test for dc offset?

[mattmcl]

Can someone explain to me how to test for dc offset? I'm just finishing a set of Thor speakers and want to be sure I'm not running dc through them. I'm using an old Rotel amp with an mp3 player connected. Thanks.

[jackinnj]

Short the inputs of the amplifier and measure the DC component on the output with a multimeter.

[NicMac]

Originally Posted by jackinnj

Short the inputs of the amplifier and measure the DC component on the output with a multimeter.

Is it necessary to short the inputs?

[Greg Erskine]

Originally Posted by NicMac

Is it necessary to short the inputs?

Yes, to measure DC offset accurately. I my experience the DC offset will measure higher without shorting the inputs.

What if there is a earth lift resistor on one of the inputs? Should both inputs be connected to earth instead?

EDIT: Usually we measure DC Offset to test for an amp failure. This is usually indicated by the DC offset being approxiately equal to the rail voltage, so the absolute DC Offset value is not that important as long as it is the millivolts range.

[Speedskater]

Lots of amplifiers have a input coupling capacitor so shorting the input won't have any effect. But, it only takes a second to short the pins on a cheap interconnect with a screwdriver.

[fotios]

Originally Posted by Speedskater

Lots of amplifiers have a input coupling capacitor so shorting the input won't have any effect. But, it only takes a second to short the pins on a cheap interconnect with a screwdriver.

You are right in this. The input DC blocking (or input coupling if you mean the same thing) capacitor has this effect. It is better to short the input after the capacitor. My option it is to do this by soldering a little piece of wire across the out pin of capacitor and Gnd.

Fotios

[NicMac]

Same thing with balanced inputs? Both hot and return tied to signal gnd, preferably right after the coupling caps - if any.

[fotios]

Originally Posted by NicMac

Same thing with balanced inputs? Both hot and return tied to signal gnd, preferably right after the coupling caps - if any.

Yes. In the case of existing coupling caps (there are such cases) between input pins and inputs of subtracter (some times there are also voltage followers as buffers before the subtracter) you may do the same like in the single input. Short the out pins of caps to gnd.
The issue it is that, the coupling caps (yet of high quality like polypropylene foil) they present by alone a significant DC offset (up to 200 - 300mV) due to their coupling with the rest input filter parts of amplifier. Usually after the out leg of input cap it is soldered a series resistor (from 1 to 2,2KΩ) and in the out leg of this resistor it is soldered another one parallel (to Gnd) resistor equal to the feedback resistor. This network pull up the output of cap thus can not discharged due to its big value (usually from 2,2μF and above to not cut the low frequencies). If you short with a wire the out leg of cap to Gnd then is discharged and its DC offset is zeroed.

Fotios