Are the heatsinks in a power supply supposed to carry a charge?

[Jeff7]

I'm doing a custom project here, and it involves a bare PSU PCB. Anyway, with a voltmeter set to measure AC, I found that there is a potential between the two heatsinks and ground, of about 110V - slightly lower than the house current. And it measures 1.6 amps. That was on a cheap PSU though.
So I got a known good PSU - TTGI brand, 420W, a pretty well made thing, built like a name-brand unit. Its heatsink->ground measurements are different - 0V for one, 140V for the other. Why do the heatsinks show as having such a large current flow through them? I'm guessing this might be one of the many reasons they always say you should never open a power supply - electrically charged hunks of metal.

 

[dmw16]

If you dont know exactly what is going on in there perhaps you shouldnt fiddle. That said, as long as the PSU is working, I would leave well enough alone. I also would never use a cheap PSU, and I would esspecially never mod a cheap psu.
-doug

 

[Jeff7]

Originally posted by: dmw16
If you dont know exactly what is going on in there perhaps you shouldnt fiddle. That said, as long as the PSU is working, I would leave well enough alone. I also would never use a cheap PSU, and I would esspecially never mod a cheap psu.
-doug

Well, it's for a non-critical project. That, and the structure it's going in can't support the weight of a really beefy power supply unfortunately, and there's just no room. This is for my scanner-PC project. Thus far, it works fine - cheap PSU and all. More assembly will come this week. I'm designing this as I'm building it basically. But it is in the final stages. I'm finally mounting components...and discovering that, like so many, the manufacturer of this floppy drive labeled the pins backward. Pin 1 is not pin 1, but pin 33 or 34.:roll:

 

[Zepper]

Most heat sinks are mounted to regulator tabs that are connected to ground, but there is no reason that the tab couldn't be on the hot side of the regulator or chopper either... Just be careful in there.
.bh.

:moon:

 

[OverVolt]

i don't think i would ever go poking around inside a PSU if i didn't know what i was doing.

 

[Cerb]

There's a charge in the PSU, and many IC packages have the flat area that the HS attaches to be the Vout.

 

[RussianSoldier]

i like taking the fans out of PSUs

 

[nineball9]

I don't know what components are mounted on the heatsinks in your power supplies, however the package case of large semiconductors is often a lead. For instance, in TO3 transistor cases (big old ones) the case is the collector, or the drain if the device is a FET. TO66 and TO220 (IIRC) are similar cases, but smaller. Some semiconductors have metal wings or tabs for heat dissipation which are often leads as well.

These components are sometimes mounted on their own heat sinks with a direct electrical connection to the metal heat sink. A common method is electrical isolation from the actual metal heat sink which may be chassis ground. In this instance, the device does not make an electrical connection with the metal, but the mechanical mounting allows for heat dissapation. For instance, a TO3 mounting package uses nylon or plastic washers to isolate the mounting screws and the case of the device is insulated by a piece of mica or similar material between the semiconductor's case and the metal heat sink.

I doubt your power supply uses TO3 bipolar transitors, but the concept is the same for voltage regulator IC's and other power semiconductors.

As for current - did you connect one lead to the heat sink and the other to the chassis ground with your meter on amps? If so, and if the heat sink is isolated, you created a short between the 2 circuits! This is not the current flow through them - you created your own circuit if you were using your meter to measure current. (You could measure voltages this way, but not current. The voltage readings would not mean much though - 2 different circuits.) Surprised you didn't fry something

Sometimes you may encounter a "virtual ground", but that's a different animal.

 

[VirtualLarry]

Originally posted by: Jeff7
I'm doing a custom project here, and it involves a bare PSU PCB. Anyway, with a voltmeter set to measure AC, I found that there is a potential between the two heatsinks and ground, of about 110V - slightly lower than the house current. And it measures 1.6 amps. That was on a cheap PSU though.
So I got a known good PSU - TTGI brand, 420W, a pretty well made thing, built like a name-brand unit. Its heatsink->ground measurements are different - 0V for one, 140V for the other. Why do the heatsinks show as having such a large current flow through them? I'm guessing this might be one of the many reasons they always say you should never open a power supply - electrically charged hunks of metal.

I've seen it mentioned before on here that sometimes, heatsinks in a PC PSU can be at "live" voltage levels, not ground. I've never measured it personally, but I could believe it to be true. I guess that's therefore just hearsay from me, though, FWIW.

 

[Jeff7]

Originally posted by: nineball9
As for current - did you connect one lead to the heat sink and the other to the chassis ground with your meter on amps? If so, and if the heat sink is isolated, you created a short between the 2 circuits! This is not the current flow through them - you created your own circuit if you were using your meter to measure current. (You could measure voltages this way, but not current. The voltage readings would not mean much though - 2 different circuits.) Surprised you didn't fry something

Well, if I fry anything, I'd know not to do that again. And, I've got two more of these power supplies.

Also, maybe it's worth noting - when I would be testing, I'd press just one lead of the voltmeter to the heatsink, and let the other one just held in the air, not touching anything. It'd still read about 30V. My VOM does this with other things, like one test lead in an outlet and the other lead not connected. It'd still register voltage flow.

 

[Tostada]

I remeber about 5 years ago I was replacing the fan in a PSU and decided to check if everything was working before I put the PSU housing back on. I found out the hard way that the heatsink was carrying AC. It actually burned my finger pretty badly and made me pretty twitchy for awhile.