polar patterns v.s. voltage

can anyone point me in the right direction for some reading on this?

Some VERY informal experiments seem to suggest to me that higher voltage might = a wider pattern on a k67 type capsule... 

any info is appreciated.

Thanks.

Variable-directivity condenser microphones operate with dual capsules. These are constituted of a two cardioid capsules back-to-back, which are connected in parallels AC-wise, so their output voltages are summed. By varying the bias voltage of one of the capsules respective to the other, the summation gives differnt results.
If the bias of the rear capsule is null, its output voltage is null, and then the combined voltage is that of the front capsule, i.e. cardioid.

When both capsules receive the same bias voltage, their voltages add linearly; the output is the same when the sound comes from any direction (in fact, depending on the physical characteristics of the capsule, there is some narrowing of directivity at HF, but the front and rear sensitivity are equal).
If the rear capsule bias is of opposite polarity, the front and rear sensitivity is equal but out-of-phase, and lateral sensitivity is null, giving the typical figure-8 pattern.

Several mics offer intermediate positions like very wide cardio, wide cardio, not-so-wide cardio, hypocardio, hypercardio, supercardio, megacardio, figure-of-peanut...

For practical reasons (unavailability of negative voltage), very often the backplate is at a positive voltage and the diaphragm of the front capsule at zero; consequently, the voltage on the rear diaphragm is varied between zero and twice the nominal bias. In cardio mode, the rear diaphragm must be sitting at nominal bias, in omni it must be zero, and in figure-8 it must be elevated at 2x nominal.

In a U67, the backplate sits at 60V, the front diaphragm is at zero, and the rear diaphragm is at the same votage (in fact directly in parallels) in omni mode, completely disconnected in cardio, or bias at 120V in figure-8.

Is it a dual diaphragm capsule?

If so it depends on the polarisation of all three electrodes - rear, backplate(s), front.

Think about each side as a variable size cardioid shape. The higher the voltage the bigger the volume / more sensitive the mic.

These can be combined either in phase or out of phase, depending on the polarisation voltage relative to the backplate

So, for something like the G7, in omni, you have 0V, 80V, 0V. Both diaphragms are -80V relative to the backplate, and so combine equally in phase to give the omni capsule.

In fig 8 you have  0V, 80V, 160V. Relative to the backplate that's -80V, 0V, 80V. So they combine equally anti-phase, and at the sides where they overlap they cancel.

All the other patterns are something in between these two extremes.

EDIT - was writing while Abbey was posting, hence the repetition.

Thanks very much for the replys guys.... Though I should have been a little more clear when writing my question... What I've been fiddling with is taking a k67 type single diaphragm chinese capsule and varying the voltage on it, listening to the sonic differences.  It seems as if at higher voltages the capsule starts to have more of an overall wide cardioid response.  I may be totally off, as my experimenting is not exactly scientific, and increased sensitivity makes judging polar pattern difficult.  Would one expect sensitivity increase to be linear across the freq spectrum as voltage rises?

thanks.

Increasing the bias voltage increases the electrostatic pull; the diaphragm's resonant frequency and damping increase too. The change in sound certainly makes the evaluation of directivity difficult.
The pattern is governed by the rear labyrinth and the physical dimensions. Since they don't change with voltage variations, I don't think the directivity pattern changes significantly.

After some more listening, I think you are probably right abbey, I think the changes in sensitivity, damping, and shifting of resonances are what led me to think that the pattern had changed somewhat... This was the first time I fooled with varying voltage across a broad spectrum on the capsule and really listening, and I was quite surprised to hear how significant the differences can be. 

thanks abbey and z50