YES you need an input transformer. How else could you get CV into the grids without being shorted-out by audio parts?
Gyraf 1176 is a VERY different compressor. The FET does not have audio and CV on the same pin. The FET is linear-enough without being worked push-pull.
And the transformer input will take balanced OR unbalanced drive equally well. (Some of your questions about transformers and pots seem very unclear.)
> my card was not putting out a hot enough signal
{sigh} I did ASSume you would use semi-Pro gear able to pump 7V signal, or at least a good bit over 1V. If you use some lame PC port, without any decent external gear, you may as well do your compression in the PC. (That's the only way I do it now.)
And when asking if a limiter is working, you monitor the sidechain. In this case you wudda got "no" CV, or just a slight blip on peaks. Clue!
> i have some 6bz7's
If you had DECENT Pro levels, the 6BZ7s would reduce S/N and would fuzz-up too easy.
If you are going to use lame drive, yes, 6BZ7 will have more gain than 12AU7. Do it. (Check pin-out.)
> why only 100 b+? Doesnt the lower B+ cause a tube to distort easier?
Distortion is relative to signal level. There is a gain of 10 after the tube. The maximum final output is, what, 2V? So the tube only has to make 0.2V output. A 100V supply in a normal stage would support 20V output. It appears we are at 1% of the distortion point. What distortion?
It is not that simple, because for high output we force the tube to LOW current. That's the flaw in this type compressor: when we need the most, the tube is at its weakest. It's a fine design point. However for modest limiting, this thing don't suck.
Why 100V? Remember that to get gain reduction we reduce tube current. And to get large reduction we need to force the tube very nearly to "off". How much G1 voltage does it take to turn-off a tube? Vpk/Mu. How much G1 voltage can we make? With the chip sidechain, maybe 10V, though with the simple plan, 5V is a safer bet. What is 12AU7 Mu? About 20. So the max Vpk we can run is 5V*Mu or 100V.
The other reason: "LOW cost". It uses a 120VAC winding, junkbox parts. This rectifies to about 160V. However the Threshold is affected by tube voltage (ultimately Vpk/Mu). The output limit "must" be Studio Standard, whether wall-voltage is low or high. Maybe I'm fussy, but I like a solid limiting level. So I regulated the tube supply. No snazzy LDO regulators: junkbox Zener. To get passable efficiency and stability, the Zener voltage has to be much less than the nominal raw voltage. 160V raw to 100V Zener seemed good to me.