I'm sorry this is so much to read. I hope this topic isn't too redundant. I did try searching.
I'm trying to build a very simple 3 in, 1 out mic mixer. The design is based on Forrest Mims' drawing in "Basic Semiconductor Circuits". I made a pdf of the drawing. http://www.imconfused.net/3MicMixer.pdf
The idea is to put 3 microphones via XLR into this, and have it go out through an XLR jack to my USB Line6 Toneport. As I understand it, it uses a 2n2222 as a simple adding amplifier. The 3 mics come in on XLR through some 10K pots for simple volume adjustment.
Now, the circuit works, in the sense that with 1, 2, or 3 mics plugged in, you can hear sound from the mics on the output. However, there is a hideous amount of noise. I would estimate that at the loudest you can get the mic, the signal to noise is 1:2. The noise is part hum and part crackle. The crackle goes away if I short the X to the R on the output jack; I don't seem to be able to eliminate the hum at all.
I expect that this is a ground loop problem, but I've no idea where to start. Here are some other details I think might be relevant:
IN
X directly to X out
L to circuit (input capacitor)
R to circuit low ("ground")
OUT
X directly from X in
L from circuit (output capacitor)
R from circuit low ("ground")
To eliminate part of the noise, I shorted X to R on the output jack. Not a whole lot better, but definitely better. I tried modeling the circuit in spice, and all the voltages look about like they should, but those were under very artificial conditions (putting a function generator in and reading the output on a scope).
I'm baffled. Thanks to anyone who read all this.
I'm trying to build a very simple 3 in, 1 out mic mixer. The design is based on Forrest Mims' drawing in "Basic Semiconductor Circuits". I made a pdf of the drawing. http://www.imconfused.net/3MicMixer.pdf
The idea is to put 3 microphones via XLR into this, and have it go out through an XLR jack to my USB Line6 Toneport. As I understand it, it uses a 2n2222 as a simple adding amplifier. The 3 mics come in on XLR through some 10K pots for simple volume adjustment.
Now, the circuit works, in the sense that with 1, 2, or 3 mics plugged in, you can hear sound from the mics on the output. However, there is a hideous amount of noise. I would estimate that at the loudest you can get the mic, the signal to noise is 1:2. The noise is part hum and part crackle. The crackle goes away if I short the X to the R on the output jack; I don't seem to be able to eliminate the hum at all.
I expect that this is a ground loop problem, but I've no idea where to start. Here are some other details I think might be relevant:
- There is no DC offset (0.000V on my meter) between R and L on the output. I confirmed this with a pretty nice Tektronix scope.
- There is a small (2.5mV) dc offset between X and R when they aren't shorted
- All signal wires longer than an inch are shielded. Doing this didn't reduce the noise perceptibly. As such, I did not bother to shield the ground wires.
- I am using a wall-wart dc power supply, which says 9V but is actually 13 or so. I put the low side to the circuit "ground" and the high side to the collector of the 2n2222.
- The hum doesn't seem to go away by shorting any two things together, except of course shorting R to L makes it completely silent.
IN
X directly to X out
L to circuit (input capacitor)
R to circuit low ("ground")
OUT
X directly from X in
L from circuit (output capacitor)
R from circuit low ("ground")
To eliminate part of the noise, I shorted X to R on the output jack. Not a whole lot better, but definitely better. I tried modeling the circuit in spice, and all the voltages look about like they should, but those were under very artificial conditions (putting a function generator in and reading the output on a scope).
I'm baffled. Thanks to anyone who read all this.
Hum can come from perhaps three sources. If you hear it with the controls at minimum, that eliminates the input system. It leaves power supply (if you are going into USB why not use the power you get from USB port?) and EMI, which is a long shot.
You also need to filter the 9V. Wall warts are pretty ragged. Try listening to the hum as you unplug the wall wart; if the hum disappears instantly, that's the problem. If it holds and fades down, it's something else.
You also need to filter the 9V. Wall warts are pretty ragged. Try listening to the hum as you unplug the wall wart; if the hum disappears instantly, that's the problem. If it holds and fades down, it's something else.
Yes, it definitely goes away instantly upon unplugging the wall wart. As far as filtering the power, do you suggest a linear regulator and a couple caps? I think I have some LM7809s sitting around here somewhere. I'm not using the usb because I was going to place my mixer is quite far away from the computer (in a different room, actually) and I don't have a usb cable that long. That is an interesting idea, though.
Hi
It seems as if your are choking your input signal.
I had a look at the schematic, you use a 1k 15k base divider network wich yields a transistor input resistance of appr. 900 ohms.
Also you got a 10k series resistor which acts as a divider with the 900 ohms input resistance.So there isn't much left to amplify.
My suggestion: Rise the input impedance of the amp by vastly increasing the base divider resistors.Or even better amplify each input seperatly before mixing.
Regards
Arne
It seems as if your are choking your input signal.
I had a look at the schematic, you use a 1k 15k base divider network wich yields a transistor input resistance of appr. 900 ohms.
Also you got a 10k series resistor which acts as a divider with the 900 ohms input resistance.So there isn't much left to amplify.
My suggestion: Rise the input impedance of the amp by vastly increasing the base divider resistors.Or even better amplify each input seperatly before mixing.
Regards
Arne
That's an interesting observation, yeti. I'll take it into account if I have problems going forward. As it is, cleaning the input power like bob91343 suggested (with two 470 uF caps and a 7809) cut out all the noise. I have yet to really use the mixer very much, but if I need to boost the performance I'll check out what you say.
Still don't understand why I need to connect the X and the R together to eliminate the ground loop, though...
Still don't understand why I need to connect the X and the R together to eliminate the ground loop, though...
This is not a good circuit, for a number of reasons.
The voltage on the base is barely sufficient to bias the transistor into conduction under any conditions.
The DC conditions are not fixed but vary with the pot settings.
The amplifier input impedance is low.
Use an inverting opamp summer.
w
The voltage on the base is barely sufficient to bias the transistor into conduction under any conditions.
The DC conditions are not fixed but vary with the pot settings.
The amplifier input impedance is low.
Use an inverting opamp summer.
w
- Status
- Not open for further replies.