Circuit to connect PC Headset to USB Audiointerface

Special Interest Groups Recording Equipment
1

Hello forum

I recently started recording voice (spoken only so far) on my PC and soon found out that internal/external sound cards are a waste of money, since almost all of them have low quality, noisy microphone preamps. I also wanted a portable solution, that would be usable with decent results in all realtively quiet environments, so I did some digging.
To make a -really- long story short: I found a connector circuit that was outlined in a blog post here, and since I wanted to verify what was posted there, I also found this post by steve in this forum. Since both sort of matched it gave me enough confidence that the layout would not fry my newly bought hardware, so I settled for a PC Headset (Beyerdynamic MMX2) and an USB Audiointerface (Focusrite Scarlett 2i2).

I built the circuit, and it worked like a charm. So thanks to steve and the whole forum for that!

Because I had some additional parts, I used a 10kΩ trimmer to replace the 2,2 kΩ resistor outlined in the original circuitry. Since I’m sort of an adventurous type I fiddled a little with the trimmer (I just increased the value, I’m not that adventurous :smiley:), and the increased value gave me a much higher output signal from the Focusrite.
Since it improved my input signal, which is always a good thing to have, I would like to keep the increased value. However because I only have some very basic electronics knowledge, I’m not sure if keeping the value would damage my headset, or even the Focusrite, so I would like a professional opinion on that :slight_smile:

Another thing I would like an opinion on is: The Focusrite also gave me an improved signal when I set the input channel the circuit was connected to to Hi-Z. Please mind that that was without altering the resistor value! Can I keep that setting, too, or will it be harmful to the Focusrite? I’m worried Hi-Z may not be supposed to be connected to the kind of output the custom built circuit provides?

Any help will be greatly appreciated!


Cheers

Line40

2

The 2.2K resistor is appropriate for the original 5 volt supply in the computer. Since you’re using 9 volts, it can be higher.

Variable resistors can have their own problems. Because they are physical and direct part of the sound circuit they can cause more noise than a good quality fixed value resistor. So if you get more frying than you think should be there, that may be a good place to look. If you’re using a wound multi-turn pot, that can cause frequency response issues. They have different values as the pitch of the sound goes up.

My design put an additional 33K or higher resistor on the (-) side of the capacitor to shield. It’s not low enough to cause noise problems and it prevents the explosive pop when you plug this into a mixer or the Focusrite.

There are capacitor issues, too. Electrolytic capacitors use the applied battery to make the capacitance. So putting a 10u 100v capacitor into a 5v circuit will not give you 10u. It will be seriously lower and can give you low pitch sound problems. This is why you can’t drift too far away from the listed voltage. 6v, 10v and 16v capacitors should all work OK in a 5v circuit. 16v is pushing it.

You should also be careful because electret microphone elements do have overload points. The point where it’s so loud the two elements smack together. Older condenser microphones could be destroyed like that, but I think the newer ones survive.

Another quick note since you are messing with mic-level sound. The reason for three-pin XLR type microphones is so you can have hundred-foot long microphone cables. Since you are using unbalanced, home-made adapters and cables, you’re probably going to be limited to six or eight feet before the electrical noise in the room starts to get to you. This may not bother you a bit.

Koz

3

an improved signal when I set the input channel the circuit was connected to to Hi-Z.

That’s because you have a Hi-Z microphone.

Rock Band and broadcast Lo-Z microphones like the Shure SM58 have a 150 ohm value, not 2200, 5000 or up, and they’re expecting to see an amplifier of 1500. That’s why your sound level is abnormally low with it set like that.

This is another reason you can’t have a long cable. The SM58 has a powerful, firm signal to survive a long trip. Your microphone has all the robust electrical force of a butterfly.

Koz

4

The 2.2K resistor is appropriate for the original 5 volt supply in the computer. Since you’re using 9 volts, it can be higher.

Hmm, I do not quite understand that, I thought that increasing the value on the resistor would decrease voltage for the mic capsule goes? I would imagine the output level going lower when I increase the resistor then. But I really have no idea :slight_smile:

Variable resistors can have their own problems. Because they are physical and direct part of the sound circuit they can cause more noise than a good quality fixed value resistor. So if you get more frying than you think should be there, that may be a good place to look. If you’re using a wound multi-turn pot, that can cause frequency response issues. They have different values as the pitch of the sound goes up.

Ok, the variable resistor was just for testing anyway, I’ll remove it.

My design put an additional 33K or higher resistor on the (-) side of the capacitor to shield. It’s not low enough to cause noise problems and it prevents the explosive pop when you plug this into a mixer or the Focusrite.

Is the 33k resistor in line with the cable going from the resistor (-) to the Focusrite, or parallel connecting both lines on the Focusrite end of the circuitry, like the 10k that was posted in the design in this forum? I’d like to add the resistor, since I already noticed the Focusrite gain control flashing bright red when I plug the circuit in.

There are capacitor issues, too. Electrolytic capacitors use the applied battery to make the capacitance. So putting a 10u 100v capacitor into a 5v circuit will not give you 10u. It will be seriously lower and can give you low pitch sound problems. This is why you can’t drift too far away from the listed voltage. 6v, 10v and 16v capacitors should all work OK in a 5v circuit. 16v is pushing it.

Thanks for the advice!, Now I’ll just have to find a retailer here in germany that has these kinds of capacitors available, all I could find with such a low voltage so far were SMD ones :slight_smile: I’m currently using a 50V part, since that was the lowest voltage my local electronics parts dealer had available.

You should also be careful because electret microphone elements do have overload points. The point where it’s so loud the two elements smack together. Older condenser microphones could be destroyed like that, but I think the newer ones survive.

When would that happen? If the voltage level on the microphone capsule gets too high, or if I talk to loud? Sorry for such dumb questions but I really have’nt got not much of an idea what causes what in that circuit, or of the condenser microphone design. I did look it up in Wikipedia, but from reading there I thought the moving part of the capsule would not swing “more” if the input voltage is increased.

Thanks again for the quick help :slight_smile:


Cheers

Line40

5

@Line40, not all of the sound cards are crap. Some of them are actually worth the money. The reason micro-phone recording is not that well in most sound cards is that while designing these cards, the main consideration is not the audio input quality, it’s the audio output quality and gaining maximum possible amplification from least power supply. You can also find the sound cards that radio DJs use in their systems or are used for professional recoding, they are sheer class and the voice recording clarity is stunning.

6

Hmm what sound cards are you referring to? At the time I wrote the above posts, I had done some research on sound cards and did not find any that were mentioned as being used by Radio DJs for voice recording. From my research I was under the impression they would use USB Audio Interfaces or USB Mixers.
I’m currently using a Beyerdynamic MMX 2 connected to a Focusrite Scarlett 2i2 trough the above mentioned circuit, and I’m quite happy with the setup, however if there was a soundcard that delivered the same quality but would help me get rid of the circuit I’d be happy to try it out.

While experimenting with the circuitry I did at one point connect it with input switch of the FocusRite set to “Inst”, and the gain was much better than with the Microphone however I had a strange crackling noise while recording, as if someone would pour sand on the microphone while I’m speaking, so I switched back to “Line”. Has anyone got any idea what could cause that crackling?

Cheers

Line40

7

True, but the ones that are generally included in PCs as standard are usually pretty bad for recording.
I’ve used a RME HDSP 9632 sound card (not tried it with Audacity) and that sounds terrific. At over $500 it should sound good. :wink:

8

The resistor and the mic capsule work together as “potential divider” (a “voltage” divider).

If this is two resistors ( or “resistive” components) in series:

(A)/////(B)_____/////_(C)

and if the voltage from A to C is 9 volts,
the the voltage from A to B, plus the voltage from B to C, equals 9 volts.

Increasing the resistance across A ↔ B will cause a greater voltage drop across A<->B, so the voltage across B<->C will become less.

In the microphone circuit there are two conflicting things going on.

  1. The bias voltage. This is the DC voltage across the microphone capsule. Generally, a higher voltage will provide a higher signal (and importantly a higher signal capability before clipping), but if too high the capsule is destroyed. Reducing the “bias resistor” value causes a greater voltage across the capsule, Higher bias resistor values reduce the voltage across the capsule.

  2. The signal voltage. This is the (AC) signal generated by the microphone capsule.
    In effect we have a second “potential divider” circuit, where A<->B is the bias resistor and B<->C is the “load resistance” of the “Audio out”. The mic capsule inserts the signal voltage at “B”. Reducing the value of the circuit resistor allows more of the signal current to “leak” to ground (via the battery), so the voltage at the output is reduced. Increasing the circuit resistor blocks that path allowing a higher signal voltage at the output.

Ideally the circuit resistor should be ‘tuned’ low enough to provide a sufficiently high DC voltage for the mic capsule, but still high enough to block the (AC) signal path.

9

Plus a high value static drain resistor across the output lines 33K to 100K. The value isn’t as important as to have something there.

Over time, an unconnected adapter will leak enough charge across the capacitor plates (electrolytic capacitors aren’t perfect) to deliver a very healthy pop or crack in the sound circuit when connected. This resistor prevents that.

Koz