Short circuit when probing live ATX power supply primary section with voltmeter [closed]

Question

I was probing around an old ATX PSU to inspect voltage levels in the primary section. The first 15 minutes went smooth (bridge rectifier inputs, big caps, Power IC Vcc, etc..) but as soon as I probed the solder pads of the large coil circled in red in the picture, a short happened with the usual bang, sparks and smoke.

First question: what is that large coil (Hi-Pot) connected to the primary side? Is it part of the EMI reduction circuit?

Second (probably stupid) question: my DMM has 10Mohm input impedance when used as voltmeter. How can I cause a short like that if I am careful to ensure that meter probes do not touch one another and that I do not short any PCB tracks with the probe tip?

I am extremely careful when I work on energised circuits, so I am rather annoyed by what happened as I may make the same mistake again if I do not understand exactly what I did wrong.

A last question related to general safety: is there any difference between using a 1:1 (230VAC -> 230VAC) isolation transformer vs connecting the appliance to be probed directly into the mains using a lead without the earth wire (or with such earth pin isolated/disconnected)?

Thanks to the community for your precious help.

Answer 1

I'm not sure what the big inductor is for. It doesn't look like part of an EMI filter to me, since that would typically be a common mode choke with four leads, and this apparently has only two. (That's probably a common mode choke on the circuit board next to the the torroidal inductor, labeled FL3, with the white bobbin and the ferrite core.) It's also rather large for that. My guess (and it's only that) is that it may be part of a power factor correction circuit. BTW, "HI POT" is short for "high potential", and generally just means that it has passed an insulation breakdown test.

I don't have any really great theories about the short, except to say that, especially if you're running on 230VAC mains, the peak voltages in such circuits can be quite high, especially in circuits with big inductors. So it's possible that you didn't even make true metal-to-metal contact but simply decreased the air gap just enough to cause an arc.

Re the isolation transformer, YES, there is a difference! Please use the isolation transformer! Disconnecting the safety ground lead on the equipment under test doesn't remove the potential between the hot circuit and earth ground, so you are still exposed to danger. Even with a well-insulated, floating DMM, you can still have an accident. Sometimes engineers will float the ground on oscilloscopes so that they can probe line-side circuits, but this is still dangerous. An isolation transformer is the only reasonably safe way to work on these things.

Answer 2

Amanda, and welcome to the site.

Assuming your power supply looks something like this, your coil is the primary input inductor, shown at upper left in the schematic.

Now, what happened? Having been in similar situations a few times, I'd guess that you got distracted. You were holding both probes to the board, looked away at your meter, and BANG!. When you twisted your head your body followed, your grip on the probes shifted and you did, in fact, short out the probes. This put something like line voltage across your poor little coil, and it went to that Great Junkyard In The Sky. Very sad. Of course, that sort of excitement is (deep down) what draws a lot of us to messing around with electronics.

Please excuse FakeMoustache's response. He was just a bit appalled that you're putting yourself at risk by poking around in mains-connected circuitry without knowing much about what you're doing, and he's right - it's dangerous. You can take various precautions, but sooner or later you'll get something dramatic happening. Actually, you got off lightly. Very early in my career I lost a perfectly good pair of wire cutters when I tried to cut a live power cable. No shock involved, but the cutter edges were toast.

Answer 3

There is a reasonably-good chance that in fact, your meter's internal insulation failed and arced-over internally.

I honestly don't know what that off-board inductor is used for but my guess is that it's part of the Power Factor Correction circuit. If so, it's probably part of a boost supply that generates a nominal 400 Vdc. But that's only a Wild-Assed-Guess (WAG).

Regardless, many inexpensive DMMs are NOT constructed as they should be and do not have adequate clearance and creepage distances internally. They may work properly to their rated voltage but they simply can't withstand any sort of transient spikes.

Interestingly enough, old-school analog meters ARE properly constructed internally and will often withstand many times their rated voltage without suffering insulation failure. I'm talking about the old Simpson 260 or Avo meters from years gone by.

Part of that construction is because those meters were used to troubleshoot tube-type equipment with plate voltages of 250 to 450 Vdc or higher. It was not uncommon to see plate voltages well in excess of 1KV in transmitters.

Add to that the peak voltage swing that occurs with AM modulation - that 1KV plate supply swings up to almost 2KV peak.

Those old meters could (and did) withstand those voltage levels with no problems.

Modern DMMs made by reputable manufacturers such as Fluke are built to deal with those kinds of voltage levels. However, many of the inexpensive DMMs that come out of Asian factories are not. There are documented cases readily found with Internet searches of injuries caused by those inexpensive meters being used on 480 Vac power systems.

Answer 4

Please see Benchtop circuit breaker where I address some of the questions you raise concerning isolating transformer.

^{simulate this circuit – Schematic created using CircuitLab}

Figure 1. Benchtop mains power supply - modified for 230 / 115 V operation.

For working on switched-mode PSUs risk can be reduced by using a centre-tapped transformer to give 115 V where the PSU is so rated.

A last question related to general safety: is there any difference between using a 1:1 (230VAC -> 230VAC) isolation transformer vs connecting the appliance to be probed directly into the mains using a lead without the earth wire (or with such earth pin isolated/disconnected)?

The danger with disconnecting the earth wire is that the case and heatsinks may go live due to poor insulation somewhere. Since these present much larger areas and 3D shapes they may be touched accidentally. Another concern may be that the secondary is earth-referenced and that may go live-ish too.