I've often puzzled about this. Had I learned electronics in school, I would probably have learned the answer. 🤔
With maximum signal voltage known, the numbers used to calculate the wattage for the resistor from an amp's output to inverting input, often called Rf, are obvious.
What about the numbers for the resistor from inverting in to ground?
With maximum signal voltage known, the numbers used to calculate the wattage for the resistor from an amp's output to inverting input, often called Rf, are obvious.
What about the numbers for the resistor from inverting in to ground?
The current is the same as in the other feedback resistor (assuming voltage feedback), so the dissipated power just scales with resistance:
P = V I = I R I = I2 R
It's more complicated when you are into exceedingly low distortion levels, as resistor self-heating and temperature dependence can affect distortion.
P = V I = I R I = I2 R
It's more complicated when you are into exceedingly low distortion levels, as resistor self-heating and temperature dependence can affect distortion.
That's what I have assumed, but the wattage calculated is typically so teeny, I doubted it.
Only for high gains. If the noninverting gain is two, the power dissipated is equal for the two resistors.
The voltage across that resistor is the same as the input voltage. Very small, thus small dissipation.
Jan
Jan
Yes, obviously. I should have qualified my response. The question arose because I've been mostly dealing with high-gain, high-voltage amps recently.
Aye... very long strings... gain X Rlength... but it moots the concern about locating that R close to the inverting input 🙂
From my experience, you'll get quite a minimal influence (almost negligible thermal or other distortion) when this resistor is about 8x-10x rated power than RMS power dissipated on it. That's for the metal-film type, of cause. When I say 'negligible' I mean a level at which the whole amp has about 0.001 % THD or less.
Last edited:
That makes sense. It is so low that I've never had to pay attention to it... until now.
Wishing you well in Kyiv.
IF you are fussed about distortion (slightly off topic) then I would be more concerned with the actual type of resistor rather than having multiple resistors of the same type.
If you manage to find carbon resistors and use them you will be disappointed, metal film (through hole types) to make a wild generalisation are good, I have read articles claiming SMD thick film resistors are less stable than thin film, and thin film SMD types perfrom better in terms of noise and distortion. There is a science to reistors and a bit of reading will likely be worthwhile if you are designing for that last decimal point in THD and Nosie.
If you manage to find carbon resistors and use them you will be disappointed, metal film (through hole types) to make a wild generalisation are good, I have read articles claiming SMD thick film resistors are less stable than thin film, and thin film SMD types perfrom better in terms of noise and distortion. There is a science to reistors and a bit of reading will likely be worthwhile if you are designing for that last decimal point in THD and Nosie.
Thank you.
Now I realised that I should add that my recomendation is more applicable to old fashion one MF resistors (middle or large sized). There are a lot miniature high-temperature types now and I don't have an experience with them in a precision circuits. They are possibly a bit worse then typical types.
But at the other hand, there are quite wide available 5-25 ppm MF precision types now - of cause, they are much better (10x better).
Long strings of resistors = antenna with each resistor distorting the noise non linearly (in theory). All depends on use i suppose.
Yeah, strings of equal-value R's has vey limited usefulness. Only relevant for low-gain amps, really.
You can always shield them if you are afraid of RF pick up.
Using strings of unit resistors has been a standard technique in analogue IC design for decades, but those resistors are much smaller and much more non-linear, especially in the bad old days when we still used doped silicon as a resistor. With discrete metal film resistors, it is only relevant when you are into extremely small distortion levels.
Using strings of unit resistors has been a standard technique in analogue IC design for decades, but those resistors are much smaller and much more non-linear, especially in the bad old days when we still used doped silicon as a resistor. With discrete metal film resistors, it is only relevant when you are into extremely small distortion levels.
You could also use resistor arrays like these.
Nice matched resistors for other uses too.
https://www.digikey.com/en/products...MIu56I5Jj0-gIVTceGCh3GFwtSEAQYBSABEgIKRfD_BwE
Nice matched resistors for other uses too.
https://www.digikey.com/en/products...MIu56I5Jj0-gIVTceGCh3GFwtSEAQYBSABEgIKRfD_BwE