Two Qs Regarding The Green Pre And Similar Designs

Hi

Two things I wondered about for a long time with these transformerless designs (for reference this schematic: gren_pre_cct4.gif):
* R2 and R5 are bootstrapped to provide very high common-mode input impedance (à la transformer); however R34/R35 shunt this impedance - so I do not see any advantage for doing so. A simple 68k would do.
* The first stage is essentially a current-feedback topoglogy. My textbooks, my reasoning and SPICE tell me that for best stability you should not apply lag compensation to the feedback resistor (C6/C7 and R6/R7) as we do with voltage feedback amps. Nonetheless I see them on some designs.

Any hints on where my arguments are wrong are appreciated!
Samuel

[quote author="Samuel Groner"]Hi

Two things I wondered about for a long time with these transformerless designs (for reference this schematic: gren_pre_cct4.gif):
* R2 and R5 are bootstrapped to provide very high common-mode input impedance (à la transformer); however R34/R35 shunt this impedance - so I do not see any advantage for doing so. A simple 68k would do.
* The first stage is essentially a current-feedback topoglogy. My textbooks, my reasoning and SPICE tell me that for best stability you should not apply lag compensation to the feedback resistor (C6/C7 and R6/R7) as we do with voltage feedback amps. Nonetheless I see them on some designs.

Any hints on where my arguments are wrong are appreciated!
Samuel[/quote]

In the presence of the phantom power R's the bootstrapping does seem a bit silly. I suppose if you switched phantom off downstream of C15 it would make more sense. If you want to really address common mode rejection it would work better to do it a la Whitlock's scheme, where you are a.c. driving both R's with a common-mode signal. This is now patented, though I don't know how broad the claims are. And where do you get the common-mode signal?

As far as the compensation---it's true that the structure is like a current-feedback amp. But maybe the singularity pattern of the op amp led to the feedback cap seeming to make sense. I suspect it was determined empirically. Spice may be lying about the 5532, or maybe there is a better solution.

Thanks for the answer.

Yeah, those opamp macro models are crap - I can make a 60 dB 10 MHz amp with my 5534 model by making the compensation cap small enough. :?

Samuel