noise gate idea, advice sought ;)

[pinkjimiphoton]

hi guys, i find more and more i could use a decent or even a crummy noise gate to play with, as some of the filth boxes i've got would be able to raise the dead much
more effectively without the scirrocco of noise they produce when like, 6 or 8 of them are engaged.

so me and petey were looking around at easy gates, and found one on techniguitare.. but petey said someone here said it didn't work.
well, if they built the vero in the pdf on that site, it sure won't... they have all kinds of mistakes in it.  CLOSE, but no cigar as groucho used to say.

so i took a look and saw the problem... they had reversed a couple component values from the schematic, and had pin 5 of the opamp grounded...when it should have had a 100n cap and 2.2m resistor there. i think that jumper to ground is why it wouldn't work as shown.

for reference, here's the forum page with files for the project:

http://techniguitare.com/forum/fiches-tutoriaux/noise-gate-sur-plaque-bandes-hell-gate-t9557.html

here's the original schematic:



i looked at it, and it looked easy, and doable on vero cuz it was simple.

so i aimed to overcomplicate it some.

also, they specced tl071, and i don't have any... i have 741's, but... i also have ne5534's, which are low noise jfet single opamps. an ass load of 'em.

so i figured i'd adapt it.

now, i've used a similar kind of 741with frequency compensation... and it sounded good, but without the frequency comp circuitry it's noisy as f***.

so i looked at the data sheet, looked fairly simple... a cap between pin 8 and 5 (don't know what value is good, i wrote .47 on the schem but on the vero it shows sockets to experiment with)
and a pot between pins 1 and 8, with a 22k resistor being fed by the b+ voltage.

i THINK it should enable finding the frequency for where it cuts off... not sure. need advice on this part, cap values and such please if possible.

if it's not worth f'n with, please illuminate me so i can ditch the stuff... or if i made some other dumb mistakes..

likely, with me. 

here's the data sheet or the 5534:

http://www.onsemi.com/pub_link/Collateral/NE5534-D.PDF

also added an ass load of power supply filtering/protection... current limiting resistor, snubber, diode in line and diode reverse biased to ground and a big honkin' 220u cap, hoping that by doing so it will make this thing as quiet as possible in operation.

added in a simple circuit to flash an led white the gate is operating. didn't bother  with t/b switching stuff.

here's the schematic first:



and here's the vero i worked up... a second set of eyes is always welcome in these matters of course... if someone can give me a thumbs up or down, i plan on building this puppy posthaste.



so... i would really appreciate advice, criticisms, GFY's, attaboys, whatever ya go!

thanks brothers...peace

[PRR]

NE5534 is not FET input.

"0.47" is NOT a likely compensation cap. Look at page 5 of that PDF. Likely values are zero to 100pFd. In fact you probably want 22pFd, though 30pFd or 47pFd will be fine also.

However, I do NOT understand the "original schematic".

* The opamp has an input at its negative supply, few opamps will be happy this way. It may work in LINE level apps, because the threshold pot goes up to 2V, where the opamp may work (and the offset in the 1Meg bias resistor may be moot).... but who wants to kill all signals below 2V? Or 0.2V, the lowest that threshold pot can be conveniently set?

* The connection of 2N3906 Base to the timing cap and MOSFET Gate looks like it might hold the MOSFET ON for any practical setting of the release pot.

In your hand-drawn: you do not want the "Balance" network on pins 1+8. That's for precision DC use, and you are doing adjustable audio comparison. And the Balance network is a great input for stray buzz and radio. Don't do that.

Seriously: there's better plans out there.

Real simple and easier to bebug: build a high-gain chain that will light an LED when signal is just above unwanted hiss. When that works, shine the LED on an LDR wired like that MOSFET. The LDR fast attack slow release will smooth the audio light, open sorta-fast and close kinda-slow. It's far from ideal, but as I say, it's simple and likely to lead to A Result.

[Scruffie]

Check out a DOD 230 (I think is the right number) pretty simple and meant to be pretty good.

[pinkjimiphoton]

NE5534 is not FET input.

but the pdf said it was..... for the tl071 

"0.47" is NOT a likely compensation cap. Look at page 5 of that PDF. Likely values are zero to 100pFd. In fact you probably want 22pFd, though 30pFd or 47pFd will be fine also.

thanks, i hadn't read the whole thing yet... now i understand what "cc" is... every ref is to 22pf. paul, i really appreciate you taking the time for this... thanks man..

However, I do NOT understand the "original schematic".

* The opamp has an input at its negative supply, few opamps will be happy this way. It may work in LINE level apps, because the threshold pot goes up to 2V, where the opamp may work (and the offset in the 1Meg bias resistor may be moot).... but who wants to kill all signals below 2V? Or 0.2V, the lowest that threshold pot can be conveniently set?

hmmmm, wonder if it's a typo, and should go to the positive input instead? maybe reversed?
i'm thinking a cranked dirt box may put out a couple volts peak to peak.... in that case, maybe .2v would be adequate... i totally don't understand this stuff. i DO need something that will work to chill out the fuzz collection.

* The connection of 2N3906 Base to the timing cap and MOSFET Gate looks like it might hold the MOSFET ON for any practical setting of the release pot.

can ditch that easily, i don't think it would be necessary in the first place. thought it would look cool, tho.

curse you petey getting me sucked into this stuff!!! lol

In your hand-drawn: you do not want the "Balance" network on pins 1+8. That's for precision DC use, and you are doing adjustable audio comparison. And the Balance network is a great input for stray buzz and radio. Don't do that.

gotcha. can ditch that, too easily.  would the 22p cap still go between pins 8 and 5? or just leave that out?  sorry for all the questions!!

Seriously: there's better plans out there.

of THAT, i have no doubt paul!! this just looked simple, and maybe fun to play with. i look at some of the projects out there and they look either impossibly overcomplicated,  or so simple i can't see how they'd work.

Real simple and easier to bebug: build a high-gain chain that will light an LED when signal is just above unwanted hiss. When that works, shine the LED on an LDR wired like that MOSFET. The LDR fast attack slow release will smooth the audio light, open sorta-fast and close kinda-slow. It's far from ideal, but as I say, it's simple and likely to lead to A Result.

let me see if i got this straight... you mean use the audio coming out of some stompboxes to light an led, and then have an ldr between signal and ground to the output?
or do you mean build a high gain preamp of some kind just to drive the led alone... sorry i'm not sure i'm following. i can see where the led/ldr thing would/ could work, but i'm unsure of what you mean...

i THINK you take the guitar output, buffer it, split it : send one side to an amplifier to light the led...use the gain of that stage to light the led to whatever level is just above unity... have the  ldr set between ground and the output of the other side of the splitter?  going to the amp? did i get it close?

something roughly like this? or am i still over complicating/misunderstanding?

[pinkjimiphoton]

Check out a DOD 230 (I think is the right number) pretty simple and meant to be pretty good.

thanks scrufie, i'll check it out!

[J0K3RX]

I think they made an updated version using a TL072?
http://techniguitare.com/forum/realisation-montage/tranquility-noise-gate-malheur-composants-t8976-90.html

[Buzz]

Hi Jim. I've put a gate on my to-do list for similar reasons. The simplest I've found is the MXR Noise gate. Couple of transistors to match is about as tech as it gets. I don't know if I'm allowed to link to it, but there is a "verified" vero layout for it easily found.

edit: Actually a version of it is on this site too.

[samhay]

I'm with PRR - this design would make more sense if you used a single-supply op-amp. I would imagine you will also get some distortion as the MOSFET switches off.
More gnerally, while the release pot should work nicely, the turn-on time of this design will be relatively fast, so I am not sure how musical it will be. Jimi - did you try it on the breadboard?

Either way, I am fairly sure that the status LED will not work as shown, and I note that the dual op-amp version that J0K3RX posted doesn't include this. If you want a status LED then replace the BJT with another mosfet with source to ground - this will turn the LED on when it is happy to pass signal.

[Mark Hammer]

If one is not going to use an SSM2166, then my vote for next best simple solution is to avoid JFETs like the plague, and go for LDRs.  Especially if you're going to use half-wave rectification.  The absolute enemy of unobtrusive noise gating is a superfast response time.  That' s generally what transistors have and what LDRs lack.  You don't need a LOT of lag in the LDR; just enogh to smooth over the ripple, like smudging oil pastels with your thumb.

When the Gaines noise gate project ( http://hammer.ampage.org/files/GainesNoiseGate.PDF ) first appeared in Modern Recording magazine, subsequent issues contained letters from folks who could not get it to behave well, and Jon's replies to those letters revolved around reducing the glitchiness of the envelope follower.  Some of that involved increasing the value of the averaging cap, or tweaking the decay time, but in truth a good choice of LDR can take care of all of that for you.  Plus, they are quiet and do not distort.

[petey twofinger]

i can not find a vero layout for the SSM2166 chip , i have all the components to built the hotamp gate , hells gate , and the totogate .

the ssm2166 , the one diagram i noticed there was a single pot for the ng . i would THINK having a release control would be extremely important .

this pedal is going last in the chain .

the reports i read on the mxr where it is difficult to get it working properly , then another person said it doesnt do much ..

i also read that some folks felt the dod 230 was very slow , unusable .

which are the reasons i was drawn towards these designs .

thanks for the info to all 

[Mark Hammer]

I am guessing that the release time could simply involve a variable resistance in parallel with the AVG cap on pin 8. Failing that, just use different cap values for different time constants.