Better 180° rejection for dual diaphragm capsules (Hybrid Second Order Cardioid)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kingkorg

Well-known member
Joined
Apr 15, 2017
Messages
3,063
Location
Norway
I had an idea for a while i finally got to test yesterday.

The idea was to improve rear (180°) rejection of a dual diaphragm mic with a k67 capsule. The same can be done with any dual diaphragm capsule.

The mic i used here has dual output. Each side of the capsule has it's own circuit and output. Think of TL Sphere, Austrian Audio OC818...

This is cardioid response of one of the sides. 0° (red) and 180° (green). Typical response in k67 based cardioids. Note that the responses are not absolute, I haven't used speaker compensation, not necessary for this test.

Screenshot_20230313_172553.jpg


The idea is that if i activate the rear diaphragm, introduce some signal out of phase at just the right level, apply EQ curve so that it matches the response of the rear (green) curve but with everything cut below say 3K i should get additional cancellation from the rear and bring that HF peak down in the green line without affecting the front (red) response.

And indeed, I was right. If the signal from the rear is about 15db attenuated compared to the front, phase reversed, of the right shape, suddenly you get the purple line as the 180° response.

Considerably better rejection in the high end compared to the green line! And the best thing, it didn't affect the front response. Slight bump between 2k and 3K, but that's because of the phase shift of the HP filter. The corrective EQ curve can be tuned better to avoid this.

Screenshot_20230313_172609.jpg


And this is the EQ curve I applied to the rear membrane to get the best rejection. It is roughly similar to the green line, but fine tuned because the EQ's HPF introduces phase shift.

Screenshot_20230313_172618.jpg


I hope that made sense. I did some audio testing as well, and i've never heard a mic reject that well from the rear.

It is important what EQ you use because different plug-ins handle phase shift differently. With Q3 i could test different modes, and it worked best in zero latency mode.

The result would be even beter if time delay was adjusted between front and back so that you virtually bring rear diaphragm closer to the front one.

This is so different to using a hypercardioid. Hypercardioid changes the shape of the 0 degree response compared to a cardioid capsule of the same construction, and lets more of the signal through at exactly 180° in order to get more rejection from the sides. I would call this for a hybrid second order cardioid. I'm not aware of anything like this commercially available.
 
Last edited:
Do you get any useful rejection with a 24db/oct high pass at 1k5 and a 6db/oct low pass at 15k?
If so, is there any hope of an analog approximation of this approach inside the mic?
 
Do you get any useful rejection with a 24db/oct high pass at 1k5 and a 6db/oct low pass at 15k?
If so, is there any hope of an analog approximation of this approach inside the mic?
I experimented with various hpf slopes, this gave me best results. It could be it's not necessary to go that steep, and some other capsules might need way different shapes. So probably possible, but very tricky. This was quick and dirty test. It would need some serious R&D. I threw the idea out there, hopefully someone will find it useful, and possibly take it further. This could easily become my go to approach for miking snare.
 
I wish more of us had access to the separate sides of double-diaphragm capsules. I worked on a Lomo (Russian) mic some years ago that amplified both sides of the diaphragm separately, then took them into an external pattern select device. I particularly appreciate that you used EQ to match response patterns of the two diaphragms and I can easily see some of the potential uses for adding the two signals together using custom tailored responses.

I wrote up a short description of how to "optimize" cardioid response in mics that sum the two diaphragms together before the amplifier circuitry, but it only deals with balancing the signal strength of the two diaphragms. For those who might be interested, here it is:

I'm wondering if builders on this forum "optimize" the cardioid pattern of their double-diaphragm, variable pickup pattern designs. When I'm setting up the variable polarizing voltage for the rear diaphragm, I experiment to see what exact voltage yields the least output from the rear of the microphone. I consider that to be the “optimized” cardioid response available from that capsule. Many of you will know that quite a few double diaphragm microphones are slightly hyper-cardioid when the rear diaphragm is held at exactly one-half of the variable polarizing voltage (VPV). Some of the Neumann microphones with K47 capsules come to mind as a highly regarded and well known examples of this.

If a microphone has a stepped system with nine pattern options, I put the pattern selector in the cardioid position. In most commercial designs this will feed an exactly equal amount of polarizing voltage to both diaphragms. Then, I place a steady sound source directly behind the microphone (180 degrees off-axis), usually at a distance of 8-16 inches. [Old fashioned guitar tuners with an annoying 440 Hz tone work very well for this.] Then, I write down the preamp gain setting and monitor the output of the microphone to get an accurate measurement of its level. This needs to be done in a very quiet environment, so my heater, refrigerator, and any other noise sources are off when I’m doing this part of the work.

I then move the pattern select switch one click toward the omni setting. If the output falls, I know that the capsule has a slightly hyper-cardioid response when in the cardioid position. If I move the selector one click toward the bi-directional setting and the output voltage falls, I know that the capsule has a slightly omni directional response when the pattern select is in the cardioid position. (I won't go into why this happens, but trust me, it does. Better yet, test your multi-pattern mics and see what they're doing.)
To modify a stepped pattern select circuit so that the microphone has the greatest amount of off-axis rejection when the pattern selector is in the cardioid position, add a potentiometer to the pattern select circuit in the following manner (usually 250K or so will do the job and using a linear response pot, rather than one with an audio taper, will make your testing easier and more accurate). If the capsule showed a slightly hyper-cardioid pattern when in the cardioid position, then detach the polarizing voltage source from the pattern select circuit and attach the voltage source to the first or third leg of the potentiometer. The wiper (the middle leg) of the potentiometer gets connected to the point in your pattern select circuit where the voltage source had previously been attached. When this has been done, make sure the pattern select circuit is back in the cardioid position and twiddle with the potentiometer until the output from the rear of the mic falls to its lowest level. This is the capsule’s optimized cardioid response. Remove the potentiometer and measure the resistance that got dialed-in between the wiper and pattern select circuitry (make sure that you don’t jostle the potentiometer and change its setting before you make this measurement and that you’re measuring between the wiper and the same leg of the potentiometer that you attached to the voltage source). Find a fixed resistor that is as close as you can get to the resistance you measured on the potentiometer and ADD IT IN SERIES to the resistor that is closest to the “bi-directional” setting of the pattern selector (this resistor is already part of your pattern select circuit). The added resistor will cause the circuit to deliver less voltage to the rear diaphragm than it did previously, except when it’s in the omni or bi-directional setting. Summary: you’ve added resistance between the last two “clicks” of your pattern select circuit, the ones closest to the hyper-cardioid setting. Now, your patterns will still go from onmi through cardioid to bi-directional, but when set in the cardioid position, you’ll get the same minimum output from the rear of the microphone as you achieved when experimenting with the added potentiometer. This is as close to a definitive cardioid pattern as your circuit can achieve.

If your capsule has a tendency to be a bit omni-directional when in the cardioid setting, add the test potentiometer to “the bottom” of your pattern select circuit (i.e. between the resistor that is nearest the omni setting and ground, though this is not likely to be the case with most double diaphragm capsules). Then do the same measuring at the rear of the mic to determine what setting on the potentiometer gives the best rear rejection. When you find the optimum resistance, add that amount of resistance IN SERIES to the first resistor in your pattern select circuit (the one closest to ground).

If the pattern select circuit is continuously variable and uses a potentiometer with a center detent to indicate where cardioid is, then you just set your pattern select knob at the center detent and use the test potentiometer to add resistance to whichever side your pattern select circuit requires (between the existing potentiometer and the voltage source to fix a hyper-cardioid tendency or between ground and the existing potentiometer to fix an omni tendency).

Another perfectly good approach is to take a one-meg (or higher) linear potentiometer and substitute it for the entire pattern select circuit. Somewhere near the middle of the rotation of the potentiometer, you'll see that sounds arriving at the rear of the microphone produce the least amount of signal. Without turning the potentiometer AT ALL, turn off the mic, unhook the test potentiometer from the circuit and measure the resistance from pin one to the wiper (pin two), and from the wiper to pin three. The ratio of these two resistances is the exact ratio you want to establish in your pattern select circuit when it is in the cardioid position.

Will “optimizing” your cardioid pattern mess up your other patterns? No. So long as you hold your backplate voltage at one-half the available variable polarizing voltage, you’ll get an omni directional pattern when you feed 0 Volts to the rear diaphragm and you’ll get a bi-directional pattern at full polarizing voltage. And not only will the cardioid pattern have been compensated, but all the in-between positions were also compensated when you re-balanced the resistors in your original circuit.

Have fun with it!
 
I wish more of us had access to the separate sides of double-diaphragm capsules. I worked on a Lomo (Russian) mic some years ago that amplified both sides of the diaphragm separately, then took them into an external pattern select device. I particularly appreciate that you used EQ to match response patterns of the two diaphragms and I can easily see some of the potential uses for adding the two signals together using custom tailored responses.

I wrote up a short description of how to "optimize" cardioid response in mics that sum the two diaphragms together before the amplifier circuitry, but it only deals with balancing the signal strength of the two diaphragms. For those who might be interested, here it is:

I'm wondering if builders on this forum "optimize" the cardioid pattern of their double-diaphragm, variable pickup pattern designs. When I'm setting up the variable polarizing voltage for the rear diaphragm, I experiment to see what exact voltage yields the least output from the rear of the microphone. I consider that to be the “optimized” cardioid response available from that capsule. Many of you will know that quite a few double diaphragm microphones are slightly hyper-cardioid when the rear diaphragm is held at exactly one-half of the variable polarizing voltage (VPV). Some of the Neumann microphones with K47 capsules come to mind as a highly regarded and well known examples of this.

If a microphone has a stepped system with nine pattern options, I put the pattern selector in the cardioid position. In most commercial designs this will feed an exactly equal amount of polarizing voltage to both diaphragms. Then, I place a steady sound source directly behind the microphone (180 degrees off-axis), usually at a distance of 8-16 inches. [Old fashioned guitar tuners with an annoying 440 Hz tone work very well for this.] Then, I write down the preamp gain setting and monitor the output of the microphone to get an accurate measurement of its level. This needs to be done in a very quiet environment, so my heater, refrigerator, and any other noise sources are off when I’m doing this part of the work.

I then move the pattern select switch one click toward the omni setting. If the output falls, I know that the capsule has a slightly hyper-cardioid response when in the cardioid position. If I move the selector one click toward the bi-directional setting and the output voltage falls, I know that the capsule has a slightly omni directional response when the pattern select is in the cardioid position. (I won't go into why this happens, but trust me, it does. Better yet, test your multi-pattern mics and see what they're doing.)
To modify a stepped pattern select circuit so that the microphone has the greatest amount of off-axis rejection when the pattern selector is in the cardioid position, add a potentiometer to the pattern select circuit in the following manner (usually 250K or so will do the job and using a linear response pot, rather than one with an audio taper, will make your testing easier and more accurate). If the capsule showed a slightly hyper-cardioid pattern when in the cardioid position, then detach the polarizing voltage source from the pattern select circuit and attach the voltage source to the first or third leg of the potentiometer. The wiper (the middle leg) of the potentiometer gets connected to the point in your pattern select circuit where the voltage source had previously been attached. When this has been done, make sure the pattern select circuit is back in the cardioid position and twiddle with the potentiometer until the output from the rear of the mic falls to its lowest level. This is the capsule’s optimized cardioid response. Remove the potentiometer and measure the resistance that got dialed-in between the wiper and pattern select circuitry (make sure that you don’t jostle the potentiometer and change its setting before you make this measurement and that you’re measuring between the wiper and the same leg of the potentiometer that you attached to the voltage source). Find a fixed resistor that is as close as you can get to the resistance you measured on the potentiometer and ADD IT IN SERIES to the resistor that is closest to the “bi-directional” setting of the pattern selector (this resistor is already part of your pattern select circuit). The added resistor will cause the circuit to deliver less voltage to the rear diaphragm than it did previously, except when it’s in the omni or bi-directional setting. Summary: you’ve added resistance between the last two “clicks” of your pattern select circuit, the ones closest to the hyper-cardioid setting. Now, your patterns will still go from onmi through cardioid to bi-directional, but when set in the cardioid position, you’ll get the same minimum output from the rear of the microphone as you achieved when experimenting with the added potentiometer. This is as close to a definitive cardioid pattern as your circuit can achieve.

If your capsule has a tendency to be a bit omni-directional when in the cardioid setting, add the test potentiometer to “the bottom” of your pattern select circuit (i.e. between the resistor that is nearest the omni setting and ground, though this is not likely to be the case with most double diaphragm capsules). Then do the same measuring at the rear of the mic to determine what setting on the potentiometer gives the best rear rejection. When you find the optimum resistance, add that amount of resistance IN SERIES to the first resistor in your pattern select circuit (the one closest to ground).

If the pattern select circuit is continuously variable and uses a potentiometer with a center detent to indicate where cardioid is, then you just set your pattern select knob at the center detent and use the test potentiometer to add resistance to whichever side your pattern select circuit requires (between the existing potentiometer and the voltage source to fix a hyper-cardioid tendency or between ground and the existing potentiometer to fix an omni tendency).

Another perfectly good approach is to take a one-meg (or higher) linear potentiometer and substitute it for the entire pattern select circuit. Somewhere near the middle of the rotation of the potentiometer, you'll see that sounds arriving at the rear of the microphone produce the least amount of signal. Without turning the potentiometer AT ALL, turn off the mic, unhook the test potentiometer from the circuit and measure the resistance from pin one to the wiper (pin two), and from the wiper to pin three. The ratio of these two resistances is the exact ratio you want to establish in your pattern select circuit when it is in the cardioid position.

Will “optimizing” your cardioid pattern mess up your other patterns? No. So long as you hold your backplate voltage at one-half the available variable polarizing voltage, you’ll get an omni directional pattern when you feed 0 Volts to the rear diaphragm and you’ll get a bi-directional pattern at full polarizing voltage. And not only will the cardioid pattern have been compensated, but all the in-between positions were also compensated when you re-balanced the resistors in your original circuit.

Have fun with it!
Excellent stuff!!!
 
These ideas are really innovative, i love it :)
I could imagine a system with soliloqueens K47 capsule, my MicUlli input stage for SDCs (doubled) and a VST plugin with EQ/Mixing/delay to get all the responses fit.
 
These ideas are really innovative, i love it :)
I could imagine a system with soliloqueens K47 capsule, my MicUlli input stage for SDCs (doubled) and a VST plugin with EQ/Mixing/delay to get all the responses fit.
I have no idea what happened to the images, the op doesn't make much sense without them.

Edit:
The op updated with the images, hopefully they will stay there this time.
 
Last edited:

Latest posts

Back
Top