Antec 350 Recap Epic Fail

[bigbeark]

This 350-watt Antec was non-working. Three fuyuhu caps in the secondary were vented, so I recapped this PSU, removing every blasted fuyuhu cap and replacing with quality caps.

I had securely reassembled the PSU, but the top was off, and as I attemped to push the power cord in (switch was off) there was a loud pop and some sparks. The primary side heatsink was apparently live and touched the case.

Visual inspection confirmed this, slight heatsink melt and blackened area on inside of steel case.

So I'm guessing one of the components on the primary heatsink was shorting directly to ground. However, I repeat, the rocker switch was in the OFF position.

What happened to cause this? I went over all the caps I replaced and everything was done properly. There is no burning on any components that I can see.

I have now removed the wires so I could remove the heatsink, but the wiring was untouched when the short happened.

How do I test the Mosfets for short to ground? There is no continuity between any of the pins on the big MOSFETS and no continuity to the heatsink except at the mounting screws The smaller one has continuity between pins 2 & 3, but none to the heatsink except at the mounting screw.

[ben7]

Re: Antec 350 Recap Epic Fail

This 350-watt Antec was non-working. Three fuyuhu caps in the secondary were vented, so I recapped this PSU, removing every blasted fuyuhu cap and replacing with quality caps.

I had securely reassembled the PSU, but the top was off, and as I attemped to push the power cord in (switch was off) there was a loud pop and some sparks. The primary side heatsink was apparently live and touched the case.

Visual inspection confirmed this, slight heatsink melt and blackened area on inside of steel case.

So I'm guessing one of the components on the primary heatsink was shorting directly to ground. However, I repeat, the rocker switch was in the OFF position.

What happened to cause this? I went over all the caps I replaced and everything was done properly. There is no burning on any components that I can see.

I have now removed the wires so I could remove the heatsink, but the wiring was untouched when the short happened.

How do I test the Mosfets for short to ground? There is no continuity between any of the pins on the big MOSFETS and no continuity to the heatsink except at the mounting screws The smaller one has continuity between pins 2 & 3, but none to the heatsink except at the mounting screw.

Because the switch is off, that does NOT mean there is no voltage in the circuits. It likely was a N-GND short, or possibly they messed up big time and put the switch on the neutral, so the components are always live (on hot wire)

Is the fuse measuring good?

Check the bridge rectifier to see if it is shorted.

I would think the MOSFETs would be OK. (usually the heatsink is connected to the positive DC rail. but I may be wrong)

[goodpsusearch]

Re: Antec 350 Recap Epic Fail

Imagine your finger conducting the primary heatsink and the case... Yeah that happened to me... It was an Antec Smartpower 450 and the switch was at the off position but it was connected to neutral..

[bigbeark]

Re: Antec 350 Recap Epic Fail

Imagine your finger conducting the primary heatsink and the case... Yeah that happened to me... It was an Antec Smartpower 450 and the switch was at the off position but it was connected to neutral..

Just after it happened, I said to myself, lucky I didn't touch that heatsink...

That must be frightening and painful at the same time. OUCH!

[goodpsusearch]

Re: Antec 350 Recap Epic Fail

This saved me: http://www.conrad.com/Siemens-RCD-Fa...-int&pi=628032

[c_hegge]

Re: Antec 350 Recap Epic Fail

^
I've done a similar thing, to a similar PSU (an Antec SL350) and was saved by a similar device.

[bigbeark]

Re: Antec 350 Recap Epic Fail

Imagine your finger conducting the primary heatsink and the case... Yeah that happened to me... It was an Antec Smartpower 450 and the switch was at the off position but it was connected to neutral..

Did you find the shorted part?

I have some other PSUs where a screw is threaded from the case into the Primary heatsink, obviously this is intended as a safety feature.

If the heatsink can "go live" why is the heatsink not directly grounded to he case, which of course is grounded to the power cord?

[bigbeark]

Re: Antec 350 Recap Epic Fail

Because the switch is off, that does NOT mean there is no voltage in the circuits. It likely was a N-GND short, or possibly they messed up big time and put the switch on the neutral, so the components are always live (on hot wire)

Is the fuse measuring good?

Check the bridge rectifier to see if it is shorted.

I would think the MOSFETs would be OK. (usually the heatsink is connected to the positive DC rail. but I may be wrong)

This PSU uses an IC instead of a fuse. I used the Ohms setting on my multimeter, but did not detect a short to ground on the bridge rectifier.

I did see a slight deflection of the needle between the negative and positive legs but that
was not a repeatable reading. I assume the rectifier became charged and would discharge over time.

[ben7]

Re: Antec 350 Recap Epic Fail

This PSU uses an IC instead of a fuse. I used the Ohms setting on my multimeter, but did not detect a short to ground on the bridge rectifier.

I did see a slight deflection of the needle between the negative and positive legs but that
was not a repeatable reading. I assume the rectifier became charged and would discharge over time.

On the left side of the picture, there is a fuse, it is standing upright, in some plastic shrink-wrap tubing.

[delaware74b]

Re: Antec 350 Recap Epic Fail

I see carbon tracking around the small transformer in the lower left corner of your first photo, reattached and circled in yellow.

Also, removed -any- and -all- of that brown glue on that board. I had a power supply in an HP DC5100 small oven short out because of the glue.

Was there a plastic shield installed under the circuit board?

[bigbeark]

Re: Antec 350 Recap Epic Fail

Thanks everyone, really appreciate the help.

I repeated my testing of the bridge rectifier with a better ohm meter with better probes. I placed one probe on the solder spot where the heatsink was attached to the PCB and the other probe on each of the legs of the bridge rectifier in turn. This time every leg showed almost no resistance between the leg and the heatsink anchor point. Can I conclude that bridge rectifier KBU8J is internally shorted?

The fuse (beside the green disk) shows no continuity - it is blown. That rubber sleeve will be destroyed getting it off to see what the rating is. Thanks Ben7. I totally missed seeing it.

What appears to be carbon tracking is not. There is some sort of clear plastic lacquer which ends right above the duller black plastic of that coil.
I will indeed remove the big caps and clean that gunk off.

[retiredcaps]

Re: Antec 350 Recap Epic Fail

This time every leg showed almost no resistance between the leg and the heatsink anchor point. Can I conclude that bridge rectifier KBU8J is internally shorted?

Probably, but desolder to verify just in case something else in circuit is affecting the readings.

[Shocker]

Re: Antec 350 Recap Epic Fail

If the heatsink can "go live" why is the heatsink not directly grounded to he case, which of course is grounded to the power cord?

To prevent a shock if:

1. The ground is missing
2. The silicone pads break down

They figure that they'd prefer live heatsinks to a ground-fault.

[bigbeark]

Re: Antec 350 Recap Epic Fail

Probably, but desolder to verify just in case something else in circuit is affecting the readings.

So I removed the bridge rectifier and if the positive probe of the ohmmeter is conected to the positive leg of the rectifier and the negative to each leg in succession, there is continuity. if the negative probe is connected to the positive leg of the rectifier and the positive to each leg of the rectifier in succession there is no continuity. So the brdge rectifier only allows currrent to flow in one direction.

I checked another working PSU and connecting the negative probe to the heatsink and the positive to each leg of the rectifier shows continuity on every leg between the rectifier and the heatsink.

I must assume that I blew the fuse when the heatsink contacted the case, and that the heatsink is supposed to be live!

[ben7]

Re: Antec 350 Recap Epic Fail

So, you will need a new fuse, and possibly a new NTC (the green thing - check if it is open circuit)

[bigbeark]

Re: Antec 350 Recap Epic Fail

To prevent a shock if:

1. The ground is missing
2. The silicone pads break down

They figure that they'd prefer live heatsinks to a ground-fault.

Thanks for pointing this out. If someone plugs in a bad power cable and the heatsink is grounded to the PSU case, the whole computer is "live".
Yikes! So whoever takes the lid off a PSU should be **very** careful about plugging it in, or better, avoid doing so.

[bigbeark]

Re: Antec 350 Recap Epic Fail

So, you will need a new fuse, and possibly a new NTC (the green thing - check if it is open circuit)

Thanks Ben7! The NTC is good! I have learned a valuable lesson about working "under the hood":

Never plug the PSU in with the lid off!

[sam_sam_sam]

Re: Antec 350 Recap Epic Fail

I had securely reassembled the PSU, but the top was off, and as I attemped to push the power cord in (switch was off) there was a loud pop and some sparks. The primary side heatsink was apparently live and touched the case.
:

I am glad that you shared this I was working on a Dell Switching power supply and thought about not putting the top on it because of a fan problem

After thinking about this for a few minutes I though this might not be a good idea because it might go

I have another Dell Computer that I will be recapping all of the cap in the power supply and mother board

[sam_sam_sam]

Re: Antec 350 Recap Epic Fail

Thanks for pointing this out. If someone plugs in a bad power cable and the heatsink is grounded to the PSU case, the whole computer is "live".
Yikes! So whoever takes the lid off a PSU should be **very** careful about plugging it in, or better, avoid doing so.

I agree with this

[larrymoencurly]

Re: Antec 350 Recap Epic Fail

I did the exact same thing with an original version 300W or 350W Antec SmartPower, which also used high voltage MOSFETs, but after I unplugged the AC and plugged it back in (after installing the case top), the PSU worked normally. Neither fuse blew (this design had 2).

I haven't had too many MOSFETs fail on me, but every one that did had a short between the gate pin and the other pins, regardless of which direction the meter leads were connected.

I've read that the heatsink is connected to 170VDC in order to provide shielding from RF and to make it safer than connecting it directly to ground because sometimes there's no earth ground. The only grounded heatsinks I've seen were in the low voltage section of the PSU, namely in FSP (Fortron-Source Power) brand products.