MXR Noise Gate Build

[oldrocker]

I finally got this thing to work.  I was going to try and use the Tonepad PCB but I'm not sure that would've worked.  I found in the Schematics section of this forum where Aron posted a working schem of this circuit from the FIS version using two 741 single op amps instead of one duel op amp.  I tried it and it fired right up.  I didn't change any other components and just routed wires around to split the duel into the two 741's and that's all it needed to work.  I almost gave up on it (actually I did give up) but I came back to it with a different approach and I finally got it working.  Now when I'm using heavy compression with distortion and I put my hand over the strings to mute I don't get that huge swell of white noise followed by excessive feedback.  It's just nice and quiet.   I don't know why a duel op amp won't work but I'm convinced it never would've worked unless I spit the duel into two single op amps.  Does anyone have any insight to why that would be?  Why is there a difference between op amp style set ups?  Shouldn't they work the same?

[oldrocker]

Oh I almost forgot.  I used 3904's for the bi-polars and a MPF 102 for the JFET.  I have some 5952's and tried them but they seem to cut down on the overall volume of the pedal.  So I put the MPF 102 back in.  And yes I 180'd the 5952 when I installed it.

[jimbob]

I had this board done for some time but I keep hearing the voices in my head from those in this forum awhile back saying, "It (noise gate's) will color your tone" and that scares me off enough to move onto something else.

Do you notice anything like that?

[oldrocker]

I'll have to get back with you on that.  The reason I say that is I tried it out with compression and some heavy distortion.  I'll need to test it with a clean signal and do some A-B tests.  I'll try that in the morning.  I can say this though from what I heard of it so far if it is coloring the sound it's not too bad.

[Cliff Schecht]

They do color your tone. I have a Rocktron Hush pedal that doesn't really cut down on noise too much, but makes everything sound ridiculously tight.

[oldrocker]

I guess it's going to be if the tightness bothers you or not.  I think I know what you're saying.  With the gate turned off the sound seems to open up.  When I used to use a hush in my band that color that the gating does to the signal sounded good.  My guitar didn't sound washed out while the whole band was playing.  The band sounded tighter without my guitar making all the noise at quiet spots or stops in the song.  Of course you can use volume foot controllers or a cutoff switches.  But the gate is automatic and when I'm in the zone I don't want to have to come out of it to turn off my guitar signal.  I saw a local band the other night and everytime a song ended the guitar was still making all that high gain noise until he stomped on a switch.  Very unprofessional and annoying.  So it depends on the situation and how you use it.  I only use a gate when I'm using a lot of compression and some heavy fuzz-distortion.  If I'm playing clean especially with no compression I never use gating since you really don't need it.  When I'm not using compression and I'm only using a TS or a OD250 I never use gates either. It's the added compression with higher gain pedals that I built the noise gate for.  The only way to find out if it's acceptable or not is to try it.

[oldrocker]

It's finished.  Works great. And was the toughest pedal I've built yet.  Remember if you're going to perf this effect like I did use the Tonepad schematic but substitute the TL072 duel op amp with two LM741 single op amps and it'll work.

[Fp-www.Tonepad.com]

You are making statements that are wrong.

The tonepad version works, I have built a few of them, all worked, and I know of dozens of people who have also built it successfully.

Congratulations on your build.

Fp

[oldrocker]

I'm sorry if my statements aren't correct.  It seemed the only way it would work for me was to do it the way I did.  I went over this circuit I swear a hundred times and I couldn't get anywhere with it.  I even rebuilt the whole thing twice with the same result.  I'm sure you're right there shouldn't be any difference between using two single op amps or one duel.  I've used duel op amps many times with no problems and I swapped out several of them but it made no difference.  My apologies if I am making false statements here.  Thanks Tonepad.

[KMS]

The Tonepad version works great....with no tone color added that I can detect none at all.

I used all the mods at tonepad too...they all work great.

It was my first build......the only two things I had problems with were I didn't ground the battery to my grounded guitar signal (beginner mistake) and once I got that right it seemed to be choppy on the attack particularly on high frequency notes (provided that I had it on a very light attack setting).

I then took the approach that if my FETs had identical HFE measurements that it might reduce the choppy attack.  So I bought a bag full of maching pairs....where my supplier let me check his entire stock for matching pairs.   HFE 150 on each transistor seemed to produce the smoothest results.

Your right about the idea that a dual op amp "should work"....because it does.

I can still get the choppy setting if I set the attack real light.....and that is a good FX if you find a need for it, but most of the time I rather not have that and I now have what I want.

[Fp-www.Tonepad.com]

This project has had a few problems because the FETs don't all have the same pinout, and for some, the outline in the layout image doesn't match the pinout of the FET (solution: look for datasheet for the FET you're using and match it to the layout).

Also, there is a schematic on the web (I based tonepad's project on that schematic) that had the FET traced wrong (due to the pinout issue mentioned above).

So, this project has it's share of things that could go wrong, and that has persuaded some people from building it, I just wouldn't want that to continue to happen and that's why I replied to this message.

The project is verified, in the tonepad way So you can be confident that if it doesn't work... (you can finish that sentence!)

Fp

[oldrocker]

Ok you've convinced me that I had a build problem with my duel op amp for some unknown reason.  KMS I'm a little confused about matching Jfets in a noise gate.  Are you saying that you had to find the Jfet that matched the specs for this effect the best?  Matching fets is normally for when using two or more of the same fet type to get them to match each other.  The noise gate only has one Jfet.  Right?  To clear things up I had no problem with my jfet install.  I had used a MPF102 originally and then when I got it working I tried some 5952's.  Since this was a perf build I didn't have the confusing trace problem with the layout.  All I had to do was install the 5952 180 degrees facing the other way.  It worked but not as good IMHO.  My problem had nothing to do with the Jfet.  When I read many past threads I noticed many had the exact same symptoms as mine did and they never resolved it.  I know a lot have got it working but do a search for past threads using "MXR Noise Gate" I think there's a majority that didn't.  It's hard to believe so many gave up just because they could'nt install one Jfet correctly.  I'm sure they would've figured that out sooner or later but many never did.  I finally got mine working and figured if what I did could help others get theirs working by doing what I did then it was worth taking the heat.

[Cliff Schecht]

I remember building the Callate Gate 2 and having nothing but problems... I gave up on the thing after a while. I should probably try to fix it up when I go back to Austin next time.

[petemoore]

So you can be confident that if it doesn't work... (you can finish that sentence!)
  Nice pass, I could probably use a good noise gate.

[Mark Hammer]

Okay, here's the deal about noise gates.

They do NOT "colour your tone".....when they're on.

The problem is that noise gates have to detect a signal in order to turn on, and turn off when the parameters of the signal change enough.  In a perfect world, all noise gates would be able to easily distinguish between noise and signal, and would turn on within 1msec, so that you wouldn't lose any of the initial attack in a note.  Note only that, they would be able to tell if a note is still lingering and wait until the note has truly died out before shutting off.

That's a pretty tall order.  Not only that, but if the gating action is controlled by a rectifier/envelope-follower circuit, and the rectification is not velvety smooth, the tails of fading out notes can be pretty ugly due to the envelope ripple.  Sadly, a great many designs use half-wave rectifiers (similar to the Dr. Q circuit; op-amp into a diode, with cap to ground) whee the use of a bigger capacitor to smooth out the decay phase has the side-effect of slowing down attack time - you can have your quick onset OR your smooth sustain, but not both.  You can have it set up for responding to quick strumming OR you can set it up for responding to sustained notes, but generally not both.

THAT is where the "coloration" comes in.  There is no change in the tone of what you get to hear, but there can often be a change in how much of the uncoloured note you get to hear.

In general, ALL noise gates, regardless of design, can be made to behave better if you give them an easier task to perform.  So, if it is easier to distinguish between noise and signal, whether because the signal is hotter or the noise is low, or both, then the threshold can be kept mercifully high (or low) without jeopardizing the notes.  Traditionally, noise gates perform best when used with punctate intermittent sources.  A kick drum is probably the textbook case.  It gets hit every so often rather than continuously, doesn't sustain, and the dynamic contrast between signal and ambient noise is huge.

The irony of noise gates is that, while we probably need and tend to use them at the tail end of a bunch of pedals that generate cumulative noise, hum, whine, and hiss, in terms of accurate note detection they are actually best placed at the start of the pedal chain, where the difference between noise and signal is much clearer.  There was a schematic of the DOD FX30 that was posted around, and from what I gather, it either includes a send-return loop or can be easily modified for one.  Here, the envelope-detection occurs using the clean unprocessed input signal, but the gating action itself is applied after all the processing where it is needed.

In some respects, what this calls out for is some sort of generic modular circuit in which a well-rectified envelope signal is obtained immediately after the guitar, and fed to both a compressor circuit and a gating circuit, which can then be inserted wherever in the signal chain you like or need.

[petemoore]

The irony of noise gates is that, while we probably need and tend to use them at the tail end of a bunch of pedals that generate cumulative noise, hum, whine, and hiss, in terms of accurate note detection they are actually best placed at the start of the pedal chain, where the difference between noise and signal is much clearer.  There was a schematic of the DOD FX30 that was posted around, and from what I gather, it either includes a send-return loop or can be easily modified for one.  Here, the envelope-detection occurs using the clean unprocessed input signal, but the gating action itself is applied after all the processing where it is needed.
  Very sensible approach for a well behaved, effective NG.
  But because it requires the end users knowledge or understanding [what you just wrote] ... and a couple more jacks/two more cables to route than a 'regular' NG...these reasons are probably mostly why DODFX30's Type NG's aren't a more common noise problem solver.

In some respects, what this calls out for is some sort of generic modular circuit in which a well-rectified envelope signal is obtained immediately after the guitar, and fed to both a compressor circuit and a gating circuit, which can then be inserted wherever in the signal chain you like or need.
  Sounds like a 'bester' approach to noise gating, especially for those who prefer the NG's gating the end of a chain where noise levels change markedly...like when a noisy effect is switched in.

[Cliff Schecht]

The thing about noise gates is that a lot of  people are using them for the wrong reasons. If someone is running a slew of pedals and they don't use batteries or a clean power source, a lot of 60cycle hum is introduced and putting a noise gate at the end of the signal chain is just going to squash the sound, especially since loud 60 cycle hum is fairly loud and hard for the noise gate to distinguish from the actual signal. Some people expect their noise gates to be a one stop solution to their noise problems, but proper shielding and grounding have made it to where even at ridiculously high gain and volumes, I never need a noise gate. Maybe we should start a campaign to educate America about clean power.

[KMS]

Sorry....I matched up the the 3904(s) Q1 and Q2 on the tonepad schemo.   It has been a while but I do remember that hfe 150 on both was the smoothest I could get it and still keep a lot of sustain.

I don't know why it worked better.....I can postulate why.....but I don't really know.

Mark is right about putting any NG at the beginning of the chain (that's where mine is).

Cliff has a good point too.

I originally had problems with my Boss OD2 making noise with high distortion settings......and when I remounted it in a steel box instead of the stock aluminum box.....it is much quieter....but it is still a little noisy.

My 1952 Kalamazoo with two single pick ups also adds to the noise problem but mostly because the pickups are weak compared to others like my Fender or my Ibenez. The stronger the signal from the guitar the better the noise gate works (at least for me).  Also I get cleaner sound from all my pedals with the stronger pickups. So there are a lot of little things that make up the noise issue.

[KMS]

By the way...I still have several 3904(s) that are pairs...ranging from around hfe 80 to hfe 300.  I hope to put them to use someday but I probaly never will since I now have a Digitech RP7 and pleny of other pedals.  I guess I am just about out of pedal building.....kind of miss it too.