Noise in HF condenser microphone

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Guest

Guest
Hi All,
I had interesting project for christmas time: to DIY Schoeps
type omnidirectional high-frequency mic. (working at some 8MHz)
I decided to Schoeps topology (two resosnant circuits in
one oscillator, where one is tuned by capsule and one by capacitor
trimmer) because of simplicity and minimum components.

Firstly I set up capsule, original capsule was Mylar, cardioid,
I modified to Al diaphragm ( motion controlled by backplate structure/air compliancy) omni.

Capsule porn:

http://www.rudolfinea.cz/capsorig.jpg

Epoxy filling:

http://www.rudolfinea.cz/epoxy.jpg

Al diaphragm pretension (micron thick, pure Al, I use it for ribbons :))) )

http://www.rudolfinea.cz/aldiaph.jpg

After I mod capsule, I mod the mic:

http://www.rudolfinea.cz/vfmictop.jpg

http://www.rudolfinea.cz/vfmicbot.jpg


And it is result :-((((((.....((((
40 dB dynamic range. BAAAAAD.


http://www.rudolfinea.cz/mic1.wav


Can someone help me???, PLEASE...

xvlk
 
> 40 dB dynamic range. BAAAAAD.

Can you try it with a simple audio buffer, to be sure the noise is not in the capsule?

If not: FM mike systems are notorious for noise problems. I had a pair of original Stephens RF condensers and they hissed awful. I could get a little less or a lot more hiss by fooling with the trimmers, but I think everything has to be perfect to get FM noise down lower than a simple analog buffer. I am not the RF-head to figure out why. I'm not sure why you would want an RF mike for audio, since we can usually make a buffer go to 20Hz. (Sub-sonic test mikes do need FM techniques to go toward DC response.) I know one company specializes in FM condensers, but I think they just like the challenge.
 
[quote author="PRR"]
Can you try it with a simple audio buffer, to be sure the noise is not in the capsule?
[/quote]
I will do it, mainly for capsule sensitivity check.
by the way, is capsule leakage noise problem in low impedance RF mics???
[quote author="PRR"]
FM mike systems are notorious for noise problems.
[/quote]
I have read papers, where HF principle was used for measuring Brownian noise.
[quote author="PRR"]
I had a pair of original Stephens RF condensers
[/quote]
What is it??? may be interesting to get some info... Please...
[quote author="PRR"]
I'm not sure why you would want an RF mike for audio
[/quote]
One and only one reason: Recording in humid enviroment.
I am not friend of silica gel
or other hell,
used by Bruel and Kjaer.
[quote author="PRR"]
I know one company specializes in FM condensers, but I think they just like the challenge.[/quote]
Do you mean S*nnheiser with Griese/Hibbing the designers ??? Interesting, but
somewhat bulky for my mic.

xvlk
 
Simple buffer finished, but my capsule
is so "temperamental" = veryvery noisy to seriously measure some.
I designed Hf capsule, not LF.
But it is possible, that somehow noisy capsule in LF
can also {but orders of lower} be noisy in HF, but with somewhat different
noise spectra???? dont know, I must do some measurements at home...
Maybe to check HF circuit with input capacitor.

xvlk
 
I thought the main idea behind the RF mics was to get more sensitivity by putting the diaphragm closer to the backplane. The AC bias means that you don't have an attraction to the backplane, so less chance of the diaphragm collapsing. I know the sennheiser capsules are quite a bit different than their DC biased counterparts. I think they also use some active EQ to balance their mics and lower the noise.
 
[quote author="BradAvenson"]The AC bias means that you don't have an attraction to the backplane, so less chance of the diaphragm collapsing. [/quote]
What??? Tha backplate and diaphragm also attracts in HF condenser microphone. Attraction occurs, like in the electrostatic voltmeter, which
measures true effective (RMS) value, with quadrat of voltage.
If you have 40 V DC polarization, you have the some atteraction
like 40 V effective.
Sennheiser have 42 V (eff) I have somewhat about 50 V eff.
Problem can be with so called quasistatic stability:
In DC polarised microphone at normal (not charge coupled amp) topology
you have system with constant charge
In the HF microphone of (sennheiser)Griese type you have hard HF voltage on the capsule and then it behaves like constant voltage system.
But at Schoeps type problem is dependent to adjust. I still not understand it, That is, why I choose it.

xvlk
 
[quote author="PRR"]
I could get a little less or a lot more hiss by fooling with the trimmers[/quote]
PRR, thanx a lot,
after fooling trimmer, i got 57 dB SNR:
www.rudolfinea.cz/mic2.wav
:)
Yes, there is different trimmer position
for maximum signal / minimum noise in RF microphones.

By the way, if RF principle can use old temperamental capsules
veryvery optimally (as I have proved experimentally),
then someone can use old DurAL diaphragm Wente microphones
from 1920s with excellent parameters in RF connection with parameters
that Wente could not heard.

... only idea.....

xvlk
 
I'm a little confused by all this...

A HF (High Frequency) mic to me is somthing that picks up 22kHz, maybe a bit more.

I don't understand an 8MHz mic.

Why use a capsule? (thats for sound pressure, not electromagnetic rad.)

Does the mic have an IF stage? What "comes out" of this mic?

shouldn't this mic just be an antenna?

thanks!
 
brad,

the mic isnt detecting rf. the mic diaphragm (capacitor) frequency modulates a vhf oscillator. the carrier is then demodulated and presto... audio.

the idea is to get a large (audio) amplitude change with only a small change in capsule capacitance. at megehertz range it takes less than picofarad of change to shift frequency dramatically.

in essence an rf mic is a closed circuit fm transmitter. (exciter is the proper term but whatever...)
 
Wow,

this sounds really cool. What kind of scheme are you using for demodulation?

Is the point of all this to get large swings of audio output?

very cool
 
[quote author="bradb"]What kind of scheme are you using for demodulation?
[/quote]
Schoeps mics are not Modulation/demodulation.
They works rather like Grid-dip meter (one resonation
tank, tuned by microphone takes energy from other tuned
by trimmer or varactor for better stability)
[quote author="bradb"]
Is the point of all this to get large swings of audio output?
[/quote]
Yes, original Schoeps circuit works only at voltage of diode
treshold +-0.4 V. If you use additional Zeners or LEDs, output
swing can be larger. I.E. some +-2V (but it is some Hibbing s/ Sennheiser patent :-( )
here is schemo of modded Schoeps:
http://www.rudolfinea.cz/schps.tif
[quote author="bradb"]
very cool
[/quote]
Thanx, but no so cool. It had sensitivity/noise like normal small diaphragm
mic. (approx. 30 dB self noise) But for cool mic you are limited only by
brownian motion of air. cool mic must have some -(minus)10 dB self noise.
Parametric amplifier have capability to obtain -10 dB self noise from
small diaphragm mic if it is pressure and properly designed. This will be cool mic.
...I am working on it :) and I will finish work at April 1-st 2007

xvlk
 
Gee I thought the RF mics are a free running oscillator with the diaphragm acting as the variable C in the LC circuit/tank, and with some kind of discriminator. I suppose my Millen grid dip meter is, essentially, a discriminator. Trouble with lots of RF around sensitive preamps is raised OA noise floor. A good design would have the RF stage completely isolated, choked, decoupled, and bypassed, from the rest.

Pointer to the schematic?

bbob

htttp://microphonium.blogspot.com
 
[quote author="BBob"]Gee I thought the RF mics are a free running oscillator with the diaphragm acting as the variable C in the LC circuit/tank, and with some kind of discriminator. [/quote]
I never had seen that topology. Maybe problems with
oscillator frequency sensitivity to change microphone C.

By the way there is some patent for ?40 MHz? oscillator tuned by microphone, next mixer and demodulation on IF. But
will be problems with local oscillator.
[quote author="BBob"]
I suppose my Millen grid dip meter is, essentially, a discriminator.
[/quote]
yes, it can be, but there is simplier way to do it (to measure directly
grid or gate dip)
[quote author="BBob"]
Trouble with lots of RF around sensitive preamps is raised OA noise floor. A good design would have the RF stage completely isolated, choked, decoupled, and bypassed, from the rest.
[/quote]
I have it done. Without it nothing can be measured.
[quote author="BBob"]
Pointer to the schematic?
[/quote]
www.rudolfinea.cz/schps.tif does not woks??? For me it works normally.
[quote author="BBob"]
htttp://www.microphonium.blogspot.com[/quote]
Nice site :)

xvlk
 
Thanks. Yes that tif works but I thought that was just the mod and didn't know if it was the same/similar to the base design.

The old James Millen grid dip meter reacts to the change in grid current as the resonant circuit is loaded by whatever it gets coupled to. The coupling in that case is mainly inductive but as you increase frequency it acts more like cap coupling and therefore is even more proximity (motion) sensitive. There are a bunch of different coils in the kit that you plug in for the various ranges.

The reason I'm interested in all this is because I am working on a nanotube microphone now that is basically a carbon whisker array that moves in response to sound, and the array (a forest of nanotubes) act like variable caps in parallel. It works fine at DC, but I'm thinking of RF also.

bbob


http://microphonium.blogspot.com
 
[quote author="BBob"]the base design. [/quote]
I dont know base design of Schoeps.
I know only principial block schemo, where only
two tuned circuit and inductively coupled detector
was drawed.
My mod are that LED diods for make larger output voltage.
If you omitt the two transistors (oscillator) and short that LEDS, you
have "original" schemo.

And good luck to develop nanotube mic. It is nice work and lot of work,
and much of work was forgotten, i.e. piesopolymer microphones :)

xvlk
 

Latest posts

Back
Top