Mic preamp - questions and problems

[ukman]

Hello All.

I am trying to construct a simple preamp for a dynamic mic (to allow me to connect to the mic-in port of a PC) and have run into some noise problems. I ought to point out that I'm, at best, an electronics amateur, and generally clueless at circuit design. I'm also working somewhat blind as I'm relying on some PC based scope software that gets its data from the sound card. I'm not sure how reliable it is.

Here are a few facts:

1. The mic I'm working with is a Behringer XM8500 with an impedance of 250R. It produces a signal of about 10 mV. It has a differential o/p. (I think audio guys call this "balanced").

2. I have a very simple design: I'm using (half of) a NE5534 as a difference amp (i.e one of these), with a voltage gain of about 50, 35dB. This gives an acceptable signal to record at the PC end.

3. The CMRR of a difference amp is apparently pretty sensitive to the ratios of R3/R1 and R4/R2 so I've selected these carefully so that they're better than 1%. R1 = 3351R, R2 = 3347R and R3 = R4 = 178K3

4. The h.f. 3dB point is about 18KHz. (For l.f. rolloff see below ..)

5. The circuit is battery-powered (two 9v batteries) so I think I don't have to worry about ground loops (I think ..)

The problems:

P1. The dynamic mic seems to be astoundingly susceptible to 50Hz hum. I was hoping that it would all be common mode noise, and the difference amp would eliminate almost all of it, but I'm wrong: with the mic connected to the amp, I see a minimum of 5mV p-p hum. Given that my o/p signal is about 500-600 mV typical, this is a poor SNR.

Can anyone suggest why I'm getting so much hum in the amplified signal.

P2. The self-noise of the amp. (with no mic attached) seems to be about 1 mV p-p. This seems to be far higher than I'd expect from the spec of a 5532. This is broad-band noise, with the same power density all the way up to 20KHz or so. This manifiests itself as background hiss when I record.

What could be causing this noise ? Interference due to poor shielding ? I can't believe that it's all generated in the circuit itself. The noise from a 5532 + any Johnson noise wouldn't seem to be enough, given such a simple circuit.

P3. I am currently coupling the 5532 directly to the mic o/p. I have tried adding a couple of caps to give low freq. rolloff (3dB at 100Hz) to reduce the 50Hz hum, but this had no noticeable effect. This confuses me - surely if I roll of the gain from 100Hz I should substantially reduce 50Hz noise, no ?

P4. In addition to this, I also see 120 Hz hum - now I'm in the UK, not the States, so this can't be a mains harmonic. Cany anyone suggest where I'd be picking up 120Hz hum from in a domestic environment ?

And more generally - I see mic preamps advertised with say 70-80dB gain, but very good noise figures - however do professional designers achieve these ? Does a lot of this come down to good shielding ?

 

[Nigel Goodwin]

There are hundreds of suitable mike preamp designs out on the net - why try and start from scratch?.

The hum will be 50Hz mains hum, and it's because you're either not connecting it correctly, or you're not screening it correctly - and possibly because you've not got a sensible input impedance?.

Here's one such example:

**broken link removed**

Otherwise, post your circuit so we can see what you're doing.

 

[audioguru]

Your preamp has a gain of 50 and its output feeds a mic input?? Then the total gain is way too high. The preamp's output is line-level, not mic level.

A coupling capacitor as a highpass filter has a very gradual roll-off. If it is -3dB at 100Hz then its output at 50Hz is -6dB that sounds only slightly less than the level of the 100Hz. You need an 8th-order active filter with a cutoff at 100Hz to reduce 50Hz to -48dB which might still be audible.

I think your mic does not have shielded audio cable or your circuit is built on a breadboard that picks up all kinds of interference.

 

[Nigel Goodwin]

Your preamp has a gain of 50 and its output feeds a mic input?? Then the total gain is way too high. The preamp's output is line-level, not mic level.

I missed that - he should be connecting to the line input, not the mike input.

 

[ukman]

Your preamp has a gain of 50 and its output feeds a mic input?? Then the total gain is way too high. The preamp's output is line-level, not mic level.

I'm reasonably confident the gain is about right. this Dell Inspiron audio spec specifies 280 mV p-p for mic-in, and in practice, I've found I need 300 mV + to start clipping when recording. This may not be a standard mic-in level, but it seems to be what the soundcard is looking for.

A coupling capacitor as a highpass filter has a very gradual roll-off. If it is -3dB at 100Hz then its output at 50Hz is -6dB that sounds only slightly less than the level of the 100Hz.

Right. But 6dB down is 50% of the signal - that is both easy to hear and to measure. I don't appear to be seeing that kind of attenuation. It's baffling me.

I think your mic does not have shielded audio cable or your circuit is built on a breadboard that picks up all kinds of interference.

Well, there are two signal degradations, 50Hz hum and the rest of the higher frequency noise. The 50 Hz hum comes from the mic - if I disconnect the mic from the cable but leave the cable attached to the preamp, there is negligible 50 Hz hum at the mic o/p(though plenty of the noise - about 1 mV p-p). If I connect the mic to the cable, then there is about 5 mV p-p hum with noise on top. If I move the mic, I can reduce or increase the hum reliably. I'm pretty certain the mic itself (i.e. the mic coil) is picking up the hum.

If I plug the mic directly into the mic in port, and look at the signal with the PC scope (I have nothing better, I'm afraid), I see a few hundred uV of balanced 50 Hz hum i.e. I'm getting a nice common mode hum signal that a difference amp should largely eliminate, but it appears not to do so.

The circuit is currently on veroboard, and probably not adequately shielded, but the circuit itself is not picking up most of the hum.

 

[ukman]

There are hundreds of suitable mike preamp designs out on the net - why try and start from scratch?.

Well, I'm implementing a difference amp - that's hardly starting from scratch. It's merely starting from something simple. This makes more sense for me as a) I have all the components to hand to implement it b) I understand what each component does c) most importantly, I know what component values to change if I want to change the gain or roll-off frequencies. If I use someone else's circuit all of that is more difficult.

The hum will be 50Hz mains hum, and it's because you're either not connecting it correctly, or you're not screening it correctly

Oh, it's certainly 50Hz hum. And it's coming from the mic, most of it. What I'm trying to figure out is why the amp isn't eliminating more of it.

- and possibly because you've not got a sensible input impedance?.

Well, the minimum input impedance is about 3K3 to the inverting input, which should be fine for a 250R mic, I think ? I guess the input impedance to the non-inverting impedance is higher though, due to the 178K resistor to ground. Hmm. Does that break things, I wonder ? Maybe thats screwing up the theoretical CMRR ?

Otherwise, post your circuit so we can see what you're doing.

How do I do that ?

 

[Nigel Goodwin]

Well, I'm implementing a difference amp - that's hardly starting from scratch. It's merely starting from something simple. This makes more sense for me as a) I have all the components to hand to implement it b) I understand what each component does c) most importantly, I know what component values to change if I want to change the gain or roll-off frequencies. If I use someone else's circuit all of that is more difficult.



Oh, it's certainly 50Hz hum. And it's coming from the mic, most of it. What I'm trying to figure out is why the amp isn't eliminating more of it.



Well, the minimum input impedance is about 3K3 to the inverting input, which should be fine for a 250R mic, I think ? I guess the input impedance to the non-inverting impedance is higher though, due to the 178K resistor to ground. Hmm. Does that break things, I wonder ? Maybe thats screwing up the theoretical CMRR ?



How do I do that ?

Click on 'Go Advanced' and then 'Manage Attachments', then upload your graphic file (either GIF or PNG for a computer drawn diagram).

 

[audioguru]

The input resistance of your inverting input is 3.3k and for the non-inverting input is much higher at 181.3k so the higher resistance side picks up much more radiated interference than the low resistance one.
Balanced mic preamps use a transformer or 3 opamps in an instrumentation amplifier IC for near-perfect balance. The transformer provides voltage gain for lower noise.
Do you have shielded mic cable with two wires inside a shield?

 

[ukman]

The input resistance of your inverting input is 3.3k and for the non-inverting input is much higher at 181.3k so the higher resistance side picks up much more radiated interference than the low resistance one.

Yes, I need to look at this - having thought about this, I guess the unequal input impedances present a non-balanced i/p to the op amp.

Balanced mic preamps use a transformer or 3 opamps in an instrumentation amplifier IC for near-perfect balance. The transformer provides voltage gain for lower noise.

I don't want to have to get a transformer if at all possible. Is it possible to fix this by buffering the op amp inputs with emitter followers ?

Do you have shielded mic cable with two wires inside a shield?

Yes. It's a shielded XLR to 3.5 mm jack cable.

 

[Nigel Goodwin]

Yes, I need to look at this - having thought about this, I guess the unequal input impedances present a non-balanced i/p to the op amp.

The whole point of the balanced input is to reject common-mode signals, if you start off with the opamp not balanced it obviously makes it pretty pointless.

I don't want to have to get a transformer if at all possible. Is it possible to fix this by buffering the op amp inputs with emitter followers ?

Did you check the working circuit I posted a link to? - no need for either a transformer, or emitter followers.

However, from what you say about how bad the hum is, it sounds like a much more basic error.

Post your circuit, and post a picture of how it's built.

 

[audioguru]

Maybe it is another case of a breadboard causing interference pickup.

 

[ukman]

The whole point of the balanced input is to reject common-mode signals, if you start off with the opamp not balanced it obviously makes it pretty pointless.

Right. Which is why I suspect the problem is related to the different input impedances seen by either side of the mic.

Did you check the working circuit I posted a link to? - no need for either a transformer, or emitter followers.

Yes. If I can't get any significant improvement with my circuit, I'll probably put that together. I don't understand the circuit operation though - it seems to have a complex feedback arrangement and I can't tell what controls what. I prefer to work with something I understand.

However, from what you say about how bad the hum is, it sounds like a much more basic error.

Possible but I think the evidence points away from this. The circuit suffers no hum until the mic is attached to the cable, and the hum can be changed by moving the mike location => it's the mic that picks up the hum, and the circuit fails to reject it. Anyway, I have to put this on hold for a few days but I'll let you know if I make further progress.

BTW, one of my posts is missing. It seems to have disappeared into a moderation queue or something.

Post your circuit, and post a picture of how it's built.

I'll do this in a few days - I have to put this aside for now.

 

[Nigel Goodwin]

Right. Which is why I suspect the problem is related to the different input impedances seen by either side of the mic.



Yes. If I can't get any significant improvement with my circuit, I'll probably put that together. I don't understand the circuit operation though - it seems to have a complex feedback arrangement and I can't tell what controls what. I prefer to work with something I understand.



Possible but I think the evidence points away from this. The circuit suffers no hum until the mic is attached to the cable, and the hum can be changed by moving the mike location => it's the mic that picks up the hum, and the circuit fails to reject it. Anyway, I have to put this on hold for a few days but I'll let you know if I make further progress.

Sounds like you don't have the screen of the XLR connected?.

BTW, one of my posts is missing. It seems to have disappeared into a moderation queue or something.

Sorry, missed that one - done now.

BTW, I've got a Behringer XM8500 as well, it's a decent mike - however, I've also got six XM1800's (which are 1/3rd of the price), and to be honest I can't tell any difference.