breadboard report : DOWNWARD EXPANDER NOISEGATE

[rogeryu_ph]

Guys,
First of all, I like to quote that this DOWNWARD EXPANDER is not the ultimate solution to your noisegate requirement This may not suit your needs maybe because we have our own individual set up, gears and sound system. I benchmarked and compared it with my BOSS NS2 stock and my own built MXR noisgate.
I used it at the back of the chain going to AMP and set it with moderate noise level
GUITAR single coil -> Tscreamer -> Noisegates -> 65w AMP
Here's the schematic from hammer.ampage.org
http://hammer.ampage.org/files/Expander.gif
Here's the breadboard using NPN 2n5088 and PNP 2n3906 with Zener 1n5228
http://i186.photobucket.com/albums/x139/rogeryu_ph/DSC04714.jpg
Test video
http://www.youtube.com/watch?v=rOL67HcKtik

I Rated :
expander as 95% good in front and back, Can tame higher noise level compared to other two noisegate and guitar volume is 1:1 both switch off and on.

NS2 90% which is good in the front but not at the back (I don't understand why boss can handle the noise in front but not ont the back of the chain! ) Guitar volume is 1:1 both switch off and on but can not tame moderate noise level.

and 85% for MXR my own built. MXR can tame a moderate noise level but has loss in guitar volume when swicth on

still not 100% satisfied due to increasing farther the gain and amp volume expander can no more tame the noise.

Guys, is it possible to cascade both the MXR circuit and the expander circuit into one circuit with SEND and RETURN feature like BOSS NS2 but using two threshold control? Buffered output of the guitar then fed to both circuit one is for the guitar control threshold sensivity and the other control is for SEND and RETURN  pedal chain. This is because if we can tame first the guitar noise this will not aggregate more noise to the   gain noise of the pedal and controlling is far smoother.

Here's a ISP DECIMATOR Pimer which will give you an idea. I'm not saying the same circuit or technology but the same semantic.
http://www.isptechnologies.com/DECIMATORTECHNOLOGYPRIMER.pdf

One also option is this:
http://www.muzique.com/ssm2166.htm

I'm still on the quest for an ultimate noisegate

Thanks,
Roger   

[Paul Perry (Frostwave)]

Maybe the ultimate would be a multi-channel downward expander array.
It's not going to be cheap, though.
Maybe a graphic EQ unit could be converted - but you will need the downward expander for each channel.

[kugua]

very nice test, I wanna make the isp noisegate too.but  has no schematic,I  show you the isp guts shot.got it on the web,

so I have one SSM2166,five years ago i make this one(the schematic has used by muzique web),but i dont test it ,and in that time i dont have enough knowledge to learn .maybe test it last week? hehe

[rogeryu_ph]

Good idea Paul though very expensive. I was thinking of two threshold control by integrating both MXR and EXPANDER circuit into one. Is it possible to put SEND output on MXR circuit going to PEDAL CHAIN and RETURN in EXPANDER then merge its output. MXR threshold will take care of guitar and EXPANDER threshold job is for gain pedal noise?

Finally a mythbuster, Someone has SSM2166 Give us feedback on your test Kugua

Thanks,
Roger

[rogeryu_ph]

The other thing is that there is a difference between a noise reducer and a noise gate.

A noise gate operates on the principle that when accompanied by signal, whatever ongoing noise is mixed in with the signal will be barely noticeable and indeed "masked" by the signal.  Once the signal level goes below a certain minimum level, however, the noise becomes very noticeable.  Consequently, the circuit completely shuts/blocks the signal from passing through when it goes below a certain preset level, and opens up again (i.e., lets signal flow through) when the signal once again reaches the required threshold.

If there is a huge level difference between the actual noise floor and the "true" signal, and if the signal envelope is just right, this approach works relatively well.  Sadly, not everything in life is a mic'd kick drum (where the signal comes and goes in short, high-contrast, bursts), and this approach can be a problem for things like controlling hum from a single-coil guitar for someone who likes to play like Mark Knopfler and has many notes that need to sustain for a while or are lightly picked.  A noise gate will either let far too much noise through in this instance, or if you set it to eliminate the noise, it will chop off the beginnings and tails of notes in an irritating way.

This leads to a slightly improved approach, which is what's officially called "downward expansion".  This is, in a way this complete opposite of what a limiter does.  Where a limiter works to *reduce* any volume differences between signals above a certain level (i.e., the dynamics of everything UP TO that level are not affected; only what is above that level), an expander works to *exaggerate* volume differences of everything below a certain level.  So, it doesn't really shut anything off or on, but as the signal level subsides at the end of a note, and the note+noise goes below a certain threshold, the circuit makes it sound MUCH lower than it is, and the extent to which it exaggerates the volume drop increases as you go even lower.  This manages to preserve the start and end of softer passages a bit better than a gate, though it still isn't perfect.

A third strategy is a "noise filter".  This operates on the premise that the irritating noise accumulated in the signal path is going to be primarily hiss.  Liek a noise gate, whent he signal level is high enough, you don't really "hear" the hiss, but as the signal level drops, the hiss becomes more irritating.  To combat the hiss, but retain the signal dynamics and not forfeit the onset and tail of the notes, a noise filter uses a variation of an "autowah" circuit to trim the upper treble off as the note decays.  So, it applies a treble rolloff during softer passages to make the hiss less noticeable.

More complex versions of this combine the one-two punch of a downward expander and a noise filter to both downplay the parts where the "signal" consists primarily of diverse noise sources, and turn the treble down.  This is actually a bit better, since the noise can be both hum AND hiss.  The noise filter is fine for rolling off hiss, but since it starts from the top of the frequency spectrum, it does nothing for the hum way down in the lower regions of the spectrum.  The expander/filter combo helps to attack all sources of unwanted noise throughout the spectrum.
Quote by Mark Hammer
Mark, Now that I have an EXPANDER, what Noise FILTER can you suggest for me to combined with? BTW Are you saying separate circuit or one combination circuit only when you said "combine the one-two punch of a downward expander and a noise filter"

Roger

[Mark Hammer]

If you stick a noise control device at the very beginning, it cannot do anything about the accumulated hiss over your pedals because that hiss comes from the pedals.  If you stick a noise control pedal only at the very end, there is usually a) an amplification of whatever hum and other noise was introduced at the beginning of the signal path that raises the "noise floor", and b) a reduced dynamic range from all the pedals along the way (compressor, distortions, etc) that makes it harder to pinpoint the appropriate threshold that separates the noise floor from the actual valid signal without letting noise through or losing musical signal.

What I'm saying is that if you have a noise gate early in the signal path that is set to deal with hum, then when you have a noise filter near the end of your signal path it does not have to act quite so brutally.  The problem most people have with noise is that they use one solution in one place to deal with it.  That forces them to set the controls in a way that produces all the negative aspects of gates and filters (loss of attack or decay, and significant loss of high end).  The reality is that noise comes from multiple sources and requires multiple solutions.  So, what I am saying is "Divide and conquer".  Use of two noise solutions allows each one to be set with greater subtlety.  Hum at the point immediately after the guitar is as low as it is going to be, and dynamic range is as high as it is going to be, making it easy to identify the best threshold for gating out hum without affecting anything else.  If hum is gated out at the beginning, then it will not be amplified by your pedals.  Any hiss from the guitar will not be amplified either, making any hiss at the last pedal almost entirely pedal-generated hiss.  Of course, you still have the issue of reduced dynamic range at the last pedal.  The answer to that is to have the envelope signal that control both the first noise-control device and the second one be derived from the direct guitar signal.

It would be sweet if someone made such a pedal.  Plug your guitar into it, and run a cable to your amp from it.  It has one control for setting hum-gating threshold at the front end, and another for setting hiss-reduction sensitivity at the output end, both driven by a common sidechain.  In between you have a buffered loop send for all those TB pedals and a return jack.  Master bypass is optional.

[rogeryu_ph]

Mark,
I did what you advised. Boss NS2 at front with moderate threshold for guitar control then I used the send to all dirt box last is the EXPANDER back to Boss return and out to Amp and I'm very very much satisfied Combination of both punched all the hiss and hum even I turn volume full the amp and pedals is dead quite. Decay I could control from my Boss and still has very good release, sustain and feedback. Thanks all for your kind support and long live HAMMER.AMPAGE.ORG

Roger

[Mark Hammer]

Proof of concept! 

Like I said, divide and conquer.

R.G., if you are reading this, think "signal hygiene".

[Ardric]

It would be sweet if someone made such a pedal.

Mark, ISP makes one: http://www.isptechnologies.com/decimatorGpedal.htm

[Mark Hammer]

Close, but not precisely what I had in mind.  It is certainly a step up to detect the envelope at the start of the signal chain and apply whatever the threshold is at that point to other places in the signal path.    What I had recommended, however, was two stages of noise reduction, one oriented towards hum at the start of the signal chain, and another oriented towards hiss at the end.  But you are correct that the ISP pedal comes closer to that ideal.

[raulgrell]

That's actually a project I've been meaning to look into, and your idea satisfies exactly what I need! I'll work on some designs and see if I can make something like you said.

I think it was you who was talking about a single chip that is overlooked in guitar effects that has an expander, a noise gate and a compressor, or something along those lines... What chip was that?

[rogeryu_ph]

Yes Mark,
DIVIDE and CONQUER the only solution I made. One threshold for the front and other threshold for the back end. 
At last, my quest for noise gate has end. 

Roger

[Mark Hammer]

That's actually a project I've been meaning to look into, and your idea satisfies exactly what I need! I'll work on some designs and see if I can make something like you said.

I think it was you who was talking about a single chip that is overlooked in guitar effects that has an expander, a noise gate and a compressor, or something along those lines... What chip was that?

The SSM2166 is designed to provide dynamic control at both ends of the amplitude spectrum.  At the low-amplitude end, it provides downward expansion that reduces noise (it is NOT a noise "gate"; common misconception).  At the higher-amplitude end, it provides limiting/compression.  The overall dynamics of your output signal will depend on where you set the point where the downward expansion kicks in and where the compression/limiting kicks on.  Do yourself a favour and look at the datasheets for the chip.  VERY instructive.  http://www.analog.com/UploadedFiles/Data_Sheets/SSM2166.pdf

[raulgrell]

Merci, mon ami!
I'll look into the circuits using that chip, try adapt them to guitar. Try some hacks to see if I can come up with some nice effects.

[Ardric]

How about this rack mount one then?  http://www.isptechnologies.com/decimatorprorackG.htm  I'm pretty sure there's two VCA's there:



I haven't used this and can't comment on how effective it works, but I'm just pointing out that it's available.

[Mark Hammer]

After reading their blurb, yeah, that IS pretty much what I'm talking about.  Nice to know someone's on the right track!   Of course, it would also be nice to have a stompbox-sized version of same.

[km-r]

Built it!
very nice noisegate without the fuss with FET's....
slight volume "swell" on some settings of the threshold. roger_yu suggested a mod to cure this.
adding a 1.5v-3.0v zener  in series with the second diode from left [refer to schem]. cathodes together.

i did not solder in the 5k trim to get full volume from the circuit, it is slightly higher than dry signal.

used all 4.7uf electrolytics instead of 3.3uf
used 4558 instead of TL061 with satisfactory results. [4558 has different pinout]

[rogeryu_ph]

I used 1.5 volt just enough for attack and 100k log pot for much control threshold. I already boxed it up and tried in our practice last night. I'm so much satisfy.... NS2 front and Expander back chain going to AMP 

http://community.epektos.com/photos/rogeryu_ph/picture543.aspx

Thanks,
Roger

[DimebuGG]

[rogeryu_ph]

Nice PCB Dimebugg,
Regarding the mod, add a series 1.5v zener after the (470ohms and 4.7k resistor with signal diode) junction and before signal diode with circle T and circle C. The zener should be anode pointing to TR3. This help to smooth threshold gated sound and decay more natural. Depend on your ear just experiment from 1v-4.5v zener. This is one good if you built a project by breadboard first and later mod and tweak to your taste 
BTW Km-r, I haven't yet started the INSANITY distortion project which we both plan to work next... I've been busy and crazy nowadays and having fun using and recording with my new GT10 MFX 
 

Roger