Help: Measure Condenser Capsule Voltage Signal

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

tynkerd

Member
Joined
Oct 17, 2018
Messages
5
Hi guys, I'm new on the forums here, and I'm just getting started with mics as a hobby.  :)

I bought a cheap Chinese condenser capsule (appearance is similar to the RK-87 by micparts), and am trying to figure out how to read the output voltage, to make sure the thing even outputs a signal voltage.

Basically, I know these are high-impedance, so they can't deliver much current. Because of this, I can't directly hook up power and use my oscilloscope to measure AC signal variations, because the 1M ~ 10M ohm load draws too much current.

So, I've made a super-simple circuit (attached), with:
-1G-ohm resistor connecting the capsule and +40V
-100nF capacitor to separate only the AC signal
-1G-ohm resistor to pull the bias of the JFET to 0V
-JFET has a 10k-ohm resistor to +40V at the drain (the JFET draws ~3-4mA @Vgs = 0V)
-Vdrain sits at around 2.xxx[V]

I was expecting to see a signal when I clap or shout or do something, but I don't notice ANY changes.
All my oscilloscope shows is this 50Hz jibberish, probably from the power supply or something...

Any hints on how to measure a mic capsule to see if it is even outputting signal changes when sound pressure waves pass through, would be hugely appreciated! Thank you so much!

I look forward to jumping into the community here, and being a benefit once I get the basics out of the way. 
 

Attachments

  • mic.png
    mic.png
    16.7 KB · Views: 34
If the capsule is in free air, it might be picking up voltage fields from the surrounding environment, and not just the capsule itself. If you put the capsule into a grounded mesh enclosure, similar to what a capsule is normally mounted inside of, this random noise coupling might get shielded away.

I have no direct experience with this, but also, I've never biased up an unshielded condenser capsule and listened to the output. All condenser capsules I've listened to were mounted inside of screened enclosures. So, it seems worth a try. Besides, who's gonna run a capsule flapping in the breeze anyway? Best of luck!
 
With 0V Vgs the JFET is turned on hard....

You are probably missing a component or more...

How about a few K ohm resistance between JFET source and ground? This will give the gate (at 0V) a negative bias voltage wrt source that is now slightly positive.

or not...

JR
 
Monte McGuire said:
If the capsule is in free air, it might be picking up voltage fields from the surrounding environment, and not just the capsule itself. If you put the capsule into a grounded mesh enclosure, similar to what a capsule is normally mounted inside of, this random noise coupling might get shielded away.

I have no direct experience with this, but also, I've never biased up an unshielded condenser capsule and listened to the output. All condenser capsules I've listened to were mounted inside of screened enclosures. So, it seems worth a try. Besides, who's gonna run a capsule flapping in the breeze anyway? Best of luck!

This is an interesting consideration. I'm currently working on 3D printing a mount, so I can mount it in a donor body I have lying around.

I will also try adding a resistance at the source, and try playing around with the voltages.

What gets me though, is that I was seeing absolutely no change to the output signal I was getting, no matter how hard I clapped or spoke or w/e...

Let's see what these mods do.
 
I got it!!! Kind of...

So, long story short...full shielding of the capsule solved my issue.

In summary, I have a BM-800 donor body mic, and after 3D printing a stand and affixing the new capsule, I ran a 1G-ohm resistor from the phantom power supply to the capsule, and then used a 10nF cap to the JFETs gate.

Then, I screwed the body back on so it was fully shielded, and plugged the mic in to my preamp. A quick recording in Audacity showed a low noise floor, and the mic was able to pick everything up great!

It sounds a little muddy, but that is because the filtering for the BM-800 circuit doesn't match what this capsule needs, so I need to revamp the circuit.

It works!

I'm attaching a screenshot.

Next up, making a circuit, now that I know the capsule picks up sound waves alright!
 

Attachments

  • Screenshot_1.png
    Screenshot_1.png
    641.2 KB · Views: 38
You cannot measure the DC or AC voltages of a mic capsule using a conventional probe. You need something that is ultra-high impedance like 100G. Also any capacitance at all is going to throw off your measurements.

Just thinking out loud here but this is a back-of-napkin sketch of something that might work:

MicCapsuleJfetProbe2.png


In practice I would put the jfet by itself at the end of a very well shielded 2 conductor cable. You could actually just use the gate pin sticking out as the probe tip.

I have no idea if this would work though. And the JFET would have to be able to handle 60V (or at least a few volts above whatever voltage you want to measure).
 
squarewave said:
You cannot measure the DC or AC voltages of a mic capsule using a conventional probe. You need something that is ultra-high impedance like 100G. Also any capacitance at all is going to throw off your measurements.

Just thinking out loud here but this is a back-of-napkin sketch of something that might work: ....

Just checking, but the sine wave generator you have in the schematic is representing sound waves varing the capsules voltage, right? (aka, you aren't using a frequency generator to apply an AC voltage?)

First off, I don't have any voltages higher than 48V, and my JFETs are Vgdo and Vgso of 50V...so I'll have to fiddle around with things and see what I can get.

But I figured out that as long as the capsule is properly shielded, I should be able to measure it with the right setup. I still want to measure it, but I'll be MIA for a few days. When I get back, I'll dig into it again.

But for now, I get a pretty clean sound. (Solved the muddy-sounding issue with a 1G-ohm resistor pulling the JFET gate to ground).
 
tynkerd said:
Just checking, but the sine wave generator you have in the schematic is representing sound waves varing the capsules voltage, right? (aka, you aren't using a frequency generator to apply an AC voltage?)
Yes. That drawing is from a circuit simulation program called LTSpice. So the AC voltage generator, capacitor, 1G and 48V DC is a minimalistic model of a LDC microphone.

tynkerd said:
First off, I don't have any voltages higher than 48V,
That's a show-stopper. You would need at least a few volts over 48 for this scheme to work (if it works at all).

tynkerd said:
and my JFETs are Vgdo and Vgso of 50V...so I'll have to fiddle around with things and see what I can get.
Actually if you're not concerned with killing a few JFETs experimenting, a lower voltage one might work ok because the JFET source resistor is going to limit the voltage drop across it. But you might have to be very careful about startup. Meaning it might survive if you turn on one supply or the other in a certain sequence. But if you just haphazardly turn on whatever power and start touching the JFET gate to things, it will probably die quick.

tynkerd said:
But for now, I get a pretty clean sound. (Solved the muddy-sounding issue with a 1G-ohm resistor pulling the JFET gate to ground).
It's probably heavily attenuated. The 48V is supplied through 1G. So by adding 1G to ground, you're making a voltage divider so the phantom supplied to the capsule is going to be only like 24V which is going to affect the output level and polar pattern. And the capacitance of stuff will attenuate everything further. Even a tiny capacitance of a few pF will attenuate. So even the JFET scheme I've proprosed is flawed (JFET gate capacitance can be pretty low but everything as a little capacitance). I'm not an expert in such things but my impression is that it's actually not really possible to measure a capsule directly. It's all just such high impedance that attaching anything to it will significantly change the behavior of it.
 
Back
Top