Discrete Opamp Open Design

[bcarso]

Originally Posted by scott wurcer

The 22K to ground was a nod to those that think 20KHz open-loop BW improves things, what I'm not sure. This should probably be left for listening experiments.

Those interested should also see Putzeys' article in Linear Audio for his take on having more gain at lower frequencies, and its effect on settling time.

It would be interesting to add overload behavior to the repertoire of tests in sim and in reality, including behavior under excessively high slew rate signals.

Virtually all conventional feedback amps (single-ended ones) will overload asymmetrically, the cure for which is making them faster, or bandlimiting ahead. Of course really poor slewing performance may require a lot of bandlimiting! But such performance is easy to avoid. According to Zanfino, Miller determined that Otala selected particularly poor designs to support his arguments about TIM. This does not invalidate the thesis, but it exaggerates a bit.

Once you make the bandwidth flat out to the highest frequencies of interest, does this guarantee equal distortion performance at all in-band signals? No, as there are other mechanisms besides those that the loop gain alleviates. One of the things I like about what linuxguru posts is the inclusion of the 1k resistor in series with the source, which is a reasonable start for what may be feeding the amplifiers. Those who omit any consideration of the source impedance are inviting sweet Spice lies.

[PMA]

Originally Posted by bcarso

Those interested should also see Putzeys' article in Linear Audio for his take on having more gain at lower frequencies, and its effect on settling time.

You know, it depends. This example has no GNFB, and THD below 0.0005% in audioband. I suppose no one would also predict TIM problems

[bcarso]

Originally Posted by PMA

You know, it depends. This example has no GNFB, and THD below 0.0005% in audioband. I suppose no one would also predict TIM problems

When I said conventional feedback amps, I meant conventional feedback amps with global feedback.

[Chris Hornbeck]

Originally Posted by bcarso

Those interested should also see Putzeys' article in Linear Audio for his take on having more gain at lower frequencies, and its effect on settling time.

Putzeys, in a later discussion in the John Curl thread here in DIYAudio, also made the point that the transfer function of the fedback signal differs above and below the open loop pole. Above has additional phase shift, of course, but this is a different transfer function.

Does this matter? And is it even possible to generalize? Dunno.

Thanks,
Chris

[PMA]

Settling time of GNFB design depends on frequency compensation method used.

[scott wurcer]

Originally Posted by Chris Hornbeck

Aren't those p-channel FET's?

Thanks,
Chris

Yes, sorry PFET's were the only thing compatible so it's one flavor of each device.

[bcarso]

Hope these don't get Wiley's panties in a bunch, as they say. If so I'll take them down.

The spine is a little the worse for wear, but I've gotten a lot of use out of this book.

[RNMarsh]

Originally Posted by PMA

Settling time of GNFB design depends on frequency compensation method used.

Also, depends -- if CMF circuit - no Cdom.

[scott wurcer]

Originally Posted by RNMarsh

Also, depends -- if CMF circuit - no Cdom.

Most of the circuits I've seen have ample junction capacitance to act as such.

Dick you would like the Wilson current mirror Wiki page. Barrie Gilbert and George had an overnight bet on who could make the best current mirror out of only three transistors, and George won.

I find it strange since given the first two there are not many ways to connect a third.

[jan.didden]

Originally Posted by bcarso

Those interested should also see Putzeys' article in Linear Audio for his take on having more gain at lower frequencies, and its effect on settling time..

Bruno Putzeys lives here.

jan