The resistance value might be extremely stable, but it does generate a significant thermal voltage however.
Gee, the most expensive resistor you tested had poor results. So I take it actually testing things was more interesting than theory?
I am never afraid of a good real test.
In this case, my opinions were based on experience: I already had to design sub-microvolt processing circuits where, unlike audio, the smallest DC shifts are of huge importance, and although thermal emf's were a constant concern, normal metal-film resistors didn't pose any particular problem.
Now, I realize there are some notable exceptions, but at the time, they passed me by completely, thanks to luck and probabilities.
That said, I still do not think such voltages could have a real effect on the performances of an amplifier, even in an extreme case like this one, but being aware of such effects is never a bad thing, nor is the challenging of commonly accepted truths.
Actually it was in microphone preamplifiers I found the resistor thermoelectric problem! A DC servo was used and it being of no great importance was placed near the power supply regulator heatsink. The amplifier being a direct coupled output going into a transformer coupled input, the change in DC was a very big issue!
When an expert says something is possible, it usually means they know how to do it. When the expert says it is impossible they are often wrong!
- Status
- Not open for further replies.