Making a tremolo project. Could somebody help with JFET volume dump circuit?

[rockgardenlove]

So, I'm working on this tremolo.  I've got my oscillator all set up, so I'm good there.  I'm having trouble with the volume dump setup though.  Here's what I'm trying to do:

The oscillator is fed into the gate.  That pot feeds some voltage into the gate to set the operating point.

So, all is going well, except I'm getting monster treble+volume suck.  I can't get the FET to close adequately.  Anybody got an idea?

Thanks!

[R.G.]

Let's think for a minute.

If you remember some posts on the forum here, you'll remember that JFETs are depletion mode devices. They are normally on, and you have to turn them off.

The J201 is an N-channel JFET. To turn an N-channel JFET off, you have to make the gate more negative than the source. I don't see any way your J201 can ever turn off, unless the LFO feeds a substantial negative DC voltage to it and overwhelms the pot.

[rockgardenlove]

Right.  Oops.  I feel dull now.

Hmm.  I stole this circuit idea from another schem, I'm trying to improve it really.  The problem is that it sucks monster treble.  I suppose I could just do it EA-trem style, eh?

[R.G.]

The EA tremo has the same problem. They just ignore it and use a JFET with a bigger Vgsoff, so that all of the tremolo is in the JFET-off direction. The J201 has such a low Vgsoff that you're going to get off/on banging with Vg=0. In your circuit, with the grid pulled positive, you may never turn it off, hence the terrific treble loss by loading down whatever drives the input, especially if it's a guitar.

You're not dull - just inexperienced. But you're a bit more experienced now! Your next one will be be better.

You know what experience is, don't you? Experience is what you have left when you've forgotten her name.

[markm]

Experience could also be considered a series of mistakes that one has learned from! 

[rockgardenlove]

So it does, at a closer look.  (The EA, I mean.)
Hmmm...

I tried to just make JFET emu of the Fender bias vary trem, I'm talking this:
http://www.ampwares.com/ffg/schem/champ_vibro_aa764_schem.gif

Didn't do anything.



Any suggested methods?  I've been looking around for schems to learn from and nothing new is coming up.

[A.S.P.]

that`s a very different animal,
but:
do you see those resistors between cathode and ground?
they got a purpose...

[gez]

So, I'm working on this tremolo.  I've got my oscillator all set up, so I'm good there.  I'm having trouble with the volume dump setup though.  Here's what I'm trying to do:

The oscillator is fed into the gate.  That pot feeds some voltage into the gate to set the operating point.

So, all is going well, except I'm getting monster treble+volume suck.  I can't get the FET to close adequately.  Anybody got an idea?

Thanks!

I've used exactly this set up with a MOSFET in the past - though I AC coupled the LFO to its gate using a large NP cap and used a 1M bias resistor to a smaller value trim across the rails - so just sub one of those in instead.  No guarantee it won't distort, but you can compensate for that in your design.

[rockgardenlove]

MOSFET, ok, I'll try it tomorrow.  Thanks very much!

[markm]

I've used exactly this set up with a MOSFET in the past - though I AC coupled the LFO to its gate using a large NP cap and used a 1M bias resistor to a smaller value trim across the rails - so just sub one of those in instead.  No guarantee it won't distort, but you can compensate for that in your design.

ROG EA?

[gez]

I've used exactly this set up with a MOSFET in the past - though I AC coupled the LFO to its gate using a large NP cap and used a 1M bias resistor to a smaller value trim across the rails - so just sub one of those in instead.  No guarantee it won't distort, but you can compensate for that in your design.

ROG EA?

Yeah, similar to the input stage of the EA, but the drain is connected (not DC) to the signal path/feedback loop (if used in a T-filter), and the source is grounded.  I no longer use this method, but it works well.  If the trimpot is decoupled as in the EA preamp, then adjusting things can be a bit of a pain as there can be a delay before any adjustment kicks in.  Also, mosfets that have been DC biased don't need a huge signal on the gate - 0.5V to 1V peak-to-peak is good - so you might need to divide down the LFO signal (see EA trem for inspiration).

If distortion occurs, then the audio signal will need to be divided down somewhere in the input stage and boosted after the modulation.

[gez]

PS, you only need to divided down the LFO signal (if too large that is) if you're using a phase-shift oscillator.  With triangle LFOs, you can just vary some of the existing components.

[sfx1999]

Here's a good trick:

Instead of hooking that FET directly to ground, put a diode or two there. The voltage drop will cause a positive voltage at the source, therefore 0 volts would be negative with respect to the source.

[rockgardenlove]

I tried the diodes.  Didn't do much, except I had to put a capacitor around them to dump the audio.

Thanks for the suggestion though!

I'll try the MOSFET next.

[gez]

Here's a good trick:

Instead of hooking that FET directly to ground, put a diode or two there. The voltage drop will cause a positive voltage at the source, therefore 0 volts would be negative with respect to the source.

In this application, the drain isn't DC coupled so the source isn't going to be more positive than the gate of the JFET.

[sfx1999]

Oh, oops.

The other option is to rig up a voltage inverter, but just go with the MOSFET.

Or if all else fails, an LDR/LED.