Tired with MXR fet NoiseGate? Try this DOWNWARD EXPANDER NoiseGate!

[rogeryu_ph]

Diyers,
Hi, this maybe our solution to getaway with FET type NoiseGate.
I saw this on Mark Hammer website and maybe Mark may give us more on how the circuit functions and the design. I think this is the same circuit with ISP decimator pedal is. I already read a primer on ISP on the website. I'm not 100% satisfied with my clone MXR FET noisegate and Boss NS2 that I bought. Also, I got spare CA3080 for this project.
Paging Mark Hammer................ 
http://hammer.ampage.org/files/Expander.gif
[/url]http://hammer.ampage.org/[/url]

Roger

[Mark Hammer]

Hold your horses, there, pal.  Not all expansion is the same.  The conversion of the Flatline from dynamic compression to dynamic expansion is not the sort of expansion that results in less audible noise.  Those sorts of circuits only apply expansion to that part of the signal below a certain threshold.  The Flatline conversion applies expansion to the whole audio signal, regardless of actual amplitude.  Big difference.  I appreciate the nod, though.

If you DO want a true downward expander, you need look no further than the SSM2166 chip and the Q&D Compressor project at Jack Orman's site (download the datasheet for more ideas).  Most folks think all the chip does is compression, but it does a damn fine downward expansion, and is easily adapted for a top-notch noise-control pedal too.

[rogeryu_ph]

So you mean we can not use this as a pedal noisegate?

Thanks,
Roger

[Mark Hammer]

Correct.

I suppose it might be possible to alter the circuit so that the gain is only turned on when the input signal exceeds some preset level, but that would involve more than the mod I suggested in the posted redrawing.

[MikeH]

So...  Is there a non-fet noisegate project out there that works?  I'm sick of trying to make my MXR work 

[Mark Hammer]

I'm telling you, the SSM2166 makes a DANDY noise reduction unit.  Lots of flexibility in setting the point where it provides downward expansion.  No futzing around with FETs and their ever-bamboozling pinouts.  If you are using it only for noise reduction, you can also play with the averaging cap from pin 8 to set the shut-off time.  Dead quiet.  Jack has a layout, but this is one of those projects where the properties of the chip lend themselves extremely well to perfing.  Really, just a dip in the middle with a passive component here and there going from the chip outwards.

Can't recommend this as a noise-control unit highly enough, so buy 2 chips from Small Bear or wherever, one for the compressor at the front end, and another for the nloise-control at the other end.

[MikeH]

Ahhh... I didn't realize that project was a noise reducer as well as a compressor.  I will check it out.

[MikeH]

Interesting- a couple questions:

1)  What is a "rotation set point"?

2)  I assume the "Noise Gate" control acts like a "threshold" control would, is there any way to mod this design to implement "attack" or "release time"?

[Mark Hammer]

A lot of people didn't realize that.  Wasn't until I downloaded the datasheet ( http://www.analog.com/UploadedFiles/Data_Sheets/SSM2166.pdf ) that I realized it myself.

The SSM2166 is a sort of generalized signal-dynamics controller that applies compression/limiting to any signal above a designated threshold (the Rotation point), and expansion to any signal below a designated threshold.  The Q&D compressor just uses one side of that.  It's your choice to use one side, the other, or both.

It's a good thing that both features come on the same chip, because one of the regular complaints about compressors is the manner in which they tend to increase noise by boosting the signal.  The downward expansion aspect of the chip simply shuts the signal down when the signal is small, killing all that objectionable noise.

Two comments about the chip and the Q&D circuit:

1) It was designed to be a stage of a larger device like a "mic strip", so it assumes the sorts of levels and impedances typical of a voice mic, which generally aren't at all like those of a guitar.  You will want to stick some sort of high input-impedance buffer stage at the input, to keep all that delicious bandwidth you started with.  It also assumes you need more gain than you really do.  You can consider increasing the value of the ground resistor from pin 6 to reduce that initial gain below x2, and make the "gain adjust" resistance a modest value, like 4k7-12k (it provides way more gain than you'd need).

2) There is no volume control on the output and certainly no terminating resistor.  I have no idea if the internal circuitry provides DC blocking of any kind, so stick a 10uf electrolytic cap on the output (+ side to pin 13), and a 10k-100k log volume pot after it.

Optmizing it for downward expansion and noise-reduction involves setting the Rotation point (where compression is imposed) as high as possible with a fixed resistor, eliminating the compression amount with another fixed resistor, and installing the 1M threshold control.  As indicated earlier, you will likely want to provide for several averaging cap values, whether with a 3-position toggle or rotary switch.  Smaller values yield faster turnoff times, and larger values longer turnoff times.  Varying the threshold and decay time will get you most of the control that a great many other pedals provide.  My trusty old grey MXR Noise Gate / Line-driver only had one knob for Threshold.  This one works better.  If you set it for a bit of gain with either a fixed resistor or internal trimpot, and provide an output level pot as described above, you will have the option to set the threshold low enough that you can't hear it, permitting you to treat the pedal as a clean booster which you can crank with the volume pot.

[MikeH]

Of course, also in optimizing for noise reduction, rather than hardwiring the compression and rotation point "off" you could just set the controls all of they way "off" (perhaps using a bit of a "more or less than necessary" pot, just for good measure) so that you have the option to use it as a compressor if need be.

[MikeH]

Actually, how about using a dpdt toggle to turn off the compression function?  I'm thinking one side to switch in a fixed (overkill) resistor in the rotation point adjustment, probably just switching out the 1k for a 100k or 1M, whatever's necessary.  But to turn the compression off would I want to just break the connection at pin 10?  It appears that higher resistance to ground causes more compression, so perhaps just grounding pin 10 would do it?

[brett]

Hi
some smart fellow should be able to adapt the top bit to make a MOSFET "bypass to ground" noise gate.
cheers

[brett]

Actually, it should work if you (i) ignore T1 and throw away cap X, and (ii) connect the drain (top) of the MOSFET to the signal path (RHS of 0.1uF cap).  To the right of that connection, all you need is a volume control (e.g. a 100kA pot).

The circuit works by turning the MOSFET off (and letting a signal pass) when a signal is present.
cheers

[DimebuGG]

I've heard rogeryu ph claims that this works better than the MXR noise gate and BOSS NS-2 as he tested it.

[rogeryu_ph]

Guys,
Yes it's verified,  sorry for the delay i'm so busy at the office finishing all delivery co'z tomorrow's holiday. I'm also having difficulty uploading to PHOTOBUCKET the picture of downward expander breadboard that I have finished. I benchmarked all with the same set up with my TScreamer and AMP. First on the FRONT using single coil guitar-noisegate-TS-AMP. TS set to full drive and AMP loud enough to hear the noise. Result.: Expander- just minimum threshold good fade at the end no trailing noise, guitar sound one is to one when on and off
Boss- full threshold with minimum or considerable stuttering or trailing noise, guitar sound one is to one when on or off
Mxr clone- 95% threshold with no stuttering but lost in volume, guitar sound not one is to one when on and off.
Now I put the noisegates at the BACK chain Guitar-TS-noisegate-AMP.
Result: Expander - max or full in threshold with minimum stuttering when fading but very considerable.
Boss - full threshold loud hum can not tame the noise defintely failed in test it is only good at the FRONT of the chain.
Mxr clone - full threshold hum still audible but not louder than Boss, guitar sound fairly low

My Rating
Mxr 80% due to volume lost
Boss 85% good on FRONT
Expander 95% but not 100% or perfect.

I'm not saying this is true to all MXR clone co'z this one is my own built. And i'm not saying Expander is superb but BEST in compared to other.  Boss, I say BETTER.

If I got time, I like to post these CLIP on Youtube so I could link it here.

BTW,  If I could also get an SSM2166 chip i'll go give also a try. Maybe this is the ultimate one,i'm still on quest for a perfect NOISEGATE. 

Thanks,
Roger