Simplest Noise Gate layout?

[mr.adambeck]

I'm looking for some simple noise gate layouts. 
The MXR Noise Gate seems completely out of my league.    Anyone have anything easier?  I did a search but I can't seem to find any easier options.  I'd definitely prefer vero, but perf is fine too - however I'm not setup to make my own pcb's.

In case it helps to know why - I have a circuit bent casio that gets a tone of hiss, and a piezo mic'd floor tom that I want to only hear the tom on (it often feedbacks and picks up a lot of the other instruments).  Neither of them needs anything too fancy at all, I'm just trying to clean up the noise floor at band practice a tiny bit.

[anchovie]

The simplest noise gate in the world is the pair of Germanium diodes after the first op-amp stage in a Boss HM-2. Peavey use a similar setup in some of their amp heads - the JSX head has silicon diodes and a 1M pot wired as a variable resistor in parallel to act as a simple threshold control.

[deeleyer]

I was looking for something similar to put in a diy amp.
I saw that gate from a Peavey JSX but looking at the scheme a question came: diodes will make clipping too?

[kupervaser]

Try this:


Here is the link for the layout:
http://www.newtone-online.nl/forum/index.php?action=dlattach;topic=2590.0;attach=3524

[anchovie]

The diodes in this simple gate are sometimes known as coring diodes because they cut out what is within the threshold of the diodes, whereas if the pair went to ground they would clip the signal that goes beyond the threshold. So in the case of the HM-2, any hiss or hum that is within +/- 0.3v will not pass through as the diodes will fail to conduct. The amplified guitar signal will get through easily.

[earthtonesaudio]

I believe the series-connected antiparallel diodes will cause crossover distortion, but reverse leakage might help reduce it somewhat.  That's probably part of the reason for using Ge.  If you just cared about the low forward drop, Schottky diodes would be preferable, because their forward drop is much more constant over a broad range of signal currents.

[Mark Hammer]

That is correct.  Diodes in series square up the waveform at the sides, leaving the top intact, whle diodes to ground square up the waveform at the top, leaving the sides intact.  The HM-2 uses both in order to produce a VERY square wave.

Although diodes in series DO possess the property of blocking any signal passage below a certain amplitude, they have two major drawbacks:

1) Diodes have absolute thresholds, not continuously variable ones.  You could use a series of Schottky diodes and a switch to essentially "step" between a series of thresholds, but that's cumbersome, and cumbersome is the opposite of what you were looking for.

2) Diodes will produce clipping, but more importantly, not necessarily the type you were looking for.

The general topic of noise reduction comes up here again and again.  And each time, my best advice is not to address noise at only a single point of intervention - like a noise gate at the end of the signal chain - because it will disappoint you every single time.

So, before we take another step, what exactly IS the nature of your noise concern to be addressed?  There are many noise scenarios that are addressable via the same strategy, but no single strategy works in every case.  So, what's your scenario?

[dschwartz]

i´ve always thought that antiparallel diodes in series with a pot is a great fix for dr boogie´s hiss when not playing.

of course there´s a tradeoff between crossover dist and floor noise, but dialing the pot should get you to a nice sweetspot.

maybe the crossover dist may add little more realism..since AB tube amps generate some crossover dist when cranked loud.

[DDD]

Couple anti-parallel diodes in serial do excellent noise reducing in hi-gain circuits. If the gain is lower, you hear the sitar parody.
At the same time even with the hi-gain stages the diodes produce weird-sounding and unnatural 3-rd harmonic at the release phase of the sound, which is too much audible.
I've tried 1n4148, Ge diodes as well as the Shottky 1N5817 with the similar influence on the sound. +
Also, the power chords sound much harder and "angry".

[MikeH]

Mark, this is where you usually bring up how great the Q and D compressor works as a noise gate 

It is rather simple to build, and the dual functionality (noise gate/compressor) is a nice bonus.  I never actually boxed mine up, so I don't remember, but did the Q and D do the Noise Gating after the Compressing?  I think it must; bad design otherwise.

[Mark Hammer]

I heard my train a-coming.....

The SSM2166 (and Q&D derived from the appnotes) works well because the envelope used for both compression and downward expansion (NOT gating) is derived from the untampered input signal, where the dynamic range would be greatest, and the selectivity of the thresholds set greatest.  You are quite right in thinking someone would have to be nuts to compress and then gate.

As to how the envelope is applied in a strict electroic sense, you've got me there.  My sense is that the same envelope voltage is distributed to two points so as to effectively tell the circuit "Here's what you do with any signal moments/segments above this value, and here's what yo do with any signal moments/segments below this value."  They may not be sequential in any strict sense of the word.

Certainly cleaning up noise at the outset, before any subsequent high-gain device has a chance to boost that noise, makes it a piece of cake for anything way downstream to clean up whatever noise remains before the amp's input jack.

[mr.adambeck]

wow.  A lot of posts on this.  However, as my initial post said, I'm looking for some non-pcb layouts...  Does anyone know where I could find some?

[edvard]

Non-pcb?
As in dead-bug or point-to-point layouts?

[anchovie]

Non-pcb?
As in dead-bug or point-to-point layouts?

Vero or perf. They don't count as PCBs because you still have to do some work with them to complete the circuit.

[mr.adambeck]

Yeah, vero or perf.  I'm not setup to make my own etched pcb's.