Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[zmix]

.
Bruno,

I would like a bit of clarification... on Slide # 27 of this series it is stated:



By "low frequency loop gain" do you mean some specific low frequency corner or overall closed loop gain (below 20kHz)?

If the answer is the latter of those choices, then would it be accurate for me to interperet this to mean "Limit closed loop gain to the same value as the open loop gain at 20khz"?

Thanks,

[bruno putzeys]

If there's no way of making loop gain so large that distortion becomes entirely negligible, make the loop gain constant for frequencies <20kHz. It's a bit counterintuitive because there's a net increase in THD but the spectral composition seems more benign. Again, it's not scientific at all - I'm only saying that whenever I have had to put up with distortion, I had a subjective (=sonically speaking) preference for frequency-independent THD over somewhat lower but increasing THD. Saved the day a few times.

[KSTR]

Would this work with off-the-shelf IC op-amps also, by shaping a noise gain compensation so that noise gain is parallel to open-loop gain, in graphical terms?

- Klaus

[bruno putzeys]

A choke across the inputs of an op amp doesn't really sound attractive to me . Noise gain is called that for a reason. You'd end up with a 6dB/oct "red" noise source with tremendous DC offset and drift. Besides, if an IC isn't up to scratch it makes more sense to replace it by one that is. "Dirty tricks" like these are only worth trying if distortion levels are too high to be called negligible and there's no other option left.

[John Roberts {JR}]

Have you considered the phase shift caused by inadequate loop gain margin, since the open loop is typically 90' lagging?

This effect should be modest in most situations, but not zero.

JR