Optical mic noise

[crazydoc]

Here's an update. In very brief testing with the original sensor (Fairchild QRE1113.GR) I've found no significant change by varying the resistor values for either the LED or the phototransistor.

Anyway, I've put together a prototype system to be able to test various devices and configurations.

This first photo is an overview of the system.
Power supply, preamp and headphone amp in the background.

These are the boards for the five different sensors I'll be testing. They were some little boards I found in a drawer that I've had for 30+ years. If you keep something long enough, and live long enough, you'll find a use for it.

Here's a closeup of one of the sensors, a TAOS TRS1722 (photodiode.)

Here's another, the Marktech MTRS9520.

Here's one of the boards, mounted on its base.

This shows a thumbwheel (castor cup) and the spring loaded screw for moving the sensor closer to or farther from the membrane.

This shows the five sensors mounted and wired up.

This is a membrane assembly, using aluminum leaf, a section of rubber PVC connector, and a ring of 1 1/2 inch PVC pipe. The leaf is glued onto the rubber ring while it is compressed with a hose clamp, and then stretched as the hose clamp is released.
This stuff is really difficult to work with as it is so light and fragile. I'd really like some one micron aluminized mylar to work with for its tensile strength.

Here's the bottom side of the assembly.

And here's the finished assembly, ready to fire up. The membrane unit just sits over the sensor board, whose distance from the membrane is controlled by the thumbwheel.

 

[Guest]

Crazydoc,
you have serious problem with microphone
construction. This problem had explained to you by Marik,
but you not accepted it in your latest designe.
Here is microphone classification:
(you must cut and paste link below to the browser to view it without
blahblahs)
http://mujweb.cz/www/xvlkxvlk/mic.gif
Your design is optical and principial mechanical value which is converted
to electrical is displacement, your convertor /not transducer-transducer
is reciprocal or antireciprocal/ is condenser-like.

If you can have you transducer nondirectional and frequency independent,
you must have stiffness control of membrane in the acoustic band and
resonancy of membrane UPPER that higest frequency and critically damped.
This is done in capacitor measurement omnidirectionals, where backplate have two function. To damp highly stretched membrane and to transduce into elecrical signal.

With aluminium membrane you can do in the better way gradient microphone /directional/. This microphone must have two inputs = make holes in the backplate. You must also damp membrane to obtain resistance control. And you must tune membrane to the middle of acoustic band.
The only method to do it is to use acoustic resistance material very close
to back of the membran /several micrometers/. This acoustic material
can be glass-sand porously glued with epoxy resin. Into this material you
can bake on your reflective optocouplers.
opposite side of the resin must be open=gradient mic.
If you will bake-on backplate with optocouplers on glass plate,
you can obtain micron tolerances. If you stretch MICA, you can tune
it in the middle of acoustic band and with some MICA spacer you will have
high quality mic with a bad sensitivity.

xvlk

 

[crazydoc]

Thanks xvlk.

I will file away your comments for later use where they may be helpful.

At this point in my attempts to make this mic, capsule construction and tuning are far down the road. If I can't get a handle on the noise produced by the transducers, the rest is moot - there will be no mic. The problem I am having is the noise produced by the phototransistor, so I have built this apparatus to be able to vary the electrical parameters and components I'm trying out, including photodiodes. The membrane is obviously quite primitive and has no tuning or other acoustic modifications applied to it. At this point it's just there to give me a signal I can evaluate for noise.

If you have any suggestions for noise reduction I'd be very interested to hear them. Thanks again, and I appreciate your input.

 

[Guest]

[quote author="crazydoc"]Thanks xvlk.

I will file away your comments for later use where they may be helpful.

At this point in my attempts to make this mic, capsule construction and tuning are far down the road. If I can't get a handle on the noise produced by the transducers, the rest is moot - there will be no mic. The problem I am having is the noise produced by the phototransistor, so I have built this apparatus to be able to vary the electrical parameters and components I'm trying out, including photodiodes. The membrane is obviously quite primitive and has no tuning or other acoustic modifications applied to it. At this point it's just there to give me a signal I can evaluate for noise.

If you have any suggestions for noise reduction I'd be very interested to hear them. Thanks again, and I appreciate your input.[/quote]

You have no problem with noise in photodioed etc. I think, that
signal to noise ratio of your photo is good. In signal to noise
ratio you must compute noise amount to all dynamic range of device.

If maximal dynamic range of your device is 1 inch, you must
compute dynamic ratio with output noise scaled to 1 inch.
If it is 120 dB, it is O.K. I believe, that it is.

You have problem with circuit dynamics.
Rephlective optics of your optocouplers is designed to
respect displacement range 10000 micrometers.
You must use optics /i.e fibre optics/ designed to range 0.1 micrometers.
This is 100 dB gain. simply add this value to your measurement and

Problem is, how stabilize and set this coupler. But if you vant to set it
in the laboratory, it need not be problem /for 10 sec. measurement/.

only do this: use fibre optics baked on the acoustic stone
and you will have perfect, but some dead and fragile, mic.

xvlk

 

[crazydoc]

[quote author="xvlk"]
You have no problem with noise in photodioed etc. I think, that
signal to noise ratio of your photo is good. In signal to noise
ratio you must compute noise amount to all dynamic range of device.

If maximal dynamic range of your device is 1 inch, you must
compute dynamic ratio with output noise scaled to 1 inch.
If it is 120 dB, it is O.K. I believe, that it is.[/quote]
Sorry, you lost me with the 1 inch dynamic range. Please explain. I don't have much of an electronics background, so please use as simple terms as possible.

You have problem with circuit dynamics.
Rephlective optics of your optocouplers is designed to
respect displacement range 10000 micrometers.

Again, I'm not sure what you mean. It looks to me like the device (whose graph is in this thread) has a range of distance from the reflective surface of 0 to about 0.5mm (500 micrometers) and I can easily approach it to within 200 micrometers or less of the membrane's surface.

You must use optics /i.e fibre optics/ designed to range 0.1 micrometers.
This is 100 dB gain. simply add this value to your measurement and

Are you saying that the surface of the fiberoptic lightguide must approach to within 0.1 micron of the membrane? First, I don't see how this is possible (and I'm fairly sure the membrane excursion in response ot acoustic waves would be greater than this and "bump" into the sensor), and second, I don't understand what advantage is obtained by closeness to the membrane. If the sensitivity of the device is the same at 1 or 400 microns (that is, the slope of the current vs. distance curve is the same), then the distance within this range should make no difference.

Maybe you're saying that the sensor should be able to detect displacements in the 0.1 micron range? Yes, that would be nice - I don't know what the actual sensitivity of these devices is.

Problem is, how stabilize and set this coupler. But if you vant to set it
in the laboratory, it need not be problem /for 10 sec. measurement/.

only do this: use fibre optics baked on the acoustic stone
and you will have perfect, but some dead and fragile, mic.

xvlk

Sorry, you lost me here again on the "10 sec. measurement" phrase in particular.

Again, thanks for your input. Please explain your concepts to me as you would to a child, so I have a better chance of comprehension. :green:

 

[Guest]

[quote author="crazydoc"][quote author="xvlk"]
I understand and I had some hyperbolas in previous mail, but
I can explain it you more rudimentally.
Yes, you can have better response of UNDAMPED membrane that
1 um, and have also unstability /this is your big bumps/.
But if your will have membrane resistance-controlled /and you must
have it for directive mic with uniform frequency ratio/ you must have
displacement in units of micrometers for maximal acousticaly strong
source. If you can record semtex explosions, displacement can be bigger.
If you will have regulary damped membrane, you can pick up displacement of it with standard electornics. And standard electronics have
LIMITED dynamic range. And if you want to use economically this limited dynamic range, you must use some as level-matching.
And to do level match of membrane displacement to standard optoelectronics is to use good optics. If you not want to do this
level-matching, you can not use standard electronics. Yes, here are
parametric amplifiers, quantum amplifiers and van-de Graaf generators and so on... .
Yes, you can have 1000 cubic meters of studio volume and 10 000 cubic meters of amplifier after mic volume.

xvlk

P.S.: Some circuits is impossible to import to the United States and
otrers civilised countries, mainly because are builded mainly for atomic research.

 

[crazydoc]

Thanks xvlk

I guess what you're saying is that I need to use a more sensitive optical system to improve the signal to noise ratio if I'm going to used standard acoustically tuned and damped membranes. You are probably right. However for now I will forge on along my current path until I come to a dead end. Then I may have to backtrack and take another direction.

As one of our famous American philosophers has said: "If you come to a fork in the road, take it."

 

[Guest]

[quote author="crazydoc"]Thanks xvlk

I guess what you're saying is that I need to use a more sensitive optical system to improve the signal to noise ratio if I'm going to used standard acoustically tuned and damped membranes. You are probably right. However for now I will forge on along my current path until I come to a dead end. Then I may have to backtrack and take another direction.

As one of our famous American philosophers has said: "If you come to a fork in the road, take it."[/quote]
...yes, good idea, it is as algorithm of the mouse in the labirinth
(is scholar example for programmists) to go along one side of the wall.
But mouse can leave this labirinth, but only if it is finite /and
only if it is without loops/.
But in the real word some things are infinite. /example- space/.
And we /peoples/ live only in the order of 10 years.
I think, that posibilites to make microphones are so infinite and
can not be solve by hard-force.

xvlk