Not understanding Boss NF-1 noise gate schematic ( wrong schematic corrected)

[Max999]

Update: I provided the wrong schematic at first, an NS-2. I have changed it to the NF-1.

I don't understand how the actual gating is being done. Is it Q3 used as a jfet switch that grounds the audio when it closes through the 2.2uF cap above it? Or does is has to do with the emitter of Q2 being connected to the base of Q3?

I have tried to LT Spice it but it has limited use because I don't know where to set the trim pots.

I see Q1 is a buffer, which splits the signal. The sense pot controls how much signal is fed to IC1, which is a non-inverting amplifier with 62dB(!) of gain. From there on I am lost. I didn't got very far eh 

Who sees more?

[PRR]

IC1 is a voltage controlled chip. Since no type-number is given, I dunno the details.

I do not see where you find 62dB gain. IC2a does give up-to 73dB gain, so it can bring-up small hiss to the detection circuit.

[amz-fx]

IC1 is an M5207L01

regards, Jack

[Mark Hammer]

It's one of those chips that are probably not impossible to get, like the BA662, if you know the right folks in the Akihibara.  But, like the BA662 is pretty well next to impossible to get.

[amz-fx]

Also, isn't that the NS-2 schematic and not the NF-1?

regards, Jack

[Max999]

Uh oh .. that is very stupid of me. It is indeed the NS-2 schematic, here is the NF-1.

[PRR]

[Max999]

Wauw PRR that is beautifull! Once again you delivered something nice to a soul that is blind where you can see. Thank you for being on this forum, I really appreciate that you took the time to edit a schematic for a fellow forumite!

Ok, so where I am now with the NF-1 is: Input gets buffered, branch goes into high gain amplifier, ac gets rectified into dc, which is split: one branch is fed to the gate of Q3, another branch goed through the decay knob, is being trimmed down and then applied to the source and drain of Q3.
Q3 is used as a Jfet switch, and it is open when the gate is at less voltage than the source. The 470K resistor next to Q3 is then out of the circuit, so all ac flows to ground through the capacitor with unclear markings. With Q3 open, there is still 4.5% attenuation through the 22k/470K voltage divider.

So far so good?

The source and drain of q3 are also fed with a 9v voltage that goed through a 3.3K and a 4.7v zener, D3.
So there must be about 4.7v coming from that, which function does this have?
And what is happening at the emitter of Q2, why whould it be usefull to apply the "control voltage dc" there?

[rankot]

It's one of those chips that are probably not impossible to get, like the BA662, if you know the right folks in the Akihibara.  But, like the BA662 is pretty well next to impossible to get.

I found this interesting source of BA662 clone: http://wiki.openmusiclabs.com/wiki/BA662

[PRR]

> So far so good?

Good to a point. I think you should study simpler circuits a while. There's some subtle points in this one.

> must be about 4.7v coming from that, which function does this have?

To turn the JFET off we have to pull the Gate "negative". But there is only a positive supply. That would be awkward. However we can stand the JFET up on a 4.7V point, and pull the Gate negative of 4.7V.

> emitter of Q2, why would it be useful to apply the "control voltage dc" there?

You need to look at impedances. How much stuff flows?

Look at my town. All the roads are "connected", but some matter much more than others. Beer Highway has 1000 cars an hour 3:30-4:30. Cross Road maybe 100 cars in the same hour. Bear Poop Trail gets one walker a week, maybe. Road traffic planning and backups are all about the big high-flow roads, nobody cares if Bear Poop Trail is washed out.

Yes, control voltage "does" flow through two 1Meg resistors to Q2 Emitter. Which is probably under 100 Ohms impedance. That's like the trickle of traffic coming over Bear Poop Trail onto Beer Highway. The folks on the highway don't even notice. What this really is, is a trick which reduces distortion in Q3. The circuit will work fine without it. I'm not sure it is even audible in a noise gate; it matters more in a Limiter.

[amz-fx]

With Q3 open, there is still 4.5% attenuation through the 22k/470K voltage divider.

I didn't calculate the amount but you are right in that there is always loss in the 22k/470k resistors. There is also a small amount of loss in the 1k/220k on the input. Since the signal path is always active in the original pedal, the small losses are not too noticeable, but if made true bypass, it could be more apparent when switched off.

Attenuation depends on how much resistance is in the jfet when it is fully on. Q3 is a 2SK30A-Y in the original Boss pedal, so the datasheet might tell us what the minimum resistance is. (I took a quick look at one that 2SK30A datasheet that I found and it did not specify the resistance). The attenuation will be from the divider formed by the 22k/Q3 resistances. A wild guess is that it will be 40dB or more attenuated.

regards, Jack

[Mark Hammer]

A friend just gave me one of these yesterday.  I'll have to tinker with it and see if there are any usable mods.

[Max999]

@PRR: I am in way over my head with this one, I agree.  The thing is that I want to break free from the booster/buffer type circuits, and I need a noisegate. I have looked at multiple schematics and this one seemed to be the one with the least parts. Thank you very much for your explainations and analogy, they have been really helpfull.

@Jack: thank you for looking at the datasheets for me. I am planning on changing the 220K's on the inputs to 1Megs, to get that attenuation out of the picture.

At everybody:
If I want this NF-1 to run at 18v, my Spice simulation sais that everything will scale up by a factor of 1.94 around Q3 without ac input to the "control voltage circuit".

9V - Base at 925 mV
       Source and drain at 1.49V

18V- Base at 1.79V
        Source and drain at 2.89

The problem I am fearing though, is that the control voltage isn't strong enough anymore to turn Q3 on and off. Is this correct?
If this happends, is it possible to modify the voltage by turning the VR-3 trimpot?

[PRR]

> If I want this NF-1 to run at 18v

Why? Have you built it? Are your signals so hot they clip on 9V?

[Max999]

It does not clip right after the guitar, but used later in the chain after a boost it clips merrily. 18v operation would be a lot more flexible.
I have an actual NF-1 in my possesion, and would like to improve it.

[pruttelherrie]

@PRR: I am in way over my head with this one, I agree.  The thing is that I want to break free from the booster/buffer type circuits, and I need a noisegate. I have looked at multiple schematics and this one seemed to be the one with the least parts.

Check out the "Elex noisegate" on "the other forum."

Not the best gate out there, but easy to recognise functional blocks and also easy to improve upon, within limitations of course!

Full disclosure: I'm the one who posted it there.

Also, I'm just reminded to look into the problem described in the second post.