Callate 2 (MXR Gate)

[jfrabat]

You may add a bigger cap @ Vb also....

Certainly worth trying.

How much bigger?  I have 22uF, 33uF, 47uF, 100uF, 220uF...

I tried to determine the nature of the noise from the audio in your video.  I took the tail part of the audio, heavily high-pass filtered it at 1kHz and amplifier it.  I head a constant noise which has an unnatural character.

I am not sure you are not hearing the ceiling fan there...  The "hiss" I hear is more of a sort of clipping or something.  Maybe this vid will help:

By the way, this is where we are at now:



220 Ohm Resistor added in series with the with 47 uF cap and added a 47 Ohm resistor to the +9V line.  100 pF replaces 47 pF ceramic capacitor.

[Rob Strand]

How much bigger?  I have 22uF, 33uF, 47uF, 100uF, 220uF...

100uF should be enough to know if it has an effect or not.

I am not sure you are not hearing the ceiling fan there...  The "hiss" I hear is more of a sort of clipping or something.  Maybe this vid will help:

OK it's much clearer now.   I would call that a "fizz".  It seems to be there when the notes are present.  Also it kinds of gets worse as the Sensitivity pot is increased.   

There's a couple of places in the video where it seems to cut-out.  Once at 2:12 (1/4 pot setting) and 2:19 (0 pot setting).  It that the pedal doing that?

So one test worth doing at this point is to lift one end of the 10uF going to the JFET.   

You should definitely try this test as it will help narrow down the search.  The gating function is disabled in this test.   The test will tell us if the noise is getting into the audio from the Sensitivity circuit  or if the process of gating is causing it.   So with that change repeat the test in the second video with the different pot settings.

[jfrabat]

Yes, the lower the pot, 5he less "fizz" there is.  Actually, now with all the changes, i cannot fully close the gate (I could before) and totally block out the guitar (not that I woukd want to, though).

I will try the test tomorrow (cant today) and report back.

[Rob Strand]

Actually, now with all the changes, i cannot fully close the gate (I could before) and totally block out the guitar (not that I woukd want to, though).

That piece of information might be useful.   If you can't full close the gate with the pot set to zero that means noise is getting into the circuit around IC1a, or, IC1a is oscillating.   Think of it this way:  If you have the pot set to zero you would expect the output of IC1a to be zero that means the gate should have no signal to turn on and the gate should stay closed.

Another thing you should try is connecting the metal body of the sensitivity pot to the circuit ground.   That can help stop noise getting into the pot and IC1a.

[Fender3D]

If you can't full close the gate with the pot set to zero that means noise is getting into the circuit around IC1a, or, IC1a is oscillating.

actually the issue is the dual op-amp: IC1a has a huge amount of gain... MXR used 2 singles back then, and kept them far apart each other

[jfrabat]

UPDATE: Here is what we have done so far:

• JFET is a K161 (2SK161?)
• Zener changed to 1N5234
• Ceramic cap replaced from 47pF to 100pF
• 220 Ohm resistor added in series with 47uF cap
• 47 Ohm resistor added to power line
• Changed power rail cap from 10uF to 100uF
• Changed the Vb cap from 10 uF to 100 uF
• Grounded the rear pot casing

Issues still present:

• Fizz inserted with guitar sound (dies away as the guitar sounds dies)
• Fizz highly dependent on pot position (the more I close the gate, the less fizz there is)
• Gate is not fully closing

Voltages (as of today):
Battery: 9.1V

IC (TL072CP):
1: 6.31V
2: 6.30V
3: 6.30V
4: 0.00V
5: 4.12V
6: 4.32V
7: 4.32V
8: 8.67V

Q1 (2N3904):
E: 5.31V
B: 5.66V
C: 0.00~190mV

Q2 (2N3904):
E: 0.00V
B: 0.501V
C: 4.63V

Q3 (K161):
D: 6.30V
S: 6.30V
G: 5.16V

Zener (1N5234):
Cathode: 6.30V
Anode: 0.00V

Sound clip (at youtube standards!):

NOTE: After gate halfway open, I turned the pot before the sound completely died out.  Sorry about that!

So what's the next step (other than giving this thing a name, 'cause there is lots of changes from the original!)?

[Rob Strand]

So what's the next step (other than giving this thing a name, 'cause there is lots of changes from the original!)?

Don't worry about the many changes.   When debugging it is common to modify the circuit to understand the problem.   Many of those changes can be removed later on.   At this point it is OK to leave them in.   If you had an oscilloscope you could check stuff without making so many changes.   It's always more difficult to debug circuits with limited equipment - I did that a lot when I was young.

As far as progressing is concerned there's still *many* things to try. 


Try doing these one at a time in this order:

- Change to 47pF  (which is now 100pF) to 220pF
- Remove the 1uF cap (the one across the 100k resistor, brown color) and replace it with a 10nF cap
- Replace the 6.2V zener with the original 5.1V zener in series with two silicon diodes.
  So in place of the zener use three diodes like this.  Note the silicon diodes point
  in the opposite direction to the zener diode.

  Vb ---- >|------ >|------|<------ 0V
             Si       Si        5.1V zener

  With that connection you should get about 5.2V on Vb.


Don't give up!

[jfrabat]

Change to 47pF  (which is now 100pF) to 220pF

The biggest I got is 180 pF; is that enough?

[Rob Strand]

The biggest I got is 180 pF; is that enough?

It would be fine.  In fact I'd just put that across the 100pF that's already there giving 100pF+180pF=280pF.
For this test a bigger is better.   

The reason why we change this part is to filter the high frequency noise more.   It can help reduce oscillations also.  In the original circuit the opamp had a low "gain bandwidth" and a high gain (gain = 1MEG / 680 ohm) which creates a filter by itself.

[jfrabat]

The biggest I got is 180 pF; is that enough?

It would be fine.  In fact I'd just put that across the 100pF that's already there giving 100pF+180pF=280pF.
For this test a bigger is better.   

The reason why we change this part is to filter the high frequency noise more.   It can help reduce oscillations also.  In the original circuit the opamp had a low "gain bandwidth" and a high gain (gain = 1MEG / 680 ohm) which creates a filter by itself.

OK, something is odd here; I added the 180 pF cap, and used my capacitance multimeter to be sure the solder was true and had joined both caps together.  Turns out, the MM reads 42 pF with both caps on.  I measured my stock caps, and they are reading more or less OK (168 pF for the 180 pF and 90.6 pF for the 100 pF caps).  Unless something in the board is playing games with me, I cannot understand why it would only read 42 pF...

As far as sound, same issue persists.  Gate closes better now (not 100%, but at least 95%, meaning that if I strum lightly, the sound does not get through; if I strum strongly, then about 1~2 seconds goes through, then it closes).

So what's the next step (other than giving this thing a name, 'cause there is lots of changes from the original!)?

Try doing these one at a time in this order:

- Change to 47pF  (which is now 100pF) to 220pF
- Remove the 1uF cap (the one across the 100k resistor, brown color) and replace it with a 10nF cap
- Replace the 6.2V zener with the original 5.1V zener in series with two silicon diodes.
  So in place of the zener use three diodes like this.  Note the silicon diodes point
  in the opposite direction to the zener diode.

  Vb ---- >|------ >|------|<------ 0V
             Si       Si        5.1V zener

So, do I move to the 1uF cap and replace with 10uF, or do we examine this weird cap issue?

[Rob Strand]

OK, something is odd here; I added the 180 pF cap, and used my capacitance multimeter to be sure the solder was true and had joined both caps together.  Turns out, the MM reads 42 pF with both caps on.  I measured my stock caps, and they are reading more or less OK (168 pF for the 180 pF and 90.6 pF for the 100 pF caps).  Unless something in the board is playing games with me, I cannot understand why it would only read 42 pF...

Don't worry at all.  When you measure capacitors in-circuit the other parts in the circuit affect the measurement.  Some meters work better than others depending on the method that meter uses to measure the capacitance.  The way to interpret in circuit measurements is if it is close to what you expected then the measurement work.  If it didn't work then it's likely the other parts in the circuit are affecting the measurement.  You can get the same problem with resistance measurements.

As far as sound, same issue persists.  Gate closes better now (not 100%, but at least 95%, meaning that if I strum lightly, the sound does not get through; if I strum strongly, then about 1~2 seconds goes through, then it closes).

The extra filtering from the cap is probably keeping the noise under the gate threshold.   

[aspangev]

Hi guys

I had exactly the same problem but it is easily solvable.

This is my first post here. I registered here only for this issue.

I work as an analog design engineer and as a hobby I'm a bass player and build some stompboxes for me and my friends.

Yesterday I built this popular MXR noise gate for me and made some minor modifications because I couldn't find the original JFET and Bipolars. Everything was ready but just before closing the box I noticed strange bizz when I increase the sensitivity of the gate. I couldn't sleep... Today I read this topic and decided to debug it.  I spent few hours trying to isolate detection circuitry but the bizz was there... I almost gave up but after few glasses of wine I've got the answer. The bizz comes through the input of the high gain amplifier (OA1a). The input buffer is very weak - around 200uA and its output impedance is relatively high. This allows the the bizz to penetrate between the input and output buffer. The problem is easily solvable by increasing the current through the first buffer.

jfrabat, please change the resistor on Q1 emitter from 22k to 4.7k. This significantly will decrease the buzz to levels that you can't hear it.



Good luck!
Greetings from Bulgaria

[jfrabat]

Hi guys

I had exactly the same problem but it is easily solvable.

This is my first post here. I registered here only for this issue.

I work as an analog design engineer and as a hobby I'm a bass player and build some stompboxes for me and my friends.

Yesterday I built this popular MXR noise gate for me and made some minor modifications because I couldn't find the original JFET and Bipolars. Everything was ready but just before closing the box I noticed strange bizz when I increase the sensitivity of the gate. I couldn't sleep... Today I read this topic and decided to debug it.  I spent few hours trying to isolate detection circuitry but the bizz was there... I almost gave up but after few glasses of wine I've got the answer. The bizz comes through the input of the high gain amplifier (OA1a). The input buffer is very weak - around 200uA and its output impedance is relatively high. This allows the the bizz to penetrate between the input and output buffer. The problem is easily solvable by increasing the current through the first buffer.

jfrabat, please change the resistor on Q1 emitter from 22k to 4.7k. This significantly will decrease the buzz to levels that you can't hear it.



Good luck!
Greetings from Bulgaria

Great!  I had this just laying around and really moved to something else, but I will give this a go!  THANKS!

[Rob Strand]

I had exactly the same problem but it is easily solvable.

This is my first post here. I registered here only for this issue.

I work as an analog design engineer and as a hobby I'm a bass player and build some stompboxes for me and my friends.

Excellent work, thanks for posting!   That problem really needed someone to get in there a fix it.

I noticed that high emitter resistor as well but I couldn't see a reason why it would cause a problem since all the loads off the buffer are relative high impedance.  However now, after looking at the Waza buffer thread, I can see the 22k load resistor to the JFET is going to make the buffer stage (with a 22k emitter) clip at a lower voltage than Vcc/2.   

So maybe the real problem is when the gate shuts off and the JFET starts to turn on, the 22k series resistor then loads down the buffer, then that reduces the swing of the buffer and the buffer clips prematurely.

I wonder why a lot more people aren't seeing that?   It seems like it should only occur then the signal is strong but jfrabat's videos show the effect when the guitar notes are decaying and guitar voltage will be low!  Raising the  zener voltage should have helped a bit too.

[aspangev]

So maybe the real problem is when the gate shuts off and the JFET starts to turn on, the 22k series resistor then loads down the buffer, then that reduces the swing of the buffer and the buffer clips prematurely.

I wonder why a lot more people aren't seeing that?   It seems like it should only occur then the signal is strong but jfrabat's videos show the effect when the guitar notes are decaying and guitar voltage will be low!  Raising the  zener voltage should have helped a bit too.

No, the problem is not overloading the buffer and clipping the signal. The real problem is that pulses from OpAmp's output (IC1a) occur also on IN- because the OpAmp is saturated and then, probably through parasitic capacitance or just electrically inside the OpAmp, the pulses occur on IN+. Once appeared on IN+ the pulses penetrate the signal chain through the cap 0.01uF. A lower impedance buffer would not allow this penetration. This was proven by some experiments - cutting the JFET circuitry - no any impact, the pulses are there. This means that the only way for back penetration is through the OpAmp's input or supply but we know that putting bigger supply cap doesn't help, so the supply is not the problem.
The other experiment I did is limiting the OpAmp's output  - I put 2 parallel diodes on the OpAmp feedback. That way the output voltage swing is about 2*0.7V and the OpAmp is not saturated - the bizz disappeared completely but the circuit was not sensitive enough.
The only solution I see is making the input buffer stronger or completely redesign the schematic. Happily making stronger buffer is not tough task - just replace the emitter resistor. The result is incredible - I can't hear any bizz!

Please share this mod on the most popular topic about this MXR noise gate.

Best regards
Pangev

[Rob Strand]

No, the problem is not overloading the buffer and clipping the signal. The real problem is that pulses from OpAmp's output (IC1a) occur also on IN- because the OpAmp is saturated and then, probably through parasitic capacitance or just electrically inside the OpAmp, the pulses occur on IN+. Once appeared on IN+ the pulses penetrate the signal chain through the cap 0.01uF. A lower impedance buffer would not allow this penetration. This was proven by some experiments - cutting the JFET circuitry - no any impact, the pulses are there. This means that the only way for back penetration is through the OpAmp's input or supply but we know that putting bigger supply cap doesn't help, so the supply is not the problem.
The other experiment I did is limiting the OpAmp's output  - I put 2 parallel diodes on the OpAmp feedback. That way the output voltage swing is about 2*0.7V and the OpAmp is not saturated - the bizz disappeared completely but the circuit was not sensitive enough.
The only solution I see is making the input buffer stronger or completely redesign the schematic. Happily making stronger buffer is not tough task - just replace the emitter resistor. The result is incredible - I can't hear any bizz!

Excellent experiments.  I was thinking of that diode trick as well.  2x Red LEDs might be better but beyond that you run the risk of the problem coming back when the battery runs down and the opamp clips before the LEDs.

I'm not 100% convinced the fizz is getting back through the opamp inputs since jfrabat still got fizz with sensitivity pot backed off from full (so the impedance from the opamp input back to the emitter is fairly high).   Capacitive coupling sounds like a good cause.  The tone-pad layout has quite a few "noisy" tracks passing near the audio signals around the opamp.   The PCB layout is really asking for trouble.

Were you using the tonepad PCB?

The circuit could be helped with some re-design.  I noticed the Boss NF-1 Noise gate is pretty much an improvement on the MXR.

Anyway, you did a great job working all that out!   I felt like building that circuit but I can't afford the time to build every circuit that has weird problems.

[Rob Strand]

To jfrabat,  you can remove all the mods we added.   I would keep the 220pF on the opamp since I believe that addition makes the circuit behave more like the original.  Obviously you need to keep the 6.2V zener.

[aspangev]

I'm not 100% convinced the fizz is getting back through the opamp inputs since jfrabat still got fizz with sensitivity pot backed off from full (so the impedance from the opamp input back to the emitter is fairly high).   

Hey Rob

I'm happy you mentioned that. I had impression that this problem occurs only in my project because I don't use the original JFET and bipolars. This actually is another problem that is also easily fixed.
The popping sound that is heard at the end of the last clip is caused by the sharp switch-off of the JFET. In order to cure this, I connected a cap of 4.7nF between the JFET's gate and the ground. This way the gate voltage is 'soften' and the popping sound disappears. Proven!

I use universal PCB but similar layout as from tonepad. I don't believe the parasitics are from the PCB layout but from IC itself. Maybe it's possible to find some OA (different type, different producer) that don't have this issue. I use TL072, Texas Instruments.

Anyway, I will draw later today all the mods I did.

[aspangev]

Here are the changes in my project:



1. I don't see benefit of using the 3 knobs: ATTENUATION, ATTACK and RELEASE so I removed them. I just increased the 'release' resistor to 680k. I prefer slower release. It's actually less than a second.
2. Increased all signal capacitors because I use it for a bass guitar.
3. Connected 4.7nF cap on the JFET gate - removes the popping sound around the threshold of the gate.
4. Decreased 22k -> 4.7k resistor connected to the Q1 emitter - removes the hizz we discussed earlier.
5. Added LED for visualization the gate. Used NMOS BS170 but can be any small power NMOS with Vth ~ 2V
6. This is the mod I'm going to try when we get back from the tour. I don't see benefit of using such big cap (10uF) in JFET circuitry. I face a problem, when I switch-on power supply the output is muted for about 10 seconds before I could start playing. This probably is caused by charging this big cap through 1M resistor. Calculated, RC = 10 sec. I believe a 1uF cap will be enough. The cut-off freq will be 1/(2Pi*1u*22k) = 7.2Hz which for me is enough.

@jfrabat, I saw you have the same problem with the popping sound around the gate threshold. Please, make the mod No.3 also. This will almost remove the popping and at the same time the attack will be enough not to have delay in deactivating the gate. Please share info when you fix your project.

Generally I'm very happy with the gate behavior with all the changes I did.
I'll be grateful if someone else prove and like the mods

BTW you can support our 'young' band (I hope is not forbidden from the forum rules):

Best regards
Pangev

[jfrabat]

To jfrabat,  you can remove all the mods we added.   I would keep the 220pF on the opamp since I believe that addition makes the circuit behave more like the original.  Obviously you need to keep the 6.2V zener.

OK, understood.

@jfrabat, I saw you have the same problem with the popping sound around the gate threshold. Please, make the mod No.3 also. This will almost remove the popping and at the same time the attack will be enough not to have delay in deactivating the gate. Please share info when you fix your project.

Will do! 

Changes will.most likely be updated in 2 weeks, as this week is a little hecktic, and next one I am taking my family on vacation...

By the way, I like your music...  will download it and take it with me on my vacation!