistors will read about 5 kOhms. The 4k resistance you got is lower... but maybe double-check that. Or maybe there are other shorted parts still.

I'm also not sure why they would install the Gate-driving transistors for the MOSFET that wasn't installed... but bad comes to worse, those should still technically be good... so maybe you can use them to replace the ones for the blown FET without having to buy new parts?
.
.
Next, look all the way to the right at the blown power resistor you circled in red - this one is likely the current sense resistor for the main switching PS. If this is bad, there's a good chance one or both power transistors next to it (circled in yellow) could also be bad. As such, you may have to remove them to check them (or at least one of them), along with their Gate driver circuits. On that note, I'm not able to determine yet if this PSU uses double-forward or resonant LLC topology, based on the pictures you provided. Thus, I can't really comment how the Gate drive components for those MOSFETs you circled in yellow should measure.

Thank you for your thoughts on this.
I may as well just remove the 10K resistor and just test it out of circuit.

yeah i was thinking the same thing about the transistors in yellow.
only thing is they will be a PITA to remove due to space restrictions of the long heatsinks.
any tricks on getting the screws out in tight spaces? i imagine a right angle Philips would work but dont have one .
ill report back when im able to tear into it some more

Those power fets are pretty pricey @ $8 a piece.
i did get these PSU's for free though

[momaka]

Originally Posted by sdinelli

only thing is they will be a PITA to remove due to space restrictions of the long heatsinks.
any tricks on getting the screws out in tight spaces? i imagine a right angle Philips would work but dont have one .

If you have the space (I doubt it, but anyways), will a philips add-on bit (like those short ones that come with boxes of new screws) fit in there?
If no, what about "extended" nose pliars - similar to needle-nose pliars, but a little ticker / more heavy-duty? I've used those to just turn the screws, despite the head being round. If the metal grip part on the pliers is not worn out, in some cases that will be enough to engage and grab the screw to turn it. I've generally had good success with this method.

Originally Posted by sdinelli

Those power fets are pretty pricey @ $8 a piece.

Ouch!
Well, you can always try smaller/cheaper MOSFETs (same voltage ratings, but lower rated current), just to see if you can get the PSU working again. After that, once up and running, substitute the proper parts. I've done this more times than I care to remember, especially with rectifiers and fuses. BJTs too, on a few occasions.

You can also do just multiple smaller FETs rather than a few big ones. Just connecting them would be more cumbersome, of course.