DEBUG: Tonepad MXR Noise Gate

[Kinetic]

1.What does it do, not do, and sound like?
-- No noticable gating at all, signal fades out as normal.
-- big volume drop when effect is engaged.
2.Name of the circuit = MXR Noise Gate (Callate GATE 2_
3.Source of the circuit (URL of schematic or project) = http://www.tonepad.com/getFileInfo.asp?id=77
4.Any modifications to the circuit? None.  I have built the circuit twice, the first time I added in the pots for attack/release/attenuation.  The second time, I did the stock build.
5.Any parts substitutions? None.
6.Positive ground to negative ground conversion? No.
7.Turn your meter on, set it to the 10V or 20V scale. Remove the battery from the battery clip. Probe the battery terminals with the meter leads before putting it in the clip. What is the out of circuit battery voltage? => 9.07V
Now insert the battery into the clip. If your effect is wired so that a plug must be in the input or output jack to turn the battery power on, insert one end of a cord into that jack. Connect the negative/black meter lead to signal ground by clipping the negative/black lead to the outer sleeve of the input or output jack, whichever does not have a plug in it. With the negative lead on signal ground, measure the following:
Voltage at the circuit board end of the red battery lead = 9.03V
Voltage at the circuit board end of the black battery lead = 0V

Now, using the original schematic as a reference for which part is which (that is, which transistor is Q1, Q2, etc. and which IC is IC1, IC2, C1, and so on) measure and list the voltage on each pin of every transistor and IC. Just keep the black lead on ground, and touch the pointed end of the red probe to each one in turn. Report the voltages as follows:

Q1
C = 9.03
B = 2.07
E = 2.27

Q2
C= 2.62
B= 0.29
E= 0

Q3
G= 2.01
D= 3.31
S= 3.32

IC1 (or U1)
P1= 3.31
P2= 3.32
P3= 3.31
P4= 0.00
P5= 3.03
P6= 4.52
P7= 4.52
P8= 9.02

Z1 (zener)
A = 0.00
K = 3.31

I have built this circuit twice as stated above, once with the tonepad mods in red on the schematic, and once without.  In both builds, the circuit did not function.
In case it is of any use, I have noticed that in both builds, there is a substantial voltage change across the 0.047 capacitor which runs from IC1a pin 1 to the 100k resistor/1uF cap combo.  The voltage on the 0.047 cap leg which is connected to pin 1 is 3.31V whereas the voltage on the other leg is 0.02V.  I have changed the capacitor, checked the solder joints and as stated, even rebuilt it, with the same result each time.  I haven't a clue what's going on.

The voltages given above are for the simpler build (no mods).  The two 2N3904s have matched hFE values.  I have socketed the JFET and Ra.

Any help would be awesome...

[oldrocker]

Tonepad has some of the best and great circuits and projects out there.  Having said that the only build I never got working was this one.  Well that's not totally true.  I got it working finally when I switched from a dual op amp to two single op amps 741's.  Why that worked is still a mystery to me.  I have a hundred or so diy pedals that I've built successfully.  The Noise Gate using a dual op amp is my only failure.  Although since I got it working using two 741's I didn't count it. LOL!  Anyway using the schem I found in this forum it will work.  BTW DO NOT USE THE FREEINFOSOCIETY schem if you decide to try 741's.  It has the jfet pinout wrong on thier website.  Good luck.

[gigimarga]

I tested in my clone a lot of JFETs like BF245, BF 244, 2N3819 which I had when I built it.
It worked only with 2N5952!
And I tried about 10 (using a socket) until I found the best

[Kinetic]

I found this vero layout by searching this board, does anyone know if it's verified?

http://www.aronnelson.com/gallery/main.php/v/cathexis/Veroboard+Layouts/MXR+Noise+Gate/MXR+Noise+Gate.jpg.html

[oldrocker]

I used an MPF102 in mine although I don't remember trying any others.  Again watch those Jfet pinouts.  They'll get you screwed up if you're not sure.  I brought this pedal back since this thread and mine stopped working AGAIN.  I've decided to rebuild it on a bread board and try it again using a dual op amp.  To be honest this project is one of those maybe not worth the trouble circuits.  Even when it was working I never used it.  Maybe with mods it might be better but I think I'll stick to my DOD G7 gate or Rocktron Hush.  If you read past threads on this you'll find this build next to the SG-1 Slow Gear (I did get the SG to work btw) one of the most unsuccessful projects in this forum for many builders.

[Mark Hammer]

It might be a non-issue, but on some drawings I have seen, the footswitch is shown as momentary.  I used to own one of those grey boxes way back when, and the footswitch was NOT momentary.   Perhaps some of those who have built it may have been working under some mistaken assumptions about the nature of the switch.  Small wonder that effect and bypass seem so....similar.

[Kinetic]

I've done a bit more reading today and it does seem that a lot of people who normally have very few problems have also been unable to get this circuit to work properly, which sounds a bit dubious.  I'm pretty new to this but I've troubleshooted this and tested it a lot and I really can't see anything wrong.  I figured I probably had a cold solder joint or something which I wasn't picking up, so I thought a rebuild would fix it.  But when it was the same again it stopped me in my tracks.  I'll double check the JFET and I might also change the 1M resistor after the JFET gate to 1.8k as I've seen recommended in some build reports to supposedly address the volume drop.  If I'm still not happy, I'll put it to the side for a while and try something different.

[caspercody]

I built this two weeks ago, and when I turned it on it did not work. I made a pcb for this and checked all my traces, no shorts, Checked all my components, all were correct values.

I replaced the 2N5952 with another Jfet, and it works. Not sure why it did not work with 2N5952, but it works with any other Jfet I have.

[Kinetic]

Can you give the part number of a jfet which worked for you?  I've socketed my jfet so that's something else that's easy to try.

I think I've got a couple of 2N5457s lying around for an Orange Squeezer I've not built yet.

[.Mike]

I have an original from 1980.

I've tried helping people with this in the past. Looking through my PM archives, here are some things that might help:

My schematic is pretty much the same as the Tonepad schematic. Most of the changes on mine are the same as other people have posted.

- Input cap  = 10nF / .01uF
- Cap listed at 1uF NP = 10nF / .01uF
- Output cap = 10uF
- FET Drain connected to positive end of electrolytic
- FET Source connected to Bias
- FET = 2N5953
- With a 9.71v battery, Bias = 4.79v

I tried to read the value of the zener diode, but I can't get a proper reading without removing it. I'm not really keen on modifying a 30 year old pedal. It has either an N or a Z on one side, and I think the other side says 163 or 165.

Sound sample / gutshots:

Strat -> MXR NGLD -> MT-2 equivalent -> Valve Junior (w/ NFB) -> ISO Cab -> SM-57 equivalent -> Yamaha mixer -> M-Audio Delta 66

I have all three pickups turned on, and I do *not* have a RW/RP middle pickup. I left my TV turned on to get as much background noise as possible, and used my crappiest, longest cables.

I start with the guitar volume down, then turn it up so you can hear the noise floor. I then strum a chord. I then turn the noisegate knob until it starts to have an impact (about 3 o'clock on the knob). I then play a couple of chords and let them ring out until the gate kicks in. I adjust the gate after each chord, turning the knob clockwise.

I then kick on the MT-2, which adds a ton of noise to the signal. It allows you to hear when the gate kicks in very easily. I do the same as the clean sample, starting with minimal gate, and slowly adding more in.

Really, this pedal is kind of twitchy. It sputters a bit when the gate kicks in. I really don't use it much.

http://www.americasdebate.org/MXR-NGLD/MXR-NGLD.mp3

I could have sworn that I had taken full voltages on it for someone at one point, but I can't find them. I could probably take them again if it would help (and if I can find a usable 9v).

I don't think it's that great of an effect. It seems to abrupt. Downward expansion seems much better suited to guitar.

Mike

[oldrocker]

If you have parts for an Orange Squeezer I would suggest building that.  The OS is a great compressor and I'm sure that we can get it working for you.  I'll also keep u posted on my Noise Gate bread board rebuild.

[gigimarga]

.......
If you read past threads on this you'll find this build next to the SG-1 Slow Gear (I did get the SG to work btw) one of the most unsuccessful projects in this forum for many builders.

How??   

[caspercody]

I ended up with a 2N5457.

[oldrocker]

Hi gigimarga, we can start a new thread or re-open a new one on the SG-1.  I will say this about the Slow Gear it is a touchy circuit to get right.  Similar to the MXR Noise Gate in that it's temperamental.   I believe finding bi-polars and jfets that will work in the circuit is pretty much hit and miss.  Even when they are of the same type.  For example I tried several 5952's and 4401's in the right combination and got lucky.  I had to keep tweaking the trim pot and knobs after each transistor swap until I heard an audible swell.  Yes a real pain in the wazoo.  With this noise gate it should'nt be as hard but I'm finding it harder for me.  I built both of these on perf and I was not as experienced when I built the noise gate so I'll give it another go.

[gigimarga]

Thx oldrocker!
I will build another version from scratch (using another PCB - not the GGG one) and I hope tp make it alive. Now I have some 2N5952, so my chances are increased.

[oldrocker]

Not meaning to hijack this thread again but one thing I wanted to mention about the SG-1.  This effect requires a whole different playing technique to initiate the swell.  The first note has to almost completely (or completely) die out before the next note is struck.  If you test it and just keep strumming you won't know that it's working.  I think many builders thought thier build wasn't working because they didn't realize this.
If you have the actual components used in the schem I would use those.  Even though mine works I ended up using a 5.1volt zenor and the transistors I stated above.  It's what I had in my parts stash at the time.  I believe though that the original components might be better.

[cathexis]

I found this vero layout by searching this board, does anyone know if it's verified?

http://www.aronnelson.com/gallery/main.php/v/cathexis/Veroboard+Layouts/MXR+Noise+Gate/MXR+Noise+Gate.jpg.html

My build works great. This circuit needs tweaking, it won't work stock, in my experience. See the notes at the bottom of the layout.
LARS

[5thumbs]

I built this pedal on Francisco's v2 board a while ago.  (If you search, you'll find my posts...maybe...if the search engine is feeling generous today. )

I had problems getting the gate to open and close in a usable fashion, so I tried matching FETs as a first tactic...and finding that low-hFE wasn't as important as having close-matching hFE on both bipolars.  I eventually ended up tweaking the resistor/cap combo in the IC1a feedback loop to take care of the oscillation and gate open/close issues.

Here's my two Tonepad.com build reports with the info about what I did to get my MXR Noise Gate working:

Name: Brett Miller (5thumbs)
E-mail: private
Date: 1/5/2008
Build Report: This is a minor update to my build report (so you can add it to my 12/28 report if you'd like, Francisco.)

I solved the issue with oscillation in my previous build report. Once again, the fix focuses upon the 680-ohm / 1uF combo in the IC1a feedback loop. By raising the value of the resistor, you can eliminate oscillations that may occur when you set the gate to trigger at lower signal levels. I ended up replacing the 680-ohm resistor with a 4.7K-ohm resistor to get rid of the oscillations and allow use of the full 300-degree sweep of the Sensitivity pot.

The by changing the value of the capacitor in that combo, you can control how quickly the gate clamps down. At the stock value of 1uF, I had problems with the gate shutting too early, even with the Sensitivity pot turned to have the gate shut at very low signal levels. While testing, I tried a 0.47uF cap in that spot and it made the problem worse. A 1.44uF (or 1.5uF) cap made things better, but I ultimately settled on a 2uF cap in that position. With the 2uF cap, the gate will allow notes to ring to natural extinction without shutting the gate (with the Sensitivity pot turned to have the gate shut at very low signal levels.)

So in the end, a 4.7K-ohm resistor coupled with a 2uF Xicon film cap got rid of the oscillations and made the entire 300-degree sweep of the Sensitivity pot usable. These changes took my MXR Noise Gate build from merely good to "I'm-Keeping-This-One-Forever" status.

Even if you're having issues on this build, don't give up on it, folks. The results can be very impressive.

Good luck!

Name: Brett Miller (5thumbs)
E-mail: private
Date: 12/28/2007
Build Report: I built this project after reading many of the build reports here. As always, the board was very nicely done and easy to populate and solder. (Thanks Francisco!) My only suggestion would be to add the 2M2 resistor and 47pF cap to the parts list in the documentation, since these parts are listed on the top-side silkscreen.

As recommended in several build reports, I replaced the 1M resistor between Q3 Gate and Attack with a 1.8K resistor. (I tried it with both values and found the 1.8K resulted in better non-gated volume output.) I tried the mods, but found that I did not need them. On my build, I only have 1 pot for Sensitivity, which works pretty well once you get all the bugs worked out.

Initially, the gate would not release except with the loudest playing, and even then it was very staccato. I quickly discovered that I hadn't removed Ra after trying the Attack/Release mods. *blush* After I removed Ra, things worked better, but the gate still kicked in way too early.

I then scoured the diystompboxes.com forum for possible causes (and there are a LOT of threads there about problems with this project.) I found a post by Mark Hammer in which he talked about the original MXR noise gate being made primarily for +4dB inputs so it could be used with mixers/etc. Seeing that low gain from the IC1a op amp could be my problem, I decided to tinker with the cap/resistor in the feedback loop. I replaced the 680-ohm and 1uF cap in the feedback loop with a 470-ohm and 1.44uF cap, respectively. This combination increased the gain of IC1a while almost matching the low-pass corner frequency of the 680ohm/1uF combination. I tried at 340-ohm/2uF combination, but it increased the gain too much, causing oscillation to set in earlier and further reducing the useable range of the Sensitivity control.)

I used a 2N5485 and two 2N3904s in this build. After reading here about having matched Q1/Q2 with hFEs of 150, I measured mine. I had 3 around 180 and one around 200. As long as I used two that were almost the same hFE, I had a smooth decay when the gate kicked in. (When I tried the 200 hFE with a 180 hFE, the decay was more jagged and stuttery.) While I don't have two 2N3904s at 150 hFE, having two that match up pretty closely seems to be the important part in getting a smooth decay.

The usable range on my Sensitivity control is from 6:30 to about 11 o'clock (with 6:30 being the "zero" value to the user.) Beyond 11 o'clock, oscillation sets in. This is a minor issue, because with Sensitivity at 11 o'clock, the guitar signal rings to almost complete inaudibility before the gate kicks in. To me, this is about perfect, as I can turn the Sensitivity control to 11 o'clock when playing quiet Strat neck single-coil parts, but then crank it down to 6:30 for high-gain stuff. I'd like to find a way to tweak the pot to make the entire 300-degree sweep of the pot useable, but since the pedal sounds so good right now, I'm not sure I'll risk screwing it up to address this minor issue.

Final thoughts on this build...if this project is going to be your first build, you'd be wise to pick something easier to start with. This noise gate is probably one of the best I've used with guitar in terms of smoothness of decay. But to be honest, the MXR Smart Gate is pretty good and doesn't cost much more to buy. (I built this because I had most of the parts sitting on my workbench.) I used film caps for everything except the three 10uF caps. (I used an electrolytic for the power coupling cap in the upper-right corner and two 10uF tantalum caps in the middle-bottom of the board.) The TL072IP op amp worked out well, giving only the slightest mid-range bump (which it adds to pretty much everything anyway.)

Matching the hFEs of the NPN transistors appears to be important in getting a clean attack/release. Tweaking the gain of IC1a appears to be important in getting proper engagement of the gate when opening and closing. If you have no idea how to tweak these parts (and aren't really willing to learn), do yourself a favor and skip this build. If you decide to pursue this build, it has the potential to be challenging, but certainly wasn't the hardest build t-shoot/tweak I've had to do. If you stick with it and do your homework, you'll end up with a nice sounding gate in the end (and the satisfaction that you built and tweaked it yourself.)

Good luck!

[MikeH]

Tonepad has some of the best and great circuits and projects out there.  Having said that the only build I never got working was this one.

This is 100% true for me.  For whatever reason, this project totally defeated me.  I tried several times; all the "fixes" I could find here and in the build reports on tonepad and NOTHING worked.  I went as far as totally starting over from scratch with all new components and I still couldn't make it work.  *sigh....

[thedefog]

I just built this as a noise gate for my Sound Lab Ultimate. It works for both the synth, and on a guitar with both single coil and hot humbuckers.

My build was V2 and I bought a PCB from Francisco (he's got fantastic boards, BTW). I followed his layout instructions exactly and didn't change any of the component values. Used the 150k for Ra. I don't have any issues with volume drop.

I used two 2n3904 and a 2n5485 in my build. I used a TL072 for the ic. 500k log used for sensitivity. I got lucky with my decision not to socket the transistors or the IC.

I jumpered all of the mod pot sections, with the exception of the release pot, where I put in a 500k trimmer.

I did continuity testing on a bunch of 2n3904s between the emitter and base until I found a pair that had matching voltages. They both read 7.05v on the volt meter.

If anyone needs it, I can poke around and check voltages on mine to compare.