Noise Gates

[craigmillard]

Hi Guys,

Im looking into noise gates and came across:

http://www.thatcorp.com/datashts/dn100.pdf

its an application note for some of the THAT chips for a full fledged noise gate.
It looks quite similar to the ISP Decimator, how have they got a patent from essentially an application note? anyways..

Anyone had ago with this? thoughts?

[armdnrdy]

If you are referring to the That Corp. drawing that you linked....I don't see mention of a patent.

I see copyrights for the drawing...which is standard.

[craigmillard]

Hey armdnrdy, I was referring to the ISP technology decimator line of noise gates which have patents..

[armdnrdy]

Got ya. 

I looked into it a bit....don't know! Unless they designed the That IC.

If you look at some of the BBE Sonic Maximizer designs....BBE actually designed the ICs.....

NJR manufactured them.

[craigmillard]

Looking into it a bit more the THAT chip contains a RMS systems to detect the levels while ISP have some propriety system using discrete components. I believe this is what is patented. ISP still use the VCA part of the chip though just in a standalone package the THAT2180.

Anyways im going to look into designing a gate with the THAT4301 chip as looks like it contains almost all you need! I have also just found Marshal use this in there 2203KK and 401HJS Amps as there built in noise gates!

https://www.dropbox.com/s/s37pv48o4lqzj7o/JVM410HJS%20Main%20Circuit%20Diagram.pdf?dl=0

Do you guys think its worth implementing the Release and hold parts to the gate as per the application note?

Im thinking of starting from the Marshal design and working out from there..

 

[PRR]

> how have they got a patent from essentially an application note?

1) If you pay the fees, you get your patent. Today's patent "examiners" don't have time to examine patents. (This has raised a bit of a fuss among patent geeks.)

2) Read the "I Claim"s with a very fine-tooth comb. Most patents are "improvements" on existing inventions. The first part of the patent covers the general field, to lay the ground for the "improvement". Then the improvement is informally described. Finally the "I Claim"s specify the specific improvements you are paying to have documented for potential infringement fights. Well-written "I Claim"s are intended to befuddle the reader and leave room for claiming improvements that the inventor hasn't really thought-up yet. Remember that ultimately the infringement law suit goes to a Judge who may not be dumb, but can't possibly be expert in ALL the inventions which come to his bench. (The AT&T and Triode patent fights are instructive.)

Say I patented the wheel. Then you patent the axle. Larry patents an iron band to improve wear. Dunlop patents the inflated rubber-tired wheel. Somebody probably patented the wheel with the hubcap which keeps spinning after the wheel stops. Larry might have to talk to me to use the wheel, but Larry can't stop Dunlop from making wheels with rubber tires. Unless Larry hinted that other tires might be mounted over his iron band, and Dunlop found he needed some such thing to hold his rubber tire.

> looks quite similar to the ISP Decimator

The basic noise-gate is very old. A simple design works but with flaws. The THAT design works in log-mode, which makes knob-taper easier, and does a NLC trick. Much smarter noise/signal discrimination is possible, though it would take a ton of work to hit on a scheme that was clearly "better" in all situations. ISP may be using a micro-computer? You could rig it so if it muted when it should be on, or was on when it should be off, you *kick* it and it remembers your preferences. Or it could auto-correllate the input, assuming that music is less random than stray noise (but good buzz is quite non-random). In any case you need a VCA and might enjoy a packaged detector. The THAT parts are well-characterized and fault-free, so might logically be the basic workhorse of a noise gate. I'm not inclined to wade through ISP's patent wondering what little bit they "Claim".

[PRR]

> worth implementing the Release and hold parts to the gate

You have to have a RELEASE. If release was zero it would cut-off the tails of your notes. In this circuit without the release parts I think it would stay on "forever", which is useless.

The Hold, I've never had that, but it does address one popular shrotcoming of noise-gates.

As the whole extra she-bang is $2 of parts plus 2 knobs, I don't see any reason to omit it for a test-build. If, after extended real-world testing in many situations, you decide it is useless or may be left fixed, that's your call.

A useful noise-gate is not that hard. A perfect noise-gate is impossible. THAT's design leans to smart user-controls for fine-trimming more things while in use.