DIY RF Condenser Mics

[global-records]

I wonder whether, rather than relying on bridge imbalance, it would be possible to add (a little bit of) the oscillator output to the the T2 output, so the FET gets the (vector) sum of the two. So when the bridge is one side of the balance point, they add in phase, and when it's the other side, it will partially cancel.

I suggested adding a constant voltage to the output of the bridge DIY RF Condenser Mics
Adding a generator signal is unlikely to work, the phase will not match.

By the way, remembering my experience of replacing the second transformer and C8 with a single coil (now you can easily check this thanks to the skeletons from Aliexpress) it worked just as well. Start with 10 turns.

 

[rogs]

By the way, remembering my experience of replacing the second transformer and C8 with a single coil (now you can easily check this thanks to the skeletons from Aliexpress) it worked just as well. Start with 10 turns.

Glad to read that your simulation worked out successfully in practice.......Did you make an alternative PCB to try it out, or did you modify an example of the exisitng type?
Love to see some pictures - and maybe hear some audio samples? - of your experimental results ....

 

[rogs]

Excuse my ignorance on this - I'm sure I should know! -- When a sound wave stimulates a capacitive condenser capsule, the positive pressure exerted on the membrane will force it towards the backplate, and increase the capacitance of the capsule.
As that pressure is removed on the following negative half cycle of that applied wave, the positive pressure on the membrane is relaxed, and it tends to return to it's quiescent setting.
But does the membrane ever extend beyond its resting state, and venture into negative atmospheric pressure territory, with a consequent reduction in capacitance below the quiescent value?
Or are the 'negative' half cycle movements always relative to the previous application of positive pressure, so that the capsule value is either its quiescent value or greater?

Technical papers tend to make comments like this: "Now, when the plates of the capacitor move closer due to an incident pressure and applied voltage is constant, according above equation, the electric field has to increase to compensate for the decrease in distance between the plates. Charges have to flow from the battery towards the plates to make that happen, so now the two plates can hold more charge. That is to say that the capacitance increases.
When the pressure is withdrawn, the membrane moves back to its undeflected position, the gap increases, and the capacitance decreases. To match this capacitance change, charges flow away from the plates towards the battery. In this way, as the membrane vibrates due to an incident acoustical wave or pressure, charges keep flowing to and away from the plates. The voltage source has to pump in more charges or take away charges depending on how the membrane deflects. This change in charges manifests as change in current."
That seems to suggest to me that the capcitance changes are always positive to the quiescent value? -- but I'm not sure.
I know I should know the answer, but I don't !

 

[Khron]

I would say that depends on the stiffness of the diaphragm and the damping? Its mass is of course next to nothing, but not zero, so it must have SOME momentum that wobbles it back beyond its resting position upon removal of the stimulus. That being said, i'm just speculating based on common(?) sense...

I would imagine that could be potentially tested with... What's the lowest-mass / quickest-decaying sound source that can be easily had? Possibly a piezo driver of some description, but i've heard of popping a balloon as well. Do it outdoors, add some windowing in REW and see the waterfall plot.

But, as mentioned before, I'm just spitballing here, and I may well be talking out of my rear end - I'll be the first to admit i'm no expert on this side of things

 

[kingkorg]

Excuse my ignorance on this - I'm sure I should know! -- When a sound wave stimulates a capacitive condenser capsule, the positive pressure exerted on the membrane will force it towards the backplate, and increase the capacitance of the capsule.
As that pressure is removed on the following negative half cycle of that applied wave, the positive pressure on the membrane is relaxed, and it tends to return to it's quiescent setting.
But does the membrane ever extend beyond its resting state, and venture into negative atmospheric pressure territory, with a consequent reduction in capacitance below the quiescent value?
Or are the 'negative' half cycle movements always relative to the previous application of positive pressure, so that the capsule value is either its quiescent value or greater?

Technical papers tend to make comments like this: "Now, when the plates of the capacitor move closer due to an incident pressure and applied voltage is constant, according above equation, the electric field has to increase to compensate for the decrease in distance between the plates. Charges have to flow from the battery towards the plates to make that happen, so now the two plates can hold more charge. That is to say that the capacitance increases.
When the pressure is withdrawn, the membrane moves back to its undeflected position, the gap increases, and the capacitance decreases. To match this capacitance change, charges flow away from the plates towards the battery. In this way, as the membrane vibrates due to an incident acoustical wave or pressure, charges keep flowing to and away from the plates. The voltage source has to pump in more charges or take away charges depending on how the membrane deflects. This change in charges manifests as change in current."
That seems to suggest to me that the capcitance changes are always positive to the quiescent value? -- but I'm not sure.
I know I should know the answer, but I don't !

I can not say i know for sure, but i can share couple of things.

I don't think diaphragm stops at it's neutral position. It continues to go away away from the backplate. If it were to stop, i believe we would also see it in the visual representation of the recorded waveform as clipping. Maybe it happens with some capsules? I view capsule as pressure sensor, it follows the air pressure, so it's not that diaphragm bounces away from the backplate, instead it continues to go away from the backplate as the negative pressure of the sinewave (to oversimplify it) sucks it away from the backplate. Hope it makes sense.

This is oversimplification, the diaphragm creates complex shapes, it's not like it's moving back and forth like a speaker. It's behaving more like water surface.

Here's an image of simulation software, sealed omni capsule stimulated with ca.20Khz. In this case center of the diaphragm goes beyond quiescent position.

 

[rogs]

Thank you both for your comments......

I think it is probaby unrealistic to think that the capsule capacitive value never goes below its 'resting' value?

This aspect is not normally of any real operational signficance of course, but in the case of the unbalanced RF bridge concept, fitting a balance capacitor that has a smaller value than the capsule capacitance will mean that positive pressure applied the capsule will tend to increase the value, and unbalance the bridge further. That is a good thing ... the concept relies on the capsule capacitance value not passing throught the centre 'balance' point.
The capsule value should never be less than the value of the bridge capacitor.

If the negative pressure that follows a positive pulse is always smaller - simply because it needs to 'subtract' the positive movement before it continues on its negative path - it could be that always making the bridge capacitor smaller is the preferred option?

I think some futher experiments on coil construction, to ensure a higher inductor values might be next on the agenda.
That should allow for a high 'Q' inductor system with smaller bridge capacitors....

Thanks again for you observations...

 

[kingkorg]

Thank you both for your comments......

I think it is probaby unrealistic to think that the capsule capacitive value never goes below its 'resting' value?

This aspect is not normally of any real operational signficance of course, but in the case of the unbalanced RF bridge concept, fitting a balance capacitor that has a smaller value than the capsule capacitance will mean that positive pressure applied the capsule will tend to increase the value, and unbalance the bridge further. That is a good thing ... the concept relies on the capsule capacitance value not passing throught the centre 'balance' point.
The capsule value should never be less than the value of the bridge capacitor.

If the negative pressure that follows a positive pulse is always smaller - simply because it needs to 'subtract' the positive movement before it continues on its negative path - it could be that always making the bridge capacitor smaller is the preferred option?

I think some futher experiments on coil construction, to ensure a higher inductor values might be next on the agenda.
That should allow for a high 'Q' inductor system with smaller bridge capacitors....

Thanks again for you observations...

This is a quote from Uwe on k67 capacitance swing under different SPL levels.

"For the U87 and 94 dBSPL the change in capacitance calculates to be 0.07 pF, 1.74 pF at 122 dBSPL and 5.5 pF at the maximum 132 dBSPL. Even though the capsules are acoustically identical, for the U87A these values are 0.06 pF at 94 dBSPL 0.08 pF at 117 dBSPL and 2.6 pF at the maximum 127 dBSPL."

Taken from here:
https://repforums.prosoundweb.com/index.php/topic,36194.msg531542.html#msg531542
So i guess as long as the unbalance is larger than that we're good. And that's not much of a swing in capacitance. IIRC you recommended somewhere around 8pF of imbalance, that should be plenty. I just hope we are on the same page, and I am not just embracing myself talking rubbish.

 

[ricm]

[...]

I think it is probaby unrealistic to think that the capsule capacitive value never goes below its 'resting' value?
[...]

Not unrealistic, just wrong. The capacitance will change periodically when excited with a periodical pressure variation (sound); the periodic variation of pressure and so the capacitance is the variation around a resting point and that variation is positive and negative at some time, in periodic form (sinusoidal in each fundamental frequency component, sin(wt)).

The capacitance bridge doesn't mind wich capacitor is mic or fixed one, the RF output is because imbalance. Indeed, if the fixed capacitor is larger than the mic, equal variations in mic capacitance give larger variations in RF output, so more sensitivity. Just plot the bridge output against microphone capacitance (V1*(C0-C)/(C+C0)), being V1 the transformer secondary voltage (each leg) and C0 the fixed capacitor; the derivative is just the Bandaxall paper equation number 1 at C=C0.

Also, to get resonance of the first transformer, not only the inductance can be adjusted, but also the capacitance. If you put a capacitor in primary side, it doesn't affect the bridge, but only the resonance frequency. So if that 4T:10T:10T transformer can be tuned with a secondary capacitance of say (mic 80pF+ fixed 70pF), it could be tuned with a secondary capacitance of (mic 32pF + fixed 40pF) and a primary capacitance of about 78pF*(10/4)^2 ; 488pF (very gross calculations, of course).

Just in case it could be of interest (if not wrong...).

 

[rogs]

..... Also, to get resonance of the first transformer, not only the inductance can be adjusted, but also the capacitance. If you put a capacitor in primary side, it doesn't affect the bridge, but only the resonance frequency. So if that 4T:10T:10T transformer can be tuned with a secondary capacitance of say (mic 80pF+ fixed 70pF), it could be tuned with a secondary capacitance of (mic 32pF + fixed 40pF) and a primary capacitance of about 78pF*(10/4)^2 ; 488pF (very gross calculations, of course)....

I remember trying an extra capacitance across the primary early in my experiments - although I only tried a selection of smaller values, as I recall.
Your 'crossed out' value of c.400pF is much larger than I had tried. (It's a downside of not being a very good mathematician, and only trying things out experimentally !)

Looks as if your suggestion should work well for smaller value capsules. I've done some preliminary experiments, and it seems as if a capsule with a capacitive value down as low as 33pF might still produce very usable results, while still using the original 5u3H transformer.
As a rough idea, it seems to require a primary load capcitor of between 220p and 330p to maintain the 'Q' at resonance, with the smaller secondary capacitive load. (Still trying out different combinations with different secondary loads )

An excellent suggestion .... thanks for that...

 

[ricm]

I'm glad it was of some help.

But no too high maths, I'm afraid; just some basic transformer theory. T1 is a transformer, so the impedances connected to each secondary can be seen from primary side as multiplied by (NPrimary/NSecondary)^2. Capacitances in secondary side are seen as multiplied by (Ns/Np)^2 from primary. There are more elements in the equivalent model of a transformer, of course, but it is a first approximation (and it can be confirmed by simulation, very useful LTSpice...). So if you reduce the secondary capacitances (lower mic cap) just add the equivalent of that reduction to the primary.

 

[Cqwet Dbdfte]

Interesting thread. I love adding complications. But there is beauty in simplicity.
Modulating an RF oscillator would allow a smaller capacitive change than a traditional condenser mike, AM is simple enough to "decode".
Is the complication overcomplicated? "Warum einfach..."
There are a lot of parts in this setup that can generate unwanted excitement, but maybe a fun exercise.

 

[rogs]

I'm glad it was of some help.

But no too high maths, I'm afraid; just some basic transformer theory. T1 is a transformer, so the impedances connected to each secondary can be seen from primary side as multiplied by (NPrimary/NSecondary)^2. Capacitances in secondary side are seen as multiplied by (Ns/Np)^2 from primary. There are more elements in the equivalent model of a transformer, of course, but it is a first approximation (and it can be confirmed by simulation, very useful LTSpice...). So if you reduce the secondary capacitances (lower mic cap) just add the equivalent of that reduction to the primary.

I can only apologise for my lack of transformer theory knowledge ... I should have given that some more thought, as the project progressed.
As you have probably realised if you've followed this thread, I'm very much a 'hands on' hobby builder. My theoretical knowledge is - to put it politely - a bit 'patchy'!
Gerard's original post on this thread some 5 years ago now prompted me to try a DIY RF bias mic. It's a subject I've been fascinated by for years, but it always seems to be the inductive components that provide the biggest challenge.

It was Umshankar's idea to try using 'off the shelf' IF transfomrers as project inductors, and it was Ricardo who wisely suggested I read Peter Baxandall's paper for inspiration.
So I dived in ..... There seemed to be very few suitable 'off the shelf' inductors - indeed the Spectrum coils have been to only ready made ones I've discovered so far. (One of the older TOKO 10k types also worked, but not very well.)
I've been quite pleased with some of my results, and very grateful for the input of other expert board members.
Now that we have discovered some cheap Chinese IF transformer 'blanks' it should be possible for more people to have a try at some alternative circuits.
As far as I know, other circuit suggestions to date have only been simulated, not actually constructed as physical microphones? -- I maybe wrong there of course......
Thanks again for your suggestions .. much appreciated.
Input like this to the thread certainly helps me with my 'patchy' knowledge ... and maybe helps other people with their experiments too?..

 

[dx7]

This is really an amazing project!

I'm interested to make long-term stereo AB recordings with a pair of omni RF condenser mic's in a humid environment. Would there be a good omni capsule that can be combined with this project? I do understand that R4 in the circuit would have to be adjusted accordingly: https://www.amx.jp137.com/index-scheme.html

 

[rogs]

This is really an amazing project!

I'm interested to make long-term stereo AB recordings with a pair of omni RF condenser mic's in a humid environment. Would there be a good omni capsule that can be combined with this project? I do understand that R4 in the circuit would have to be adjusted accordingly: https://www.amx.jp137.com/index-scheme.html

It depends on how much you want to spend on a capsule?..... Something like THIS from Aliexpress should work fine for some initial experiments --
If you want to 'move up a notch or two' then the Flat K.47 from Arienne Audio is a fine capsule.

You'll need to connect both sides of the capsule together for an omni mic, and that will of course double the capacitive value.
With a likely value of between 120 and 150pf you'll need to change C4 to match that (I'm guessing you mean C4 and not R4 in your post? ).
With those capacitive values for the bridge loading you might also find you'll get better results by fitting an 8MHz crystal, and changing the value of C8 from 47pF to 68pF ......

 

[dx7]

Thanks a lot for the detailed reply and the suggestions!

Yes, I meant C4 instead of R4.

By connecting the two sides of the capsule, the two opposite cardiods are added resulting in an omni isn't it?

I'll start with the aliexpress one and once it works as intended it is time to upgrade to the Arienne Audio K47's.

I would also be happy with a good quality SDC omni capsule but have no idea where to get it.

Is the suggestion to lower the RF frequency from 10 to 8 MHz (and increase C8) meant to compensate for the increased capacitance and still be able to use the 5u3H coils? If needed I'm happy to wind my own coils but it would be better to start with something given (the 5u3H's) and experiment further from there.

 

[rogs]

Is the suggestion to lower the RF frequency from 10 to 8 MHz (and increase C8) meant to compensate for the increased capacitance and still be able to use the 5u3H coils? If needed I'm happy to wind my own coils but it would be better to start with something given (the 5u3H's) and experiment further from there.

5u3H coils are better matched to an 8MHz centre frequency, when coupled with the kind of capacitive value a higher omni capsule would create.
These kind of values should help to improve the 'Q' of the tuned assembly ...
How significacnt these differences would be in practice can only really be confirmed by actually trying things out?
Although simulations might help to give an indication, there are a number of indeterminate variables that can make accurate simulations a bit difficult.

Trying out alternative coil windings is something that has only been discussed with this project fairly recently....
The availability of the cheap Aliexpress coil formers does suggest it might be a good idea to try out some different options, but - AFAIK - there have not been that many alternative arrangements suggested so far.....
All part of the whole project being 'experimental', which is how I have always considered it!

 

[dx7]

5u3H coils are better matched to an 8MHz centre frequency, when coupled with the kind of capacitive value a higher omni capsule would create.
These kind of values should help to improve the 'Q' of the tuned assembly ...

Thanks for all the information. I've decided to go for it and start the experiment with the suggested capsule!

In parallel I'm still curious about the idea of using a high quality SDC omni capsule. It's clear that that would go in a completely different direction with a lower capacitance and corresponding consequences.

 

[dx7]

I've done my first step in the process: I've built the microphone (v.5.0) using the Tzt RK-12 double membrane from aliexpress and only connected the front membrane (and temporarily removed the cable from the rear membrane). I've done capacitance testing and came to 74 pF for C4 for maximum output (just before phase reversal of the output). In this situation I've simply used the 10mHz crystal to stay with the original design and see how it behaves.
In general the microphone works very well.
As a test I've compared the output of the RF microphone to a Neumann km183 and have the following observation: Where the km183 has self noise that is more similar to pink noise the RF mic has background noise that is more similar to white noise. As a result the RF mic's self noise is louder than the km183's self noise from approximately 3kHz onwards. This is a rough measurement but clearly observable. Would there be a way to improve this?

 

[rogs]

I've done my first step in the process: I've built the microphone (v.5.0) using the Tzt RK-12 double membrane from aliexpress and only connected the front membrane (and temporarily removed the cable from the rear membrane). I've done capacitance testing and came to 74 pF for C4 for maximum output (just before phase reversal of the output). In this situation I've simply used the 10mHz crystal to stay with the original design and see how it behaves.
In general the microphone works very well.
As a test I've compared the output of the RF microphone to a Neumann km183 and have the following observation: Where the km183 has self noise that is more similar to pink noise the RF mic has background noise that is more similar to white noise. As a result the RF mic's self noise is louder than the km183's self noise from approximately 3kHz onwards. This is a rough measurement but clearly observable. Would there be a way to improve this?

Glad to hear you've had some encouraging results.... Regarding the noise figure. That is one aspect of the project that still causes some head scratching. The J113 infinite impedance detector self biases around the Vp 'pinch off'voltage, so that each cycle of RF carrier will turn the device both off and on.
It seems that the level of RF carrier will affect the noise figure - the lower the carrier level, the lower th noise.
The actual amplitude of quiescent RF carrier will depend on the degree of bridge imbalance......The problem is that making that imbalance too small can result in phase reversal of the signal. So the whole exercise is a bit of a 'balancing act' in itself!
I have experienced quite different noise levels among the 20 or so prototypes I've tried out... and it has been quite difficult to always accurately predict the results.....Which shouldn't be true of course?
This project is still pretty 'experimental', in some areas!

The mic with the lowest self noise figure I've built so far uses a low tension capsule, with a capacitive value of around 90pF, and has C4 selected with a value of 100pF. It also uses an 8MHz crystal. ... That mic has a noise figure of around 7dB(A).
I built a second mic, using identical components, but that has a noise figure some 4dB worse? ... I haven't discovered exactly why (so far)!

Always intrigued to learn what other folk discover in their experiments....

It'll be interesting to see how some of the alternative ideas suggested by others in earlier posts in this thread actually turn out - noise wise - once they've been moved from on from simulation to actual hardware prototypes...

 

[dx7]

The mic with the lowest self noise figure I've built so far uses a low tension capsule, with a capacitive value of around 90pF, and has C4 selected with a value of 100pF. It also uses an 8MHz crystal. ... That mic has a noise figure of around 7dB(A).
I built a second mic, using identical components, but that has a noise figure some 4dB worse? ... I haven't discovered exactly why (so far)!

Always intrigued to learn what other folk discover in their experiments....

It'll be interesting to see how some of the alternative ideas suggested by others in earlier posts in this thread actually turn out - noise wise - once they've been moved from on from simulation to actual hardware prototypes...

Thanks for the encouraging reply! I still have a lot to experiment with and will also go through the earlier posts to get some additional ideas.

Would you mind sharing the type of low tension capsule you used?