Failed TOPOWER TOP-500WB

[Pentium4]

I acquired this brand new a while ago, and recapped it because it used Asia X caps. I gave it to my buddy who needed it for a light gaming machine, which consisted of a Core 2 X6800 and a 9800GT, which should be a pretty light load for this PSU. While he was using it, he heard a loud pop, saw sparks, and no power. Wouldn't power back on, and no 5VSB. I pulled it apart and of course the first obvious part was the blown fuse. Upon further inspection, I saw a bunch of char on the PFC filer cap. I pulled the primary heat sink and saw it looked like the PFC diode had blown. It tests good though.... I tested the rest of the primary, and the bridge rectifier is also bad even though it looks fine. I think the resistor I pictured may be bad. I'm not sure what the value is supposed to be. Sometimes it reads 0.2Ω, but sometimes it fluctuates between 0.2-0.7Ω. What do you think happened to this thing? This thing looks very capable for the system it was in, I can't imagine anything in it being overloaded.

[kaboom]

Re: Failed TOPOWER TOP-500WB

Check the PFC input film cap. Incorrectly labeled CX6. It is not an X cap- we're on the DC side here.

It's right across the DC output from the rectifier. It decouples and commutates the high frequency AC on the PFC input.

If open, PFC boost mosfets explode, sometimes the bridge rectifier too.

Did you check D1 and D2 (HER-xxx)?

That resistor is 0.15 ohm, 5 percent, so your meter leads are affecting your reading.

You didn't mention it, does the booster inductor smell burnt?

Was that cap one you replaced?

[stj]

Re: Failed TOPOWER TOP-500WB

something arc'd aganist the heatsink and diode.
strangely the only thing oposite appears to be the mains cap.
is the sleeve o.k. ?

[budm]

Re: Failed TOPOWER TOP-500WB

Is the current limiter RT2 OK?
Can we see the straight shot of the top and bottom sides of the board?

[Pentium4]

Re: Failed TOPOWER TOP-500WB

Check the PFC input film cap. Incorrectly labeled CX6. It is not an X cap- we're on the DC side here.

It's right across the DC output from the rectifier. It decouples and commutates the high frequency AC on the PFC input.

If open, PFC boost mosfets explode, sometimes the bridge rectifier too.

It seems to test okay, it reads 1.00uF and 0.64Ω ESR but looks "bumpy" and "wavy" compared to that similar one. Kinda hard to tell by the pictures

Did you check D1 and D2 (HER-xxx)?

Both good

You didn't mention it, does the booster inductor smell burnt?

It still smells like it came right out of the factory.

That resistor is 0.15 ohm, 5 percent, so your meter leads are affecting your reading.

Yep, you're right. My leads were very dirty and in need of cleaning. Straight 0.2Ω now The heat-shrinked resistor reads 0.4Ω

Was that cap one you replaced?

That Teapo cap was the only one in the PSU that I didn't replace. It tested good.

something arc'd aganist the heatsink and diode.
strangely the only thing oposite appears to be the mains cap.
is the sleeve o.k. ?

Yeah the sleeve is okay. Seems bizarre. So I guess I should leave the PFC diode since it tests okay? The closest replacement I have is 10A

Is the current limiter RT2 OK?
Can we see the straight shot of the top and bottom sides of the board?

Yes, RT2 is good. MOV is good as well. Pics attached.

[budm]

Re: Failed TOPOWER TOP-500WB

So the PFC MOSFETs Q2, Q5 are not shorted? They are connected in parallel to drive PFC Coil L1.
R6 is fusible resistor connected to the (-) of the bridge and then connected to the load and (-) leg of the main filter cap.
How about D1, OK?

[Pentium4]

Re: Failed TOPOWER TOP-500WB

So the PFC MOSFETs Q2, Q5 are not shorted? They are connected in parallel to drive PFC Coil L1.
R6 is fusible resistor connected to the (-) of the bridge and then connected to the load and (-) leg of the main filter cap.
How about D1, OK?

All 4 FETs on the primary tested good. Even the PFC diode did, even though it looks charred. D1 tested good. So far, the only thing that I technically found bad is the bridge rectifier, and I'm not so sure about the integrity of the PFC diode.

[stj]

Re: Failed TOPOWER TOP-500WB

dont trust film capacitors, they can suffer from "punchthrough".
you may not see this on a meter because your not using enough power to make it leak.

[goodpsusearch]

Re: Failed TOPOWER TOP-500WB

Very interesting. I wish I could help. The film capacitor getting bad could explain the APFC failure in European version of Antec SmartPower power supplies.

[budm]

Re: Failed TOPOWER TOP-500WB

Since the fusible resistor did not blow open circuit and the MOSFETs are OK, if MOSFETs shorted out they would have blown that fusible resistor.
So the bridge just shorted out.

[Pentium4]

Re: Failed TOPOWER TOP-500WB

dont trust film capacitors, they can suffer from "punchthrough".
you may not see this on a meter because your not using enough power to make it leak.

So, are you convinced that the film cap is bad? Should I replace it with the other one anyways since it has tighter tolerance?

Since the fusible resistor did not blow open circuit and the MOSFETs are OK, if MOSFETs shorted out they would have blown that fusible resistor.
So the bridge just shorted out.

Oh well that's a good sign. Was this all just an anomoly?

[stj]

Re: Failed TOPOWER TOP-500WB

i dont nkow if your film cap is bad,
but that's the problem with them.
if the resin is cracked, or bulging or the resin is darker than normal then i'd bin it.

they used to be a major problem on thompson widescreen crt tv's for example.
it's not a rare problem if you use them for dc blocking with high current.
(inline with scancoils for example!)

[kaboom]

Re: Failed TOPOWER TOP-500WB

So, are you convinced that the film cap is bad? Should I replace it with the other one anyways since it has tighter tolerance?

This reminds me of a 12-14V fluorescent ballast I built. I found a cap to put across the primary of the transformer, supposedly up to the job. It only worked for a short time with that one. And when it quit, one of the MOSFETs ended up leaky.


I had another cap which lasted almost a year. When it died, I lost a MOSFET again.

I finally rebuilt it with slightly uprated MOSFETs and a Panasonic ECQ-HV (I think?) film cap. It's been flawless ever since.

When the old caps went, there would be some squealing/squegging when the lamp struck. Once this started, it was a week or less before the caps and MOSFETs died.

So replace that film cap to be sure. When they get "bumpy" in any power ckt, they're always suspect.

[kaboom]

Re: Failed TOPOWER TOP-500WB

Very interesting. I wish I could help. The film capacitor getting bad could explain the APFC failure in European version of Antec SmartPower power supplies.

Could be. Line voltage being higher would result in 300-340 VDC peaks across said film cap, plus the high frequency AC current through it!

[Pentium4]

Re: Failed TOPOWER TOP-500WB

i dont nkow if your film cap is bad,
but that's the problem with them.
if the resin is cracked, or bulging or the resin is darker than normal then i'd bin it.

Excellent, this actually helps me with another Topower PSU with no power that has a slightly cracked around both leads on the same brand and same value PFC film cap. Additionally, all those SMD resistors on the primary solder side tested good.

This reminds me of a 12-14V fluorescent ballast I built. I found a cap to put across the primary of the transformer, supposedly up to the job. It only worked for a short time with that one. And when it quit, one of the MOSFETs ended up leaky.


I had another cap which lasted almost a year. When it died, I lost a MOSFET again.

I finally rebuilt it with slightly uprated MOSFETs and a Panasonic ECQ-HV (I think?) film cap. It's been flawless ever since.

When the old caps went, there would be some squealing/squegging when the lamp struck. Once this started, it was a week or less before the caps and MOSFETs died.

So replace that film cap to be sure. When they get "bumpy" in any power ckt, they're always suspect.

Amazing! And, your first guess regarding this issue seems to be the solution, I will replace it tomorrow I guess the only questions now are whether to replace the PFC diode with a known good slightly lower value (10A) at least for the sake of testing? and the fuse. Honestly, does this thing really need a 10A fuse? I might have one around but 8A is probably the highest I have sitting around.

[budm]

Re: Failed TOPOWER TOP-500WB

I would use this one ECWFD. Non-inductive, high current pulse, high frequency.

[kaboom]

Re: Failed TOPOWER TOP-500WB

Excellent, this actually helps me with another Topower PSU with no power that has a slightly cracked around both leads on the same brand and same value PFC film cap. Additionally, all those SMD resistors on the primary solder side tested good.

Ok, great! And those resistors before the rectifier just discharge the film filter caps, so that if the unit is unplugged on an AC peak, the plug doesn't bite.

Amazing! And, your first guess regarding this issue seems to be the solution, I will replace it tomorrow I guess the only questions now are whether to replace the PFC diode with a known good slightly lower value (10A) at least for the sake of testing? and the fuse. Honestly, does this thing really need a 10A fuse? I might have one around but 8A is probably the highest I have sitting around.

As you know, you're actually dealing with two power supplies here- the PFC booster and the two-transistor forward converter. The booster won't be called upon to supply max current unless, of course, the main supply demands it. So try a 5A fuse for testing- 720 VA, and nearly 720W input margin once the PFC kicks in. Try some turn signal/taillight bulbs, on both +5 and +12. Or even a spare/junk HDD. You're only interested if it stays on with the PFC running.

Warning: Do not be tempted to remove that fusible resistor that returns the low side of the forward converter primary to the negative lead of the boost cap! If you do, you'll be floating the 325-380V return path...

If that fusistor were in the positive lead instead, and you could keep the aux supply powered (it runs the PFC), then you could disconnect the forward converter and watch the voltage on the boost cap as the PFC came up- you'd see it go up from its idle ~170V.

You cannot do that in this supply.

Yes, I know that fusistor is supposed to open if the forward converter draws too much, but it's far more likely they're watching the voltage across it to do "cycle-by-cycle" current limiting instead. We don't go around deliberately opening up return paths.

If you want, you can still put a meter across the boost cap and watch the action.

[kaboom]

Re: Failed TOPOWER TOP-500WB

I would use this one ECWFD. Non-inductive, high current pulse, high frequency.

Ok, looking at the "Permissible Current" in that datasheet, I think a 1.2 or 1.5u cap would offer more headroom. With the 1u, just over 3A seems too close to me.

For reference, I've attached Mouser's datasheet of the same series.

[momaka]

Re: Failed TOPOWER TOP-500WB

I will also suggest to check that Teapo cap on the primary again. You know, just for a good measure .

Also, before plugging it in next time, remember to use the incandescent light bulb series trick. Since the PSU has PFC, it may take a good 10-20 seconds before the bulb goes dim after you plug it in. But eventually it should. If not, you could still have a problem somewhere.

By the way, did I ever mention I hate PFC circuits?

[Pentium4]

Re: Failed TOPOWER TOP-500WB

You guys rock! It's back up and running I had forgotten that I had already instinctively ordered a few STTH12R06D diodes to have on hand, so I replaced that (and put the ferrite beads back on), replaced the film cap with the red one I had, had a GBU808 on hand, and actually found a 10A heatshrinked fuse. Only on the first start up, it took about 2 seconds to turn on after flipping the switch. Is this because the protection circuitry had been tripped previously? Another PSU saved by the knowledge of badcaps. Thanks guys, I couldn't have done it without your input, and now I've learned quite a few things.

By the way, did I ever mention I hate PFC circuits?

Yeah Most of the PSU's I have in service have no PFC. If the coil isn't in the way of airflow, I like having PPFC.