Such devices are as much 'art' as 'science', and I recall that it's easier to get low distortion using a bulb rather than an FET in the feedback - from HP I seem to remember?.
The key question is what is your spec for THD? -60 dB? -80dB? -100 dB?? then how will you verify it? If -60 dB, the solution to both could be a Software Sig Gen. and Spectrum Analyzer. Audacity.exe generate anything you can imagine and detect with AUX input.
The key question is what is your spec for THD? -60 dB? -80dB? -100 dB?? then how will you verify it? If -60 dB, the solution to both could be a Software Sig Gen. and Spectrum Analyzer. Audacity.exe generate anything you can imagine and detect with AUX input.
Well that's not how you measure THD, it's in percentages (such as 0.01%) not dB's, and PC hardware is nowhere near capable of the requirements for decent quality audio measurements.
However, pretty cool thing Audacity - and would be VERY useful for less critical applications.
Well that's not how you measure THD, it's in percentages (such as 0.01%) not dB's, and PC hardware is nowhere near capable of the requirements for decent quality audio measurements.
Yes of course, but Spectrum Analyzers measure in dB then you convert that to %THD . Last time I used an electret mic with tower Turtle sound card, I was surprised by the SNR and THD it recorded using Audacity. The mic also had a DC response with bias nulled. Let's see how much %THD is -120 dB for the waveform in memory.....
Except a percentage is in reference to a known point, a dB isn't - and regardless, you don't give distortion figures in dB - just as you don't give the weight of jewlery in metric tonnes.
I've somehow avoided working with db to this point and recently confronted with a similar ratio question. This THD scenario frames it well.
If measuring THD, and there is a 10% positive and 10% negative deviation from the input signal, (i.e. output waveforms deviate from input from 90% to 110% of input), would THD be expressed as 0.9db (for the negative deviation) to 1.1db (positive deviation).
Except a percentage is in reference to a known point, a dB isn't - and regardless, you don't give distortion figures in dB - just as you don't give the weight of jewlery in metric tonnes.