FSP300-60PN, failed to start

[yanz]

I bought the unit in need of building some PC system.. (see my thread here..)

Well, that FSP unit didn't work when i tested it. I looked at the innard without opening the psu casing (it's easy because of the 120mm fan hole). And I saw two bulging teapos on the secondary side (one is a tall - grey colored and the other is short-light blue, i can't see what series they are).

Let me explain what "did not work" mean :

1. I plugged in the power cable to the PSU. I turn on the PSU switch and then sense the 5Vsb (purple) and the green wire -> my DMM showed +5V. So i connected my 25W (5V) and 10W (12V) dummy load. I jumperred the green and black wire on atx connector. Then, I hear a soft "click" from the psu, the leds on my dummy load blinking for a split second, but nothing more. The PSU is still dead. Checked the 5vsb, it was still 5v.

2. I redo the step 1, but now without any load at all (zero load). The 120mm fan on the PSU is turning on and the psu lives! Then I attached my 0.48A 120mm San Ace fans while the PSU running. It turned on too. Then i attached my dummy load mentioned in the step 1. Everything works.. only for a mere 8-10 seconds and the PSU died again!

3. I redo the step 2. Nothing happened, only a soft "click" sound from PSU. And as a desperate move i unplug the PSU from the main power, left it for 1 hour, and redo the step 1 and 2 for maybe a dozen times with different kind of load or no load at all. Nothing. It is still a dead PSU. I also tried loading the 3.3v but result is the same.

I hope i dont confuse you

So, my question is : do you think it is an easy fix or it has a deeper problem that can't be fixed by changing only the capacitors (including the small ost near pwm/5vsb regulator)?

I can change this PSU for a Dell 300W (Newton made, used) so i dont really want to void the warranty label if you suggest that it will be hard to fix. Maybe if you have the similar problem with FSP based unit, or have a thought of why it failed and how to fix it, or any other thought, i would like to know...

Thanks

[Per Hansson]

Re: FSP300-60PN, failed to start

Sounds a bit risky to chanche it
Usually once a PSU is up and running it does not care that it is outputting a few hundred milivolts of ripple on it's outputs
Getting it to actually start with badcaps can be hard tho...

[370forlife]

Re: FSP300-60PN, failed to start

I would just be using the dell anyways, newtons are built like a rock.

Try a recap, it might work again.

[yanz]

Re: FSP300-60PN, failed to start

Sounds a bit risky to chanche it
Usually once a PSU is up and running it does not care that it is outputting a few hundred milivolts of ripple on it's outputs
Getting it to actually start with badcaps can be hard tho

The PSU shutted down when i put 25W load on 5v, and never come back again. I feel that it's not just about those bad filtering caps. So yes it sounds a bit risky for me too.

I would just be using the dell anyways, newtons are built like a rock.

Try a recap, it might work again.

Well, i could imaginatively heard this PSU weakly crying for help.. Sending it back to the seller is just the same as sending it to the grave. (The seller will just trash it i believe).
So okay, i'll recap it. Might worth a try.

oh, where's my free time..

(And i hope that Dell has not yet been bought by someone.. )

[yanz]

Re: FSP300-60PN, failed to start

Here the pics :

[berreke54]

Re: FSP300-60PN, failed to start

here you can find the specs for fsp300-60atv(pf):

it says that the psu will shut down if the loads to +3.3v and +5v lines are not between certain limits
hope this helps
i'm working on that one

[berreke54]

Re: FSP300-60PN, failed to start

and here is the right one

[yanz]

Re: FSP300-60PN, failed to start

Thanks for the link, berreke54

Hmmm, so the 3.3v line needs at least 0.3A load. Apparently my 30 ohm resistor wasn't enough . Will try a lower ohm resistor later..

[yanz]

Re: FSP300-60PN, failed to start

3.9 ohm resistor is connected to 3.3v line. (about 0.8A, 2.8W load)

Others line :
5v -> 31W
12V -> 10W

Result : it works! Haha, thx berreke..

I learn something new. This is the first unit i ever faced that should be loaded on 3.3v rail with a certain minimum load.

now back to recapping work...

[yanz]

Re: FSP300-60PN, failed to start

wow the resistor got really really hot...

[yanz]

Re: FSP300-60PN, failed to start

looks like i got three bad teapo caps there, no less...

[Leopard]

Re: FSP300-60PN, failed to start

I would replace all Teapos on the output side of the PSU. Teapos usually last in PSUs quite well but in this FSP they seem to fail and bulge. Maybe they are underrated for their job and cannot handle the ripple and/or heat?

[yanz]

Re: FSP300-60PN, failed to start

i have completted the recapping.. huhuhu, what take it so long? will upload the pics tonight..

@Leopard, apparently, what killed them is heat. lot of heat. and dummy resistor(s) is one of the culprit. even good caps will fail with such this stupid design of fortron, wich make me wonder how couldn't those smart heads at fsp predict this failure..

and to think that my new silverstone 500w (fsp built) inherit the problem.. oh.

later..

[yanz]

Re: FSP300-60PN, failed to start

my sister was hospitalized so it was a busy days for me and my nephew.. so, sorry for late update.. n this is what i promised back then :

[yanz]

Re: FSP300-60PN, failed to start

more.. (after recap)

[yanz]

Re: FSP300-60PN, failed to start

this is the secondary area where some caps been removed. also the 3.3v coil and some dummy resistors been removed :



those resistors are literally touching the caps. burnt marks is everywhere.

------

and this is the 5v and 12v area :



i have to unsolder the modul board (for VRM, i think). that big resistor is a 9.5 ohm, equal to stupidly 2.63W load. it's too high, too hot, to big, and too tight for it to even exist in such area. i was clapping my hand cynically for the enginer who designed this.

------

what was burnt :



^ That poor teapo caps got pregnant. And that tall black thing is a 5V coil, it's transferred the heat to everywhere.




^ ceramic cap was burnt and cracked easily while being removed. small electrolit caps were also damaged (such small caps can be pregnant too, if u never seen one).

[yanz]

Re: FSP300-60PN, failed to start

the look before recapping :

[yanz]

Re: FSP300-60PN, failed to start

i did quick math : the only resistor that has below 1W load was only -5V dummy resistor. if you think such 1W - 2.6W are just small wattage, well that amount load can heat the resistor to over 60C. (i've tried touching them with bare fingers).




the heat cooked the component near side and the pcb :

[yanz]

Re: FSP300-60PN, failed to start

After recapping :





caps for 3.3v and 5v : 1500uF 10v Sanyo WX, 1500uF 10v Rubycon MBZ

caps for 12 v : 1800uF 12v chemicon KZJ and 3300uF 16V Nichicon HE

all the small caps (470uF to 10uF) were replaced with Rubycon YXF, chemicon KZE, KMF, and the like (don't remember the details).


all the dummy resistors were dumped. never again i'll see them inside this PSU!

if such resistor have to be exist, i propose to use a flying resistor mode

[yanz]

Re: FSP300-60PN, failed to start

note that before recapping, this psu only used a single caps for 12v rails. after recapping, the KZJ took over the place of that big red hot resistor, thus there are two caps for 12v rail now.

also, i have use the "flying resistor" technique in the past for several psu that i have to be sure it will live on with no load at all (for a certain purpose).

pic is stolen from xbitlabs :