Mic to pedal

[strungout]

Hey, I got asked to build a mic pre-amp that would work well with pedals. I read that, contrary to guitar pedals, you need a low Z input and a high Z output. I,ve already built this (single supply version):

The mom switch doesn't work right, though... Either single press or hold-depressed doesn't cut the sound for more than a half-second (didn't time it, actually, but...). Connected to Vref or ground is the same. But that's another problem.

What I really want to know is if the circuit is good enough for what I'm trying to do: a basic mic preamp that works with guitar pedals. No fancy stuff, audiophile opamps or such.

I know the impedance is different. Guitars need a high Z input, low Z output. The mic requires the opposite, as I read. I'm looking to give the guy a box preamp to use his guitar pedals for vocals OR maybe build him a Stasis Leak(s) (from FC) adapted for a mic. I've already built one with stereo outputs.

What are my options? What do you guys think? Should I tell the guy to buy a vocal specific box?
He's not looking for auditory perfection. His mic has a 1/4 inch jack, so, not balanced. I've read some posts around here, about this, but still not quite sure where to go...

Any help appreciated!

[Mark Hammer]

It looks like you have sufficient gain in there to bring the mic up to levels comparable to guitar signal, which is good.  Mic preamps, however, often have balanced inputs, because applying lots of gain to a very low-level signal risks amplification of noise.  And noise in your basic signal risks amplification of noise further along the signal chain.  So maybe it's a good idea to have a balanced input.

[Fender3D]

You should connect BJT collector resistor to Vcc (18V) then connect the 1M resistor going to FET's gate to BJT collector. I don't know why you have diode to GND, it should be in series with the 1M resistor to gate... Check any Boss or Ibanez pedals schematic.
IC1b 22k from +in to Vref can be avoided, unless you have high offset voltage on op-amp output pin.
Search for mixer schematics, Soundcraft, FBT, Yamaha, etc., you choose; they have plenty of circuits, look for input section.
It's easier copying known working circuits....

As Mark says, better have balanced input, especially since you just need few resistors to get it

[anotherjim]

I'm not sure a mic (assume dynamic stage vocal type ) needs low impedance in order to work. The low impedance is more of a historical thing that's the standard and the need for low noise when there might be a lot of gain available in the preamp. However, a low impedance can reduce sensitivity to all kinds of noise being picked up (and it's intended for unbalanced you say).

Doesn't your FET need a gate blocking diode and pull up high when mute is off? (as Fender3d points out)

[strungout]

I should've linked the original schematic: https://www.circuitlib.com/images/projects/audio/mic_preamp_with_mute_schematic.JPG
My bad.

I connected the 150k Collector resistor to Vcc. And the 1M in series with the diode from Gate to Collector. Results are that I can 'gate' the signal' -- as opposed to shuting it off completely with teh mom switch (actually right now it's two wires I connect).

Right now I have the right function: the circuit mutes when the switch is depressed, but really, it's gating. Some signal does pass through. Really, I would've put a switch straight in the signal's path to shut it off with a momentary, but... that doesn't seem proper...

Here's what I have now:



Did I understand right?

[anotherjim]

Is your 1N34 a real germanium or one of the modern 1N34s silicon Schottky varieties? Germanium maybe not be good since can have reverse leakage? Plain silicon like 1N4148 is as good as anything else for the job.

The single series FET switch isn't a total mute. You can repeat it on the output to increase the cut or have one working on the opposite control state that shorts the opamp -input to Vref. So a series FET cuts then a parallel FET shunts whatever leaks past the series FET away.

[strungout]

oh, before I forget. I'm not using a balanced input because the guy's mic just has a 1/4" TS plug. He doesn'T do live, but jams with his brother and his girlfriend. XLR would've been better, yeah. I'm gonna ask him for more info on his mic. I asked him about an XLR cable but he wasn't sure if he had one for the mic. I've been testing the circuit with this old dynamic mic I have.

It was a genuine Ge diode. I just used it cause the circuit called for it. But, I changed it to a 4148. I have more of those.


So, the circuit works well now, with the addition of a second FET at the output. I was wondering some things, though:

- Do I need caps on each side of the FETs (D and S pins) going into the inverting input and coming from the output? I usually see caps there, on schematics, that use a flip flop circuit (which this is not; seems kinda like a tremolo circuit with a push button instead of an LFO), like in those Boss units and others.

- I got to wonder what the BJT was doing exactly? Follower? I wanted to see so, I totally removed it and associated components (left only the diode, 39nF and 1M from the Gate of the FET and connected the switch between the 1M and ground), and it takes a stronger signal to 'break' the mute, ie, pass some of the (gated) signal.

- The 39nF cap is there to slow down the switch to prevent sudden current draw that would result in a pop, right? A lower values would close it faster?

As for impedance, I'll have to see what kind of mic he has and test it. But, how do I measure it? I learned that Zin = Vin/Iin, but how do you measure it, exactly? Plug in my guitar, hit the strings and see what the readings are on my DMM?

[duck_arse]

the 39nF cap should be at the diode Kathode and 1M, not at the fet gate.

[Fender3D]

You don't need XLR socket, you might as well use one stereo jack socket, connecting tip and ring respectively to +input and -input, then the sleeve to GND. When inserting mono jack plug, the socket shorts ring to GND and becomes an unbalanced input

You'd rather use caps on FET's D and S pins AND bias resistors each pin. It may work without though, since op-amp bias is almost the same, but I'd keep it separate, just in case...

Long story short, BJT shorts collector to GND when you apply positive voltage at base...

Check mic datasheet or user manual for impedance.

[strungout]

Stephen: Done. Thanks. I think it helped the muting a bit. Lowered the threshold to turn the FET on some more.


Federico: Something like this?

(removed)

Took a stage from the Stasis Leak...

EDIT: hmmm can't make the full version bigger for some reason...

EDIT 2: New schematic:

[Fender3D]

Yes 

You can avoid 1uFcap and bias resistor @ IC1a input.
You might as well connect output FET in parallel with IC2b feedback resistor (it won't need bias resistors)

[strungout]

Federico: I tried the 2nd FET in parallel to the feedback resistor, it didn't mute as 'tightly' as after it.


Well, it worked. Now it doesn't. The FETs aren't fully openin their normal state, they gate. The signal gets to the first FET's Drain then gets gated. Though it was my breadboard acting up again (I think I keep bending the clamps...), but I built the circuit and after looking at everything, I get the same problem.... Noisy, gating, erh... I think I'm gonna try another circuit.

Just leaving a note, in case someone tries to work with this circuit.

[duck_arse]

I think your gates are getting confused when the switch isn't on mute. they will float on whatever the wind blows their way. the transistor switch in between the switch and gates was a better idea, as it told the gates off now as well as on now. at the appropriate moment/s.

[anotherjim]

Hmmm... Well, the JFET gates don't usually get an opposite control for the on state. Once the -Vgs from the control goes off, the channel naturally leaks into the low resistance state. What is odd in the scheme is the common gate control - there's a leaky path across the opamp that way. I think the FETs need their own gate-blocking diodes to close that coupled path and those diodes still should not be germanium!

[strungout]

Had some time to chill out...

Ok, working again. I had tried putting back the BJT section, but the extra diode did the trick. Don't worry, Jim, they are Si XD Just hadn't changed the text.

Here's what I have on BB and working:



It's actually pretty silent and clean. It plays ok with my Stasis Leak Reverb/Echo/Chorus. It doesn't distort, but it feels kinda weak. I'm gonna try upping the gain a bit. Put in a pot maybe. Other than that I think I can build and box it up

Thanks for the help guys!

[danfrank]

If the mic you're using has an unbalanced TS plug, take the resistance measurement of it with a multimeter. Chances are, it's a high impedance output mic, especially if it's an older mic with a TS plug. Many of these had impedance matching transformers built into the mic. In that case, up the input impedance of your circuit and that should make the audio out sound stronger.

[strungout]

danfrank: That totally slipped my mind (even though it's on my layout...): there's a 47k res in series with the 330r to ground, on a switch, for different impedance. I forgot to put it in on my BB. The signal level is a lot better. Thanks!

I tried mesuring the mic, but my DMM'S screen sorta flashes 0's for a few then shows a number (399r) and then flashes again and shows another number. It won't settle!
I know the impedance I'm using to test, though, I used both a 'Futura UD-119C' dual impedance (50k/600r). An old Japanese mic with a loose head... Couldn't find more info on it, and a Shure SV100 (600r).

Anyway, I'm happy.

[duck_arse]

if your transistor is wired as in yer last dia, it is pulled off by the 120k [?squint?] to ground, and then it's turned off when you mute the mute. it needs to be turned on at some stage - you could shift the resistor to ground to the left of the base series resistor, and sedt it to supply instead of ground. that way the transistor is pulled up and on until the mute pulls the 120k//120k junction to ground, and off's the base.

[strungout]

Stephen: sorry, I took a while to answer. When I connect the 120k to V+ (the collector resistor is supposed to be 150k, btw. The other two are 120k),  I get an opposite operation, grounding it turns the FETs on. I'm using a 3pdt momentary (all I have), so I could just change a wire to the opposite lug to the common. That would make it normally closed (unmuted) and pushing the button would open (mute) it.

One thing I noticed is the mute doesn't work with ANY other transistor but the same 2n2222a I used testing and building... Only that one. But with your suggestion any will work (well, those I've tried: bc549, 3904, other 2222a).

If I understand correctly: The BJT must be on for the FETs to be on, and off to shut down the FETs. The mom switch pulls the base to ground, through the 120k in series with the base, turning the BJT off. So, my problem is, the BJT is not in the right state when the circuit is turned on: it's off. It needs to be on. That's with a 120k to V+.
By my schematic it's on, and just won't turn off (with any other transistors than THE one)... err...

[duck_arse]

you have a choice - transistor is wired to ground/off, and mute pulls up/turns transistor on - or - transistor is wired to V+/on, and mute pulls base to ground/off. it only matters which way you wire the resistor. [I'm simply copying one of the transistors from a boss flip-flop, really.]