Designing a pedal for harmonica

[t1redhands]

I asked this question on Reddit as well, but wanted to get your opinions:

I've been asked to design a noise suppression pedal for a harmonica player. They use a low impedance microphone and a Fender Blues Jr. type amplifier.

So far, I've been planning on using a Fortin Zuul-type noise gate using a THAT VCA chip, a Bear Hug-type JFET compressor, and an op-amp splitter to drive the clean signal to the gate's sidechain.

What should I use for a mic preamp? Should it be a transformer, or something like this Rolls personal mic preamp modified by Mark Hammer: https://www.diystompboxes.com/smfforum/index.php?topic=91046.msg775636#msg775636

What are your expert opinions?

[amptramp]

Are you sure a noise gate is what you want?  I would prefer a sound-operated filter that reduces the treble linearly when there is no audio present so the cutoff isn't as abrupt.  This is nothing new - RCA used the Magic Monitor circuit below in the 1940's:



In this circuit, the triode is a voltage amplifier that uses a shunt diode to generate an average negative voltage to drive a pentode variable reactance stage.  The reactance stage consists of a 47 pF capacitor connected between the signal and the plate of the reactance tube.  When the signal is high, the negative voltage biases the pentode into low gain and there is little amplification of the signal coupled into the output.  When the signal is low, the reduced bias causes the pentode gain to be high and the plate end of the 47 pF capacitor (pF is labelled MMF here) is driven with a voltage opposite the voltage coupled into the grid by the 1 nF (or 1000 MMF) capacitor and the effect is to make the 47 pF capacitor look larger.  This increased capacitance reacts with the 1 megohm resistor (shunted by the 330 pF capacitor) to reduce the rolloff of the stage.

You could get the same effect with a photoresistive cell in series with the audio output with a capacitor to ground.  When the signal is high, the photocell is illuminated and at low resistance.  When the signal is low, the photocell goes to high resistance and the R-C lowpass frequency drops down.

With both of these schemes, you can add noise gate function if you need it when the rolloff is low enough that the cutoff is unobtrusive.  The Magic Monitor could be redesigned for solid state and would probably be a nice compact (1590A) stompbox, if not a permanently 'on' device.

[t1redhands]

Whoa, neat! I know - I'm not sure if a basic noise gate is the way to go. It seems like there are a few ways of approaching feedback suppression for harmonica players and I decided to start with the easy route of using an available layout of a fancy gate.
I'll have to check this out - I haven't attempted to convert tubes to FETs in an application like this, so it's a learning moment!

[amptramp]

For the mic preamp, we would need to know which mic you are using and its output characteristics before we can advise you of what to use.

I have a nice Conneaut Audio Devices mic with a built-in preamp, so it wouldn't need anything but phantom power.  It actually contains two electret condenser microphones and allows you to switch between omni, cardioid and lemniscate (figure angular sensitivity patterns, and flat or 80 Hz highpass response and since it can easily overload a mixer board, it has a -20 db/full output switch.  Internally it has OPA2107 op amps feeding an MC33178 output stage with TL062 integrator stages providing bias correction from the first to second stage.  If you have something like that, all you need is power (or batteries and it came with a pair of NiCd 9 volt batteries that can be charged from phantom power).

Not all microphones are alike.

[t1redhands]

Good point!

The harpist uses an Electro-Voice 635A https://products.electrovoice.com/la/pt/635a/

Iain recommended a differential amplifier on my Reddit post. I'm going to test a high-impedance circuit based on this guy: https://www.allaboutcircuits.com/textbook/semiconductors/chpt-8/building-a-differential-amplifier/

Thanks!

[PRR]

Noise-gate is not a primary way to tame acoustic feedback. Yes if you start from silence it won't open and won't howl. But you gotta play sometime. Once it open, it howls, and the howl will keep the gate open.

FWIW, a EV 635A (I know it well) pressed close to a harmonica needs minimal preamp. While the mike is full-balanced, you can just ground one pin and take the other like a weak guitar. Give a gain of 10 to 50, that should be really ample.

[amptramp]

The mic you have selected is a dynamic mic with an output impedance of 150 ohms.  This is relatively low impedance but low voltage output, so an amplifier with a reasonable gain should be used and Paul's suggestion of 10 to 50 should be OK depending on the mixer board you are using.  You do not need the high impedance of an instrumentation amp and circuitry based on 5532 op amps should be OK because they are suited to low impedance.

A dynamic mic is like talking into a speaker and taking the output from the voice coil - low voltage and low impedance.

[t1redhands]

Noise-gate is not a primary way to tame acoustic feedback. Yes if you start from silence it won't open and won't howl. But you gotta play sometime. Once it open, it howls, and the howl will keep the gate open.

FWIW, a EV 635A (I know it well) pressed close to a harmonica needs minimal preamp. While the mike is full-balanced, you can just ground one pin and take the other like a weak guitar. Give a gain of 10 to 50, that should be really ample.

The mic you have selected is a dynamic mic with an output impedance of 150 ohms.  This is relatively low impedance but low voltage output, so an amplifier with a reasonable gain should be used and Paul's suggestion of 10 to 50 should be OK depending on the mixer board you are using.  You do not need the high impedance of an instrumentation amp and circuitry based on 5532 op amps should be OK because they are suited to low impedance.

A dynamic mic is like talking into a speaker and taking the output from the voice coil - low voltage and low impedance.

This is the expert advice I came here for! Marvelous stuff, now that I've built the Zuul with the pricey THAT vca I'll have to start over, haha!

I've taken a crack at the Magic Monitor circuit but I'm afraid it's above my reading level at the moment. Amptramp - looks like you've been promoting the circuit on this forum for a little bit. Any further advice as to how to convert the reactance pentode to FETs? I might like to try using LND150 depletion mode MOSFETs for this.

Thanks again!