looking for an extreme (noise) gate.

[any]

Hi all,

I've done a search on the subject but it either turns up nothing or 35 pages worth of threads...
I'm looking for a (noise)gate or audio gate that can be set very precisely/extreme on the threshold.
What I'm  aiming for is an effect that will only let the notes initial "attack" through and kill any sustain or decay.
This way you would only hear little "blips", they would be clean though... no distortion.
I've tried a 19" behringer noise gate and it could get this effect pretty good, so I'm guessing it's possible.
Is there a schematic that can be adopted for such extreme gating?

Thanks for any info in advance!
ANY (a.k.a whythisreason?)

[davebungo]

envelope follower->comparator->CMOS analog switch

[any]

would such a design still allow for adjustment in the attack/decay or would this purely give an on/off approach?

[David]

Tonepad has a project for a noise gate.  Run this into a compressor/limiter that's set at zero or near-zero sustain.

[StephenGiles]

Drawmer DF320 will do that!

[any]

I know some rack compressor/limiters noisegates do that...
But would it be very hard to mod an existing circuit to do this AND fit in a 1590b? 
(a very quick gate with precise control over the threshold)

[Mark Hammer]

Buy a BBD chip for starters.

If you are working in the analog domain, one of the problems you are stuck with when it comes to lightening fast responses based on sidechains is that sidechains take a little time to do their work well.  And if what you are trying to alter is, itself, a brief event, then the sidechain may not have enough time to produce the desired effect.  (Here I am reminded of the classic Monty Python bit called "The Bishop", wherein a scarred tough-guy crime-fightin' bishop and his priestly henchmen - all in shades and cassock - always show up about 2 seconds too late to prevent the crime from happening.).

That sort of problem did not escape Electro-Harmonix and the old Attack Delay pedal involved introducing just a slight amount of delay on the signal after the envelope was extracted so that the envelope parameters to be imposed could be imposed at the true start of the note, rather than a couple of milliseconds after the note had already started.

The other thing your stated objective has to grapple with is that gates generally remain open for a little while after the initial transient.  It's a bit like trying to decide whether you ought to tell the remining customers that its 6PM and you'd like them to leave the store.  If you don't know how much they're going to spend in the next 5-10 minutes, do you really want to toss them out, or will their expenditures justify you remaining open (for them) another 5-10 min?  Your noise gate has the same sort of decision to make.  What you have indicated you want it to do, however, is to say "I'm taking the first 5 customers, and then I'm closing my doors, no matter who else is out there". 

Though the initial transient is clearly at a higher amplitude than the rest of the signal, the signal will still remain high-ish for a little while after that transient, such that whatever turn-on threshold is needed to capture the attack of the note will always be too high to permit the gate to shut off until some time after that note has been around.  So, at a more technical and signal-oriented level, you are asking for the gate to accept the initial transient by opening on the basis of a low enough threshold, but then shut off by using a different threshold somewhere between the one used for turning the gate on and the peak level attained by the transient.

I'm thinking, then, that maybe what you really want is a kind of one-shot pulse that serves as quick envelope generator.  This should, in principle, ignore what the actual signal level is at the moment, and just provide a fixed attack/decay duration. In which case, you want to use something that generates a trigger pulse to actuate such an envelope generator.

The recommendation to slightly delay the audio signal for a couple of milliseconds still holds since triggering takes time and you want to hear the initial attack not produce accelerated "Slow Gear" effects.  If the trigger extraction starts a few milliseconds before the sound comes out of the circuit, then when the envelope generator kicks in both note and envelope will start at the same starting point.  You can find examples of some of this circuitry in the first couple of issues of DEVICE at my site. (http://hammer.ampage.org)

[any]

Thanks for your thorough reply, I'll have to read and reread that article since my skills aren't yet up there... 
But I do understand the basics of what you were explaining, and It would indeed get me what I'm after.
I just can't derive from the Device No.1 article (gate/trigger extractor and envelope follower) if the circuit will be relatvely small/manageble or if it's going to be a somewhat complex build. But first I'll read myself into it some more!

[Mark Hammer]

The envelope generator in subsequent issues is certainly small enough (you'll need a 555 timer).  For trigger generator there are plenty of simpler circuits around, usually described as envelope-follower/trigger modules (I think there is an older PAiA one posted at AMZ).  Where they vary is in how reliably they work.  In most instances, you'll have to adapt your playing style to them.

Another approach to consider is the "String Damper" project posted on my site, that I scanned in from an old E&MM article that either RG or Mike Irwin sent me.  This has both attack and variable decay.  It should be able to achieve short blips.

Incidentally, if its strictly guitar that you want to do this to, a simple piece of foam rubber pressed up tot he underside of the strings near the bridge ought to do it.  I won't say that it get the specific sound you're aiming for, but it may well make a more conventional noise gate behave the way you're intending by creating a faster decay and a much greater contrast between the transient peak and the level the signal drops to only milliseconds later.  No that I think of it, that may well get past some of the problems I noted earlier.  Remember, the challenge is to answer the question "How do I make a gate open up in response to a low enough signal that I capture the onset of the note, while closing in response to a signal level that wants to stay much higher than what it took to open the gate for too long a period of time?".  If the foam (a trick used on older Gretsches and basses) can dampen the strings enough to bring the average level down quickly, then you can coax the gate into closing more quickly wothout having to resort to fancy-schmancy circuitry.

So, don't build anything just yet.  First, google up "foam bridge mute" and try out the foam solution.  For all I know, it might do exactly what you want by itself.  Alternatively, it might provide the appropriate physical signal conditioning to allow completely unmodified circuitry to do what you want.  I should point out that most gates will permit the user to vary release/decay time so you don't necessarily have to be hemmed in by the ultra-short "bink" of foam pressed up hard to the strings.  The key thing is to find a way to bring the signal level down quickly enough so that ONE threshold setting works for both onset and offset of the gating.  You can always vary how quickly the gate shuts off.

[davebungo]

If you use a fast enough attack, a slower decay and a variable threshold on the comparator (which should have a certain amount of hysteresis about the threshold to prevent "thrashing"), you should be able to get just what you have asked for.  You could do it wth an op-amp, a comparator chip, a CMOS analogue switch and a few passive components.

[whythisreason?]

I think the Compander/cmos switch route looks tempting and will make me learn the most,
and the string mute isn't feasable for me. (too many guitars, and I can't switch it quickly enough)
I'll do some more searching though now I know what to search for... 

[StephenGiles]

I know some rack compressor/limiters noisegates do that...
But would it be very hard to mod an existing circuit to do this AND fit in a 1590b? 
(a very quick gate with precise control over the threshold)

Forget 1590 size - surely excellent electronics is more important!

[MetalGuy]

The answer to your problem /or at least the closest DIY one/ is EPFM's Noise Gate project. I can assure you it can do what you're asking, even more - it can completely silence your guitar. Also it fits easily into 1590B.

[any]

Well I'm going through a truckload of previous threads/posts and it seems like
it's not as easy as I thought.

The EPFM noise gate is not readily available as a schematic, wich makes it a bit of a leap
to buy the whole book in order to find out if it'll work for me.
The PAiA Gator doesn't seem to get to much of a praise performance wise, as well as the
"string damper" by Paul Williams. Those seem to be about the only entries I've found
that employ similar circuitry.
Reading the String damper article I would be really cool if it worked well... if it worked well...

I'll read up some more on what to do with comparators and how to wire it up in a simple
circuit with it.

[whythisreason?]

I've got some LM339N's laying around wich are Low Power Low Offset Voltage Quad Comparators
Is there a way to hook up a simple breadboard setup to see what I can do with this?
The datasheet shows a truckload of possible basic functions wich kind of overwhelm me...
Should I look at the "non inverting comperator with hysteresis"?
Help is much apprieciated, I'm still learning... 

[any]

I've got some LM339N's laying around wich are Low Power Low Offset Voltage Quad Comparators
Is there a way to hook up a simple breadboard setup to see what I can do with this?
The datasheet shows a truckload of possible basic functions wich kind of overwhelm me...
Should I look at the "non inverting comperator with hysteresis"?
Help is much apprieciated, I'm still learning...   


Through a mistake I made 2 accounts at the forum (one at home, one at work...)
I'll lose the whythisreason from now on and only continue with ANY, so that all of you know....

[davebungo]

If you're taking the envelope follower->comparator->switch approach then you will need some hysteresis around the comparator threshold otherwise the output will toggle up and down very quickly as the signal goes through the threshold if there is any noise or ripple present.  Non-inverting will be fine, you just need to ensure that the signal to the analogue switch has the polarity you need.  You could operate the analogue switch in a shunt mode i.e. used as shown below: so when it is gated in it is effectively open circuit:  You wouldn't even need an op-amp to get something to work.

audio in>-------/\/\/\/---------------------------------->output
                                  |
                             ___|_____
                            |              |
                            |   switch  |<-----control
                            |________|
                                   |
                                   |
------------------------------------------------>gnd

[any]

I'm not quite getting my head around this yet...  Anyway, I am having fun exploring though...
I was toying around with the PWM from circuit snippets and found you can use pin 2 on the schmitt trigger
to light a led when you play. (or any of the spare triggers if connected)
Could I use this set up to switch? (and have control over decay time?)
I'm still in the dark about how to wire up an LM339 to an envelope follower, are there examples floating around to get started?

My goal is normal attack that cuts off to silence after about .1 / .2 seconds but attack and decay control does sound like fun
and makes it more useable. It's just that up untill now all reading I've done one such effects didn't prove that satisfying...
(string damper/EH attack/decay/PAiA gator to name a few)

btw: how can I get a page in the layout gallery? I want to post some stuff on the "mayhem!" effect I did. (layout/pcb/partslist)

[StephenGiles]

http://www.morleypedals.com/gdbes.pdf

Have a look at this, or this

http://www.silonex.com/audiohm/pdf/expgate.pdf