Seeing as the general consensus seems to be that there may be stupid people, but that they never have stupid questions, try this one:
Say I have a pair of oscillators with just 2nd & 3rd harmonics way down low in the noise floor and I want to check the IMD of an amplifier. Now the desirable thing is 19kHz and 20kHz signals to the amp, then check with spectrum analyser on the output for signals at 39kHz (sum) and at 1kHz (difference) right? Or not so simple?
I figure if the amp is good then there will not be much reason to check at 18kHz and at 21kHz. Or is that where one would look first? If one could...?
Reason I ask: I can only check (hopefully) low level gremlins if I use a notch to filter out the test signals after the amplifier. The analyser has limited dynamic range. I can't notch out the 19kHz and 20kHz test signals and still hope to see 18kHz and 21kHz because I don't know how to make such a steep notch. I can just notch out the both of them though, leaving below 10kHz and above 40-ish kHz (pretty much) intact. Am I correct in assuming that 18kHz and 21kHz distortion products can only result from 1kHz IMD product "re-modulating" with the test signals? So then 1kHz output will give me a reasonable clue of what's happening?
Of ben ik nu helemaal van de pot gerukt? Can amps generate "close by" IMD products without generating "far off" ones, with closely spaced test signals?
Say I have a pair of oscillators with just 2nd & 3rd harmonics way down low in the noise floor and I want to check the IMD of an amplifier. Now the desirable thing is 19kHz and 20kHz signals to the amp, then check with spectrum analyser on the output for signals at 39kHz (sum) and at 1kHz (difference) right? Or not so simple?
I figure if the amp is good then there will not be much reason to check at 18kHz and at 21kHz. Or is that where one would look first? If one could...?
Reason I ask: I can only check (hopefully) low level gremlins if I use a notch to filter out the test signals after the amplifier. The analyser has limited dynamic range. I can't notch out the 19kHz and 20kHz test signals and still hope to see 18kHz and 21kHz because I don't know how to make such a steep notch. I can just notch out the both of them though, leaving below 10kHz and above 40-ish kHz (pretty much) intact. Am I correct in assuming that 18kHz and 21kHz distortion products can only result from 1kHz IMD product "re-modulating" with the test signals? So then 1kHz output will give me a reasonable clue of what's happening?
Of ben ik nu helemaal van de pot gerukt? Can amps generate "close by" IMD products without generating "far off" ones, with closely spaced test signals?
