> I am thinking the leakage from the backward bias diodes are the "resistor"
That's the way I read it.
> I don't understand 100% why there are 4 diodes in each group and not just 2??
The grid-cathode swing of the tube is many tenths of a volt, and a single diode would tend to go forward-bias.
Also to raise the resistance. Assume the grid current is 1 nanoAmp. A single silicon junction at 1 milliAmp has 30Ω dynamic resistance. At a million times lower current, it would be 30 MegOhms. This is a usable value, but 2*30Meg= 60Meg (similar to the popular 70M-78M resistors often found in tube condensers) only costs $0.02 more.
> you can get the capsule voltage from the cathode bias node of the CF by how "far" above ground the bottom of the bias resistor that node sits.
Yes, and you avoid a coupling cap.
But this 4060 capsule seems to be biased up near 100V-120V on a 120V supply. We can't do that and also run the tube. Lower capsule bias reduces sensitivity and soon gets you into noise. If the grid and capsule bias is 110V, we want more than 120V on the tube plate. One thing leads to another, major re-engineering.
> more parts than are needed I think I can get rid of maybe 8 or more parts.
I'm with you, except what is there makes no drawing-board sense. Yet its gross performance is not much lower than an "optimized" WCF drawing 2.5 times the current, or a paralleled cathode follower eating 4 times the current. And where it is "worse", it is in the direction of asymmetric (2nd-order) warps that won't sound as bad as an "optimum" WCF pushed to its limits. I really think someone spent many-many hours listening and tweaking.
> no surface mount parts...it's all hole-thru construction
And at $1,500 list, they could afford a 100Meg resistor (heck, 4*22Megs is only $0.48 installed). Yet AT generally knows what they are doing. So I'd guess that a zero-bias diode is a good high-Meg resistor for this purpose.
> 150dB is ridiculously loud....
It is not likely to be found in a studio. 126 to 134 is often all you need. But the capsules can often get into that 150+ range without distorting. The AKG C 414 reaches 140dB SPL straight, 150 and 160 with -10 and -20dB pads on the capsule. (The transformer version is 6dB less at 30Hz.) The limit is often the output amplifier: 160dB SPL at the 414's unpadded sensitivity would be 25V RMS! With -20dB pad, "just" 2.5V, which fits under the internal ~9V supply rail.
If the 4060 is really 2mV at 74dB SPL (rated 20mV at 94dB SPL), then 150dB SPL is 12V. They don't give a minimum load, so I would assume I could load it in 200 ohms. 12V in 200 is almost 1 Watt! And the mike claims to only eat 120V at 2mA, 0.24 watts. Even if we assume 2K load, that's 0.072 watts, and I know you can't hardly do that with triodes eating 0.24 watts.
