Replaced an old resistor yesterday that I suspected might be part of an oscillation issue I'm trying to track down (it wasn't). As part of it, I measured its inductance, and the inductance of a couple of replacement parts. I was quite surprised:
At 10Khz:
Old carbon comp, 200ohm, 0.5W: 200nH
Metal film, 220ohm, 1W: 220nH
Metal film, 220ohm, 0.6W: 670nH
Why this huge difference? The inductance (in this case), seems proportional to the physical size of the resistor. Is that a coincidence? Is there any general rule of thumb here, and are there any scenarios where it matters enough to be aware of?
At 10Khz:
Old carbon comp, 200ohm, 0.5W: 200nH
Metal film, 220ohm, 1W: 220nH
Metal film, 220ohm, 0.6W: 670nH
Why this huge difference? The inductance (in this case), seems proportional to the physical size of the resistor. Is that a coincidence? Is there any general rule of thumb here, and are there any scenarios where it matters enough to be aware of?
How did you make the measurement?
At 10kHz, 200nH is an impedance of 0.012ohm, absolutely tiny compared to 200ohm, meaning you need very robust and accurate measurement means to extract the pure reactive part.
The carb comp resistor should have ~0, the same as a piece of wire of the same length.
Film resistor are spiralled, and can have a few nH, tens of nH at most. 670nH sounds suspiciously high
At 10kHz, 200nH is an impedance of 0.012ohm, absolutely tiny compared to 200ohm, meaning you need very robust and accurate measurement means to extract the pure reactive part.
The carb comp resistor should have ~0, the same as a piece of wire of the same length.
Film resistor are spiralled, and can have a few nH, tens of nH at most. 670nH sounds suspiciously high
I used my portable LCR meter, a DER EE DE-5000.
Measuring resistance at the same frequency doesn't show much difference from DC resistance.
Try inserting a known, pure 200nH inductance in series with one of the resistors you tested: if it reads 400nH or 420nH, it's OK, otherwise you can dismiss the measurement
I would expect a value in the same area as 100R to be flat to 1 GHz or more.
Very low values < 1 Ohm go inductive
Very low values < 1 Ohm go inductive
Yes, I should definitely do that. A 100pf ceramic cap measured 26 H yesterday. That would have been practical.
I'm concerned because I have no idea, so I ask.
That's like being concerned with one grain of sand and how it might effect the whole beach.
Inductance depends _strongly_ on dimensions, wire length, wire thickness, and of course resistance-trace winding pattern.
In general the larger the component the more inductance, as current has to flow round a larger loop and loop-area is pretty dominant.
Also the thinner the wire the higher the inductance as this increases current-density.
Inductance is just the ratio of total magnetic flux of the circuit under test to the current through that circuit.