Original Zen redecorating project questions

[Andersonix]

Fellow Nelson Pass disciples,

I'm sure all these questions were covered back when this design was new in 1994, but please bear with me.

Background:
A friend got 2 hand made (in the worst way!) Zen amps on ebay years ago, and I have squatting rights to them. Each 'monoblock' amp was actually a stereo amp with inputs and outputs wired in parallel, so I have four amps supplied by two transformers. They use the original Zen circuit and PCB layout (cut in half), with the off-board addition of the 2mH inductor and 20000uF from Return of Zen (v2). They also use NTE caps: 2x 10000uF (PS) and 2x 2200uF (output caps) on each board. The amps always sounded fantastic, but were too ugly to display for polite company, in both components and workmanship. So, as a prelude to designing and making a really nice enclosure for an Aleph or some-such, I've gotten around to making a fairly nice new enclosure and substituting some not-awful components for these Zens. The upgrade budget is already blown on 16 new 12000uF Nichicons (2 for a dollar) and four film caps from Apex Jr; and connectors, Mills 0.27 Ohm 5W resistors, and BG 47uF input caps from Partsconnexion. I will wire it to make it possible to run as 4 separate channels or paralleled into 2 (or 3) with some jumpers.

Questions:
A. Should I replace the NTE output caps (2x 2200uF) with one of my new Nichicon 12000uF plus one 4.7uF film cap? My assuption is newer bigger Nichicon is better than old smaller NTE. Is NTE any good? (These have a 93 date code.)

B. Can I add the feedback used in v2 (Return of Zen 3/94) on the outside of the caps without changing P1 from 10k to 25kOhms or affecting the amp's input sensitivity too much? If I simply add R10 (1k) and R11 (3.3k), won't it affect the input impedance by the removal of R5 (4.75k), and if so, what will happen? (Sometimes the amp is driven by an op-amp based active crossover.) I don't understand how to interpret these excerpts from the v2 article:

[from Return of Zen (v2)]

"Second, it is easy to restructure the feedback loop of the amplifier so as
to enclose both the input and output capacitors in the loop, reducing
their distortions by the amount of feedback employed. This is shown in
Fig 1, where the values of R8 and P1 have been increased, and two new
resistors, R10 and R11 form a separate feedback loop outside the
capacitors.
R8 and P1 are still necessary for the purpose of setting the DC bias
point of the system, but by making their impedance values high compare
to R10 and R11, we allow R10 and R11 to dominate the performance.
Enclosing the DC input and output of the circuit with the R10 and R11
feedback loop also allows us to adjust the gain of the circuit
independently of the DC bias adjustment, which comes in handy later."
[jump]
"Third, you can adjust the values of R10 and R11 to reduce the gain of
the amplifier, increasing the feedback."

So the question is, what should I do about the extra feedback loop? It seems worth it and simple enough to add to the board if I knew what values to use.

Many thanks!
Tosh

[Magura]

A) Maybe, only you can tell which semantics you are leaning towards in the cap question. I personally used something like you're suggesting as a starting point.

B) No.

Magura

[Andersonix]

"No"? I want a second opinion.

[Magura]

How about the same opinion twice ?

Magura

[Andersonix]

How about a more specific question:
What value resistors should I use for R10 and R11 (while keeping P1 as 10k and R8 as 22k) if I want to enclose the caps in a feedback loop as in the Zen Revisited article from 3/94?

http://passdiy.com/legacy.htm