Dell D550p Not Starting

[stj]

depends on the current going through it in relation to the heatsink.

programable loads use a fet and can run for hours continuously.

[mmartell]

Since I can't find a shorted component I was wondering if the gate of the fet was holding the fet on and blowing it up at startup.

[mmartell]

Is there any way to test the PFC inductor without a ring tester ? It measures on my dmm the same as putting the probes together.

Is that what one would expect or should it be a couple of ohms at least ?

[stj]

i would think a couple of ohms with that wattage going through it would get very hot - besides, it's thick copper that probably has a much lower resistance than your test leads.

[mmartell]

Ok I'll rule that out for now and order some control ic's from China.

Speaking of that chip, I have the riser card out and fed 19v to pin 4 of that chip (DNA1005A) and it didn't draw any current. Was trying to determine if it was functioning but not certain it would run that way as I think it starts of the rectified mains voltage.

[mmartell]

After leaving this for a while I inadvertently plugged it straight to mains (actually to an isolation transformer, was going to scope the gate signal to the pfc fets...) and blew up a ground trace and popped the fuse. No other damage that I can see.

At first I thought I had everything hooked up incorrectly but there was no damage to the scope. But so I get this right... the DUT into an isolation transformer but the scope doesn't have to be (or the other way round is ok too I think ?) Ground for the scope on main filter cap negative lead, positive lead to gate of fet. Right ?

Is the blown ground trace telling me something important as to what the actual issue is ? It went directly to the ground pins of the DNA1005 PFC controller ic.

[Agent24]

It's better to float the DUT than the scope. In terms of not blowing your equipment, either one will work, but it's safer for YOU to only float the DUT, which is the main concern.

My guess is that your probe's extra capacitance made the FET gate stay charged longer than it should have, or caused unstable operation in some other way, ultimately making it conduct for too long and blow up, going shorted. Or exacerbated an existing fault with the circuit...

That probably sent mains voltage out of the gate of the now dead FET, and into the PFC IC, which also shorted, and the voltage took the easiest path: to ground, blowing the trace connected to those pins.

[mmartell]

The fet didn't blow and the pfc ic is somewhat protected from the fet as it actually drives two smaller transistors which switch the gate on and off.

But you are right, I forgot two check those transistors again. Please refer to the schematic on post 9.

How can I check the waveform output driving the gate ? Would removing the fets do the trick ?

[stj]

Originally Posted by Agent24

It's better to float the DUT than the scope.

i disagree.
i know why you said that,
but it could mean not noticing earth-faults / shorts to earth on the DUT.
(just keep your fingers off metal parts of the scope! )

[mmartell]

I've read there are differing opinions on this.

In any event, is it safe to plug the DUT into a dim bulb tester and THEN into the isolation transformer so I can scope the gate waveform before the unit resets itself ?