Vibra matic question

[Kipper4]

Quick question.
My brain can't figure this out. It's late, I'm tired. I can't sleep.

Here's the schematic.

http://geofex.com/FX_images/vibmatic.pdf

What is B?
A bias voltage?
I'm stewed help.
Thanks Rich

[Trackhappy]

Looks to me like a bias voltage "regulated" by the two 1N914 diodes and smoothed by the 10uF cap. "B" is just used to join the two points and save a line running all the way across the diagram.

[R.G.]

Maybe I can help.

"B" is a bias voltage/reference. It's low so it can be used for both the active rectifier and the comparator sections.

The bottom section with U1A and U1B opamp holds the LFO of the EAN tremolo off if signal is not present. When signal is detected, it releases its death grip on the LFO oscillator and the oscillator starts up.

The fun stuff is HOW the LFO starts up. A Nyquist or gain-phase oscillator has just a little more gain than it needs to start and run, so it starts slowly, amplifying any noise up into a nearly sine-wave. But it starts slow, and grows into full modulation over a half-second or so, perceptibly ramping up into modulation.

[Kipper4]

Thanks Trackhappy
and Thanks RG.
Thats perfect and it sounds like a fun effect. I'll breadboard it later tonight.
I like the sound of that and it feels like something that already exists in the tremvolope.
I'm working on an envelope controlled tremolo.
Too many things are not going as expected. I have work to do.
any contributions are welcome.
Rich

[Kipper4]

B=1v2 is that right?

[R.G.]

B=1v2 is that right?

Depends on the diodes and the current through them. A diode's forward voltage is a "sloppy constant". Each diode has a slightly different one, depending on the exact nature of its doping, but they're all close, depending on the material. Normal silicon junctions are just whispering about being on at 0.45V, and are well into conduction at 0.5-0.7V.

So yeah, 0.6V per diode is about right. But don't be surprised if it's 1.1 or 1.3.

[Kipper4]

Thanks RG so its in the ballpark.

[R.G.]

Thanks RG so its in the ballpark.

You got it.

[Kipper4]

by my calcs i get U1B bias at  ballpark=0.25v

1.3v x (2.4k/12.4k)
1.3 x 0.19=0.25v

[robthequiet]

Follow 

[Kipper4]

So why not use resistors to make B?
I suspect it's more to do with the currant than voltage right?

Why use a jfet Q2 and not a bjt?
What would be needed to change it to a bjt if possible or practical?

[PRR]

> why not use resistors to make B?

The ~~1.2V at {B} is intended to bias-up U1A input and output stages out of a near-zero dead-zone(*), and to semi-cancel the two diodes near "Attack".

If battery voltage and ambient temperature were known and constant, you could do a resistor thing.

Diodes are really cheap in mass production and trivial in DIY. (You should have a 10-bag if not a 50-bag handy). Do diodes.

(*) Some not all of those opamps will go to zero in and out.

[R.G.]

So why not use resistors to make B?

What Paul said.

Why use a jfet Q2 and not a bjt?

That's what was in the original EAN Tremolo, and my objective here was to make the fewest possible changes to the tremolo consistent with the LFO funny stuff. I didn't design the EAN Trem, it's preexisting.

What would be needed to change it to a bjt if possible or practical?

Fair question. BJTs as variable resistors do work, but have a rather smaller undistorted signal range than JFETs. I suspect the original designer of the EAN used a JFET because BJTs distorted too much. There is a whole body of knowledge about use of BJTs as switches and linear resistors that's too long to type here.

But this does highlight an oddity about the EAN Trem that I've wondered about. The capacitor in series with the JFET gate prevents the DC from the LFO from reaching it,  but gives the JFET gate a control signal that's only a swing around zero. I've often wondered if the source-gate of the JFET clamps the incoming LFO AC signal to ground, and if it does, was it intentional?

[Kipper4]

Thanks guys I'm over the moon.
Fantastic stuff.
I've got plenty of diodes. Thanks for explaining Paul. I just had to ask and try to understand.

I guess that makes the envelope very sensitive and fast being such a small proportionate voltage.
(Another guess since I haven't breadboarded it yet)
I will soon and make use of the cro.to the best of my ability to gauge the response.
I have used something similar with rebasing envelope detector op amps.(growler wah) and the use of double series diodes rectifiers.
Thanks for explaining the jfet too RG. That makes sense.
I breadboarded the trem part of the tremolo last night. My first time doing a signal to ground trem.
I was surprised how quite it was in terms of lfo noise compared to a shoot the moon series ldr set up.
I still have much work to do with my experiments on the breadboard and I may never achieve my goals. I know others with a far superior knowledge have tried.
Even if I don't achieve those goals but improve my understanding.
I'm still a winner
Thanks a million.
Rich