RC-filter for electret mic and USB sound card

Discussion in 'General Electronics Discussion' started by EvenAR, Aug 3, 2014.

1. EvenAR

3
0
Aug 3, 2014
I want to filter out frequencies lower than 500 Hz for an electret microphone, but as I'm quite new to electronic circuits I'm not sure if I'm doing things right.. The attached picture shows what I have tried so far, but without luck.

Using a mulitmeter I have found that the microphone I have has a resistance of 1400Ω. For testing I have used a capacitor of 220nF and a 1000Ω resistor - that should give a cutoff frequency of about 723 Hz.

The problem is; there is no sound at all! Without the filter the setup works fine.

Can anyone help me? What am I doing wrong?

Best regards,
Even

File size:
182.5 KB
Views:
1,345
2. KrisBlueNZSadly passed away in 2015

8,393
1,271
Nov 28, 2011
Hi Even and welcome to Electronics Point

An electret microphone needs DC bias. The circuit board must be providing this. When you connect the capacitor in series, you remove the DC bias and the microphone stops working.

I can't think of any simple way to do what you want.

3. Gryd3

4,098
875
Jun 25, 2014
You can always run captured audio through a filter with a program like Audacity.

Windows has a built in EQ as well that you can use to drastically reduce 500 and below... but this only works with 'playback' devices and won't work on the captured audio.

Kris... would an RL highpass circuit work instead of an RC highpass? There would be a DC offset across the inductor and microphone in parallel.

4. KrisBlueNZSadly passed away in 2015

8,393
1,271
Nov 28, 2011
I suppose you could connect an inductor across the microphone. You would need a DC blocking capacitor in series with it. This forms a series tuned circuit but with a large capacitor, the resonant frequency could be so low as to be irrelevant. But the inductor would need to be a pretty high value. Let's assume a circuit resistance of 1k and a -3 dB frequency of 723 Hz as before. So XL needs to be 1000Ω at f = 723 Hz.
XL = 2 pi f L
L = XL / 2 pi f
= 1000 / (2 × pi × 723)
= 0.22 H

So you would need a 220 mH inductor. Digikey have one that would probably be suitable: http://www.digikey.com/product-detail/en/70F221AF-RC/M8397-ND/774937

Let's calculate the resonant frequency if we put a 1000 µF capacitor in series with it.

fR = 1 / 2 pi sqrt (L C)
= 1 / (2 pi × sqrt (0.22 × 0.001))
= 11 Hz approx

So if you connect a 1000 µF capacitor and a 220 mH inductor in series, across the microphone, you will get the following behaviour if you play a frequency sweep with decreasing frequency.

At audio frequencies, the 1000 µF capacitor has a reactance of almost zero, i.e. it's practically a short circuit to the audio frequencies, so they only see the effect of the inductor, which will give a single pole high pass response with a corner frequency around 723 Hz. That means that signals above around 1 kHz will not be attenuated much, and a frequency of around 723 Hz will be attenuated by 3 dB.

As the frequency drops, the attenuation approaches a single pole high pass response of -6 dB per octave; that is, for every halving in frequency (one octave downwards is half the frequency), the attenuation increases by another 6 dB (well, 6.02 dB actually but 6 dB is close enough).

As the frequency approaches 11 Hz, the response of the combined circuit will be increasingly affected by the capacitor, and at 11 Hz the series circuit will hit resonance and will attenuate the signal to almost nothing. Below 11 Hz the response will pick up again as the reactance of the capacitor dominates the circuit's overall response.

EvenAR, I don't know whether this would be useful to you! You could try it and let us know how it goes.

Last edited: Aug 5, 2014
5. EvenAR

3
0
Aug 3, 2014
@KrisBlueNZ: That clears things up a little.. Now I tried to connect a 4.5V battery to the micorphone like the first cicruit on this page: http://www.epanorama.net/circuits/microphone_powering.html and that worked! The problem is where to get this power from. Using a power supply/battery is not very convenient. However, the USB port has a 5V output - maybe I can use that?

@Gryd3: The microphone will be used for VOIP (live). I have been looking for software based filters, but haven't found anything that outputs live audio yet...

Thanks for the quick response!

Last edited: Aug 4, 2014

5,165
1,087
Dec 18, 2013
A large inductance like that, probably not a good idea on the input of an audio amp that could pickup external magnetic or electric fields. This might interfere with the circuits operation and effect sound quality. Why can't he just do this? C1 and R2 form a high pass filter. With that said doesn't the microphone sort of have a LP effect anyway which is the capacitance of the element and the JFETs input impedance.

7. BobK

7,682
1,688
Jan 5, 2010
The circuit above is the way the mic is normally used, with R2 begin the input impedance of the amplifier it is feeding. Simply reducing C1 would get you a high pass effect.

Bob

8. KrisBlueNZSadly passed away in 2015

8,393
1,271
Nov 28, 2011
Yes, I agree. I should have made it clear that I wasn't recommending the inductor; I was just answering the question about feasibility. Actually I was hoping that the inductor value would come out so high that I could say it wasn't practical, but unfortunately Digikey did have a 220 mH inductor!
As I understood the question, he wants to modify an electret mic that plugs into a 3.5 mm audio jack. He doesn't have access to the DC supply. The supply is provided by the board with the jack socket, and is commoned with the point where the circuit gets the signal from. This is how electret mics are normally used.

You haven't shown where the circuit's input is supposed to connect on your diagram, but I assume it's after the capacitor. If he doesn't have an external DC supply, then he is back to the initial problem. You can't put a capacitor in series with a connection that carries the DC supply current to the mic.

The mic connector on the board is actually a TRS (tip-ring-sleeve) socket, aka a stereo socket, and (I assume) the mic is mono, so I guess it's possible that the board provides a separate DC power source on the other contact.

EvenAR, can you link to the technical documentation for that circuit board?
<shrug>

5,165
1,087
Dec 18, 2013
Ok cheers Kris

10. Gryd3

4,098
875
Jun 25, 2014
Is a high-pass filter for VoIP really necessary?
Most common narrow-band codecs available don't really go below the 300Hz mark.
Take a look at G.711 a-law or μ-law, as well as G.729. I could see this being an issue if you were using a wide-band codec, but is the 300-500Hz mark causing issues for your application?

11. EvenAR

3
0
Aug 3, 2014
It's actually going to be used on a multiplayer network where they use a very old voice codec for their voice servers. Low frequencies gets extremly muffled, so my goal is to remove most of the "unnecessary" low frequencies to get the voice as crisp as possible.

It's just a cheap sound card I bought on ebay. I don't think there is any technical documentation for it...