Current in heatsinks

[Vivek]

Hi all,
I built an amp sometime ago and it has MJ2955/2N3055 at the output. The amp works normally (no hum etc.) but when working the output transistor heatsinks seem to be conducting some current. I can feel it when I touch the heatsinks. Is this a simple case of the collectors not being insulated from the heatsinks properly?
Thanks in advance for any help.

Vivek

[janneman]

Vivek,

Just for the sake of clarity: you feel that there is some VOLTAGE at the heatsink, I guess?
If you had BOTH transistors not well insulated, it wouldn't work. It can be that ONE transistor is not well insulated, and that is a disaster waiting to happen. It would mean that the heatsink is at either +V or -V, and if you connect it to gnd or another point on the amp, pooof!
So, to verify, measure the heatsing voltage with a multimeter against gnd. If there is a voltage, take care of the insulation of the power transistors before you blow up anything.

Jan Didden

PS It is possible that there is nothing wrong, that the heatsink is at zero volts, but that you touch a voltage point with your other hand. Same effect, what you feel is voltage difference, not absolute voltage.

[Vivek]

Hi,
This voltage I `feel' in the heatsink is there on both channel heatsinks. I have checked the insulation and the mica washers are all there. There is no voltage between the heatsink and ground.
Vivek

[janneman]

Vivek,

You can only feel voltage between two terminals. One apparently is the heatsink. What are you touching with the other hand? Are you maybe barefoot on the floor tiles? Arms on a chassis that is grounded or connected to mains neutral?

Jan Didden

[Vivek]

You are right. Voltage is across two points. But I feel this even if I touch it with one hand with the other hand not touching anything. I might be sitting with this stuff on the bed, on the sofa or standing barefooted on the mosaic-tiled floor but it is still there.

[Jennice]

Vivek,

It may be the case that you feel some potential relative to "absolute" ground (your mains supply ground entering the building you live in.)
Potential can couple capacitively and inductively, so the entire amp is not at ground potential. Maybe the entire device is at a potential. Try measuring the heat sink or enclosure potential relative to mains "earth", but do so with a voltmeter that has a high input impedance (digital multimeter or similar). The old fashioned voltmeters with a scale and pointer have much lower input impedance, and thus you will measure a lower voltage.

Jennice

[phase_accurate]

This voltage I `feel' in the heatsink is there on both channel heatsinks. I have checked the insulation and the mica washers are all there. There is no voltage between the heatsink and ground.

If you say ground, I assume you are talking about signal ground of the amp.
In this case there may be a voltage between the chassis and earth , that could be caused by an insulation failure. This can in fact be very dangerous !

If the chassis is earthed it could even becaused by the electrical installation of your home. So be very careful.

It could also be something less severe (like capacitive coupling in a transformer or mains filter), since such things are difiicult to judge remotely, but it always pays to be careful.

Regards

Charles

[Vivek]

Jennice,
I will measure heatsink to mains earth and let you know.

Charles,
I am talking about the signal ground of the amp. Actually, the heatsinks are not mounted on any chassis. I was just testing the amp and the heatsinks are kept separately. The PCBs are just kept in a metal chassis but on a piece of wood. Maybe I should replace all mica washers and check again.

[phase_accurate]

Jennice was a little quicker than me. Basically I wanted to say the same but maybe I was alittle less clear.

Regards

Charles

[wuffwaff]

Vivek,

are you shure you are feeling a voltage? Sometimes touching the sharp edged heatsinks can give the sensation of feeling a voltage, especially when you let you fingers run over the edges.

William