MXR Noise Gate (M-106) Corrected Schematic

[Rob Strand]

This one has always had some issues because the original schematic was wrong and there was a typo on the Zener.

I checked the schematic against PCB pics on this site:
http://www.effectsdatabase.com/model/mxr/m106#

and found some more differences!.

I encourage you to count the number of 1uF and 10uF tantalums, and to count the number of 10nF and 50nF ceramics.

Note there is *no* 1uF NP cap on the PCB.  This cap is actually 10nF, it is located on the right hand side where there is a 10nF under a 50nF.

Updated Schematic: [Obsoleted see below]


Summary of Errors and Fixes for MXR noisegate schematic:
(Robin Tomtlund's schematic circa pre 2003)

1) I believe the zener should be a 1N5231 (Vz = 5.1V, Pz = 0.5W) not
1N5331.  The zener is running ar very low currents 180uA so it's
probably only going to produce 4.7V or less.

From the MXR phase 90 the 1N5231 produces 4.82V at 360uA.
500mW Zener will be about 4.5V.
1W Zener will be about 4V

2) Error FET T3 gate and drain reversed.

3) The sensitivity pot should be 500k not 600k.
    Probably log (A) taper.

4) [2019] output cap 10uF tant

5) [2019] Cap on base of T2 in parallel with 100k is 10nF;  not 1uF

6) [2019] 1N914 diode across supply rails.

[Rob Strand]

I found another error.  Another clearer pic I have shows quite clearly the
10uF tant to the JFET drain has the cap's + terminal going to the drain.

EDIT:  FWIW,  the + to drain actually makes sense because the the drain is at Vref whereas the emitter is at Vref-Vbe ie. lower than Vref.

[Rob Strand]

Here's the 10uf "cap-flip" fix,

7) Flipped 10uF to JFET drain.

[Obsoleted see below]

[Rob Strand]

Well, well well, poking around at more pics I found yet another bug.

8 ) The resistor on the counter-clockwise side of the pot is 10k not 1k.
     What I'm not sure about is if 1k was ever used.
     I have two pics clearly showing 10k and one not so clear one where I can't tell if it's 1k or 10k.

Here's the update with the XLR connector added: [Has error]




Connections and colors
Numbering from left to right component side with chamfers at top.

1   Red      +9V
2   Brown      Pot 1 (ccw)
3   Yellow      Input
4   Blue      Pot 3 (cw)
5   Black      0V
6   Orange      Switch
7   Grey      Pot 2 (wiper)      
8   Violet      Output

[Rob Strand]

Just for the non-believers, I found this post from back in 2010 on diystompboxes!

JUANSNEIDER
Re: MXR Noise Gate 150K (Ra)
Reply #7 on: September 15, 2010, 12:35:15 AM

All schematics on the web concerning the MXR Noisegate are wrong including the one from tonepad.
Referring to version 2 from 2003 (which I have on my HD):

- Input cap is in reality 10n (0.03) input resistor 10k.
- The 10µ electrolytic between the 22k and the FET is drawn in the wrong direction -
  positive should be on the Fet-side.
- Output cap is 10µ instead of 1µ.
- the capacitor before Q2 in parallel to the 100k is a 10n (0.03) instead of a 1µ.

The only thing I disagree with is the Threshold pot has the 10k instead of 1k, not the input resistor.
I checked this against and I realized the 10k resistor is in fact the input resistor!
Will update schem.

8 ) Input resistor 10k not 1k.

Hopefully this is now correct:


Connections and colors
Numbering from left to right component side with chamfers at top.

1   Red      +9V
2   Brown      Pot 1 (ccw)
3   Yellow      Input
4   Blue      Pot 3 (cw)
5   Black      0V
6   Orange      Switch
7   Grey      Pot 2 (wiper)     
8   Violet      Output

[Rob Strand]

A small point.   DIY units vs original

The original units had a Threshold control which was wired with pins 1 and 3 reversed to what is shown.  That makes the *threshold* higher as you rotate the pot clockwise.    The Theshold pot was a  500kC reverse audio taper pot.

DIY units are wired as shown and the pot is Sensitivity.   That makes the unit more sensitive as you rotate the pot clockwise.  In this case the pot is a 500kA audio taper pot.

In this day and age the DIY version makes more sense and as a bonus that the audio taper pots are easier to get.

[GibsonGM]

Thanks for that, Rob...any suggestion for a common BJT to replace the SC's they used?

[PRR]

> a common BJT to replace the SC's they used?

Look like jellybeans.

You could probably pull them out of a 1990s VCR from the thrift-shop. "Blindly", except you want NPN on that plan. So it may be easier to use 2N3904.

[Rob Strand]

Thanks for that, Rob...any suggestion for a common BJT to replace the SC's they used?

Your could probably put anything in there and it will work.  A higher gain transistor for the buffer would be better.  If you look at the callat thread you will see the circuit might be affected in a small way by the gain of the other "peak detector" transistor.

On paper the originals had gains around 210 to 340.  Small Bear sells copies with hFE=250 and other have reported measuring gains of around 300 to 360.

[Mark Hammer]

I owned on of these little grey boxes back in the day, and I distinctly remember the footswitch having the definite click of a latching, rather then momentary, switch.  Should it be momentary?

[Baran Ismen]

May I ask for a reupload of the up-to-date schematics please, Rob? I can't seem to display it 

[duck_arse]

does this work?

[Baran Ismen]

does this work?

Thanks a bunch, it does!

[Baran Ismen]

Quick question; the 2SC1489 seems unavailable in Turkey, This website suggests an alternative for this one as BC168, BC183, and BC238, of those -183 and -238 are available locally. Can I use one of them?

The same goes for 2N5952 as well, it seems rare and available but rather expensive, 2n5457 is unavailable, maybe other suggestions?

And regarding the 741, I know its the original design but what would be the difference if a single 4558 or similar used?

[Baran Ismen]

I've redrawn the schematic on EasyEDA, I'm trying to improve my layout skills, so I'm preparing a PCB for this.

I've excluded the XLR output and Momentary switch section (as I would not use it), am I doing good ?

[antonis]

Make C1 100μF, place a 10 - 100 nF in parallel, place a 100R reistor between D1 Cathode and C1 and take +9V T to C1 positive leg..

[Baran Ismen]

Make C1 100μF, place a 10 - 100 nF in parallel, place a 100R reistor between D1 Cathode and C1 and take +9V T to C1 positive leg..

But its not the same as duckarse shared?

[Chillums]

It's just a R/C filter for the supply rail and is good practice.   The only difference would be a lower noise floor .

[antonis]

But its not the same as duckarse shared?

We DO love duck_arse a lot but reason other than his sharings..!!

[antonis]

But its not the same as duckarse shared?

We DO love duck_arse a lot but FOR reasonS other than his sharings..!!