Filters in a basic tube mic circuit

As I understand, a grid resistor and the capsule forms a high pass filter,  transformer coupling cap an other one (and a cathode resistor bypass cap a shelving filter if used).

In classic circuits the transformer coupling cap is the one that does the high pass job, right? But if you want to high pass why not doing it before the signal hits the tube? I remember reading about this in some thread here but can't find it anymore.

Second, I'm planning to put a switch in my mic for different cathode bypass cap values. Three different caps, one for flat response, one for low shelving (compensation for proximity effect) and one for high boost (very low value cap). Any drawbacks expected for this kind of switch?

Capsule C versus grid resistor isn't selected to block subsonics, it is selected so the cap bypasses thermal hiss in the grid resistor. This indirectly suggests the size and capacitance the capsule needs for adequate bass.

You still may need a cap to block DC from the transformer. And in general the C and transformer parameters are selected to work together. Since infinite inductance and capacitance are never in-stock, you tune the L and C and load R for a reasonably-damped resonance at the lowest frequency you can afford.

Make the mike as perfect as you can. If the track is boomy, fix in mix or ProTools.

Thanks again.

Still have some thoughts...

"Make the mike as perfect as you can. If the track is boomy, fix in mix or ProTools. "

I'm making my mic for the vocal purposes. That's why I'm still wondering: what is perfect?

The ELA M250/251 schematic found in web has a 30M grid to ground resistor, which makes about 50-100hz  (don't know the capacitance of a ck12) high pass filter with the capsule. It's supposed to be a great vocal mic, that's why one would/could consider lower grid resistor, even at the expense of a little higher noise? But ok, if the tube doesn't mind the sub bass you can filter it out later, I buy that.

Small (compared to modern ones) transformer coupling caps in some classic mics (U67, U47..) are not supposed to act as a filter? One would think a resonance at let say 30hz wouldn't be detriment in a vocal mic, when everything that low will be high passed anyway? Wouldn't the transformer and the pre be happier with less bass? I actually changed the 2,2uF cap for a 0,47uF one and found a improvement in sound. With bigger cap the sub sonic rumbble kind of "choked" the sound. I'll try a 1uF also. Maybe 2,2uF was too big for the load R and the trafo, don't quite grasp the theory behind "too big value cap in RLC circuit" yet?

And for the cathode cap. I find the partial bypass shelving curve to look very nice, and have a impression that it's also quite "phase friendly"? (and besides I've already ordered the switch and the additional capacitors and have room for them in the mic ) Some older engineers say that it's always better to fix thing as near the source as possible. I don't know if there's any point in that, but would explain why k67 and k87 capsules are better equalized in the mic rather than in a mixer/DAW.

Sorry if I'm being a pain in the ass.

PRR said:

Capsule C versus grid resistor isn't selected to block subsonics, it is selected so the cap bypasses thermal hiss in the grid resistor. This indirectly suggests the size and capacitance the capsule needs for adequate bass.

Click to expand...

There is no noise penalty in reducing the value of the grid resistor for a given capacitance of the capsule, since the resistor noise decreases. I don't know any example of such a scheme used to modify the LF response of a condenser mic, though.
I suspect it is probably due to some reluctance in bringing super-high Z connections to a switch.

abbey road d enfer said:

I don't know any example of such a scheme used to modify the LF response of a condenser mic, though.

Click to expand...

Any of the circuits on this page.

These are all the same design coming from one person. It would be interesting to know from the designer his analysis of pros and cons. And from users if a definite advantage appears.