Questions re: LFO to modulate dual PT2399 delay

[Radical CJ]

Hi all,

I'm picking up a dual PT2399 delay project that I put down a few years ago when life got in the way. Its essentially 2X of The Valve Wizard's Small Time delays (http://www.valvewizard.co.uk/smalltime.html) that can be switched to run series/parallel; mono/stereo; regular/exchanging repeats; etc.

I think there is room in the enclosure for a daughter board, so I guess its an imperative to add some modulation.

Fortunately for me, The Valve Wizard also has a PT2399 chorus with an LFO already designed for a similar purpose (http://www.valvewizard.co.uk/jennygreenteeth.html), but I want to make some additions such as inverting the modulation signal to one of the delays.

This is what I was thinking (see below), but I have some questions for those with more knowledge about these things:

1) The design requires 2 opamps for the LFO and 1 to invert the signal, but given I'll probably use a quad or 2 dual opamps there will be a spare. Does it make sense to use this to buffer VR for the LFO or to buffer the non-inverting LFO signal, or neither? (see opamps labelled "?" in schematic).

2) Opamp B shows VR going into the inverting input, as per the original schematic I based this on. However, my instinct is that this should be going to the non-inverting input, and this may have been a typo in the original ... Obviously, this is something that can be worked out on the breadboard, but wondered what others think.

Kind regards!

[anotherjim]

You can't invert the opamp buffer "C" that way. It will just try to hold things at VR2 and stop all modulation.
You need to use a proper inverting amplifier circuit with x1 gain by fitting equal resistors for input and negative feedback. Doing the inverting there also inverts the DC level which will probably result in the drive to the delay pin6 transistor swinging too far positive. The place to invert is, unfortunately, after the LFO and would need another Depth pot followed by another non-inverting buffer.
Yes, opamp B inputs are drawn swapped over.

[Radical CJ]

You can't invert the opamp buffer "C" that way. It will just try to hold things at VR2 and stop all modulation.
You need to use a proper inverting amplifier circuit with x1 gain by fitting equal resistors for input and negative feedback...

Thanks for the reply! What gave me the idea was the Dimension-P Chorus (below), which does invert the signal post-depth knob, although also using equal resistors for input and negative feedback (as you described), and with a number of differences in the design (including that it isn't used on Pin 6 of the PT2399). Back to the drawing board! and maybe no inversion.

[duck_arse]

your dia with the three Vr's - you can't use a stiff Vr for the schmitt/integrator, it needs a little space from the reference to jump about when it switches. either Vr2 or Vr3 would work. ahhh, no, not Vr2, what did I just say? that 10uF would stop the osc.

[edit :] I retract - see below

[Radical CJ]

your dia with the three Vr's - you can't use a stiff Vr for the schmitt/integrator, it needs a little space from the reference to jump about when it switches. either Vr2 or Vr3 would work. ahhh, no, not Vr2, what did I just say? that 10uF would stop the osc.

Thank you for raising this issue Mr Duck, as this also something I was curious about. I've previously put together and tinkered with some LFOs for tremolos/vibes, and these did not have a buffer or bypass cap on the VR. I wondered whether this would affect the signal created by the switching, but don't really have the EE knowledge or terminology to quite conceptualise or articulate the problem.   

However, the Jenny Greenteeth Valve Wizard schematic on which this is based does indeed have a bypass cap on the 4.5V VR (as per VR2 above), which is also used to bias the op amp buffering the guitar signal, and that circuit works. Something to breadboard I guess? 

   

[anotherjim]

The Dimension P has a major difference that helps with the LFO invert. The delay time modulation is applied to pin 2, not pin 6. Pin 2 is a Vcc/2 reference in the chip of 2.5v and the modulation is AC coupled to it via a cap. As its AC, The DC level isn't critical. The Depth pot is a series attenuator and won't cause a DC shift when adjusted therefore the LFO invert won't cause any DC shift and the pair of delays get equal but opposite polarity modulation swinging around the 2.5v reference in pin 2.
If the Depth pot is wired as a volume control to 0v to control pin 6, then an opamp following that sees 0v DC input. If it's a unity gain inverter then compared to the +input reference voltage it's given -4.5v below the reference and the output inverts to +4.5v above the reference which is +9v. That completely changes the drive to the control transistor and it may only respond to the LFO at high depth settings. It may be improved by inserting a cap after the depth pot then the inverter will bias at 4.5v output allowing better control swing. It would be a large cap at LFO frequencies however the opamp resistor values can be higher than normal (don't care about resistor noise here) even 1M allowing a lower value cap.

[duck_arse]

well. last time I tried it on the BB, the oscillator circuit failed [perhaps due to extreme resistor values] when I added a bypass cap to the V/2 reference. however, I look at circuits I have drawn of other BB adventures - shows the additional bypass. so, now I don't know. try it on the breader, observe the V/2 on an oscilloscope if available.

to that last point, I would not bias the audio section with the same bais feeding the oscillator. tikk tikk tikk.

[Radical CJ]

...
If the Depth pot is wired as a volume control to 0v to control pin 6, then an opamp following that sees 0v DC input. If it's a unity gain inverter then compared to the +input reference voltage it's given -4.5v below the reference and the output inverts to +4.5v above the reference which is +9v. That completely changes the drive to the control transistor and it may only respond to the LFO at high depth settings...

I follow. Now that you've pointed it out I can see why it won't work. I think the problem could be solved but with a bit of a redesign and extra components, but I'm not actually sure having an inverted modulation signal will produce a more interesting sound as compared to having two slightly different delay times with the same modulation, and I want to keep the board smallish and tidy. So I think I will abandon that idea. Thanks for the knowledge!   

... I would not bias the audio section with the same bais feeding the oscillator. tikk tikk tikk.

I've encountered this one before 

[anotherjim]

Conterwise FX modulation can work very well in stereo (even if the cabs are side by side) but isn't so great mixed down into mono. Many stereo FX allow it but when one side's output jack provides a mono mix if the other side's jack isn't used, it can actually be better to use the other jack instead to avoid the mono merge feature. I find this especially true of rotary speaker FX and even stereo pianos. Of course, you lose something but you don't get the messed-up phasing issues either. YMMV.