DiVibe. A dual LFO PT2399 chorus thing.

[anotherjim]

This has been keeping me busy lately...

Sound clips to come soon...

Sorry for the convoluted looking scheme around the PT2399. It's pretty standard stuff really.

If you aren't familiar with the PT2399 chip, note there is an inverting op-amp from pin 16 to 15 (that feeds the analogue to digital converter) that is the input mixer. The Digital to Analogue converter outputs from pin 12. There is another inverting amp from pin 13 to 14 and used here as a mixer and output driver. Delay time is set by the current to ground from pin 6 and it's here that a mix of the two LFO's is applied.

This is an offshoot of my Arcomatic Rosinator project -  hence the 2 LFO's that I've used preset  speeds for (I have preset pots in mine - they could equally well be Panel pots).

CD4049UBE should work as well in the LFO's instead of the 4069, but the pinouts are different. In either case they MUST be the "UB" un-buffered devices.

The tone switches are there in case I want a low pass response similar to early BBD chorus, many of which were as low as 6Khz.

Sorry for using an obsolete (in TO92 package) 2N5457 JFET for the input buffer, but I wanted something compact. A J201 might be a better choice here but that too is obsolete in TO92 (but still available of course).

[~arph]

That clipping on the input, will it not distort clean signals? Have you tried that? I know you made this for a squarish wave signal

[anotherjim]

It isn't there to clip! There's no gain in front of it. It will clip if the input signal gets too hot >1.2v p-p.  I'd prefer that to the PT2399 overloading. I haven't got it to noticeably clip yet. First demo's will be with clean guitar!

[midwayfair]

You can avoid clipping/overloading the PT2399 without clipping the signal at all by putting an LED from pin 7 to ground. (You can also put it on pin8 and it'll blink out of phase with the one on pin7, but it's not necessary to use both). If you want to get to exactly 2.5V, put a diode in series with an LED, but I just use a green led with Fv ~2V. Your maximum signal size peak to peak will be the forward voltage of the diode, but it doesn't seem to distort no matter how big your signal gets -- it just limits the output.

Described in here: https://docs.google.com/document/d/1P7uOYFRH6ZW5eGoquhAiMQMyUU76bod_gNKkiSlPNJM/edit and originally (with some speculation about why it works) here: http://www.diystompboxes.com/smfforum/index.php?topic=101149.0

No one seems interested in using this trick except ~arph, though!

EDIT: note that it's fewer parts than two diodes and a cap, too!

[Kipper4]

Looks good Jim. Can't wait to hear it.
Jon, Sam Hay Used that exact same trick in the Space and Time reverb too with the diode on pin 7. Neat trick.
Thanks for the link too. I'll have to swap browsers to see it. my aol browser is no longer supported.

[~arph]

Yes Jon,

I know about that trick 

@jim  Yes I figured it was not there for intentional clipping, but with this arrangement you will get clipping when the signal is hot, but still in range for the PT2399. Try the LED trick on pin 7 it's really neat.

[anotherjim]

Thank you all. Excellent idea with the LED - I'll certainly give it go. Not heard of that until now.
So I can also allow the Regen gain to get a bit above unity with the LED limiter, I hope?
I know a chorus doesn't usually have Regen or a delay time adjust, but they were just too easy to add  to resist!
Actually, because of the longer delay time and 2 LFO's, this can sound like a vibe and a chorus and a reverb all at the same time if you want.

[merlinb]

The PT2399 clips fairly ordinarily by itself at a little over 3.5Vpp. I don't see the need for external clipping measures.

[anotherjim]

The PT2399 clips fairly ordinarily by itself at a little over 3.5Vpp. I don't see the need for external clipping measures.

Yeh, I'm getting that too from further reading. It seems the input opamp clips before the converters go over.
Also, I'm a bit wary of pulling too much current from a pin. I can't find data for the 2399 pin currents, but most MOS chips allow sourcing only in the order of 2 to 3mA unless specifically intended for higher currents.

But, the designer of the  Hamlet seemed to find headroom an issue - he got unpleasant distortion without the limiting.
What I'm getting from all these different approaches, is that these chips vary a lot. For instance, I'm getting better chorusing via the delay control than Ref Pin2. VCO clock is at least 1Mhz faster than the data sheet table for a given pin6 resistor.

So, I'd probably just delete D2 and D4. R6 was to limit current in those diodes, and C7 added to make an extra filter pole. Given the speed of the A-D, the single pole on the input op-amp is probably good enough. So cast out C2, R6 and C7 too!

[anotherjim]

Demo playlist
https://soundcloud.com/ashdalestudio/sets/divibe-chorus-effect
Apart from intro & outro fades and level normalizing, no post-processing has been done. The "Nasty" fuzz bass had a BMP before the DiVibe.

[anotherjim]

Added another little Demo. NON GUITAR CONTENT WARNING!

https://soundcloud.com/ashdalestudio/divibe-stringsynth

This has the DiVibe acting as an ensemble chorus on a string machine sample with it's own chorus off. Pretty close I think considering DiVibe has only one delay - not 3!

[anotherjim]

Version 1.1...


Input simplified and a little gain added. Also installed a J201 in place of the 2N5457. It seems better suited to 9volt (actually 8.4 volt due to protection diode drop) operation.

These changes don't radically alter the sound, although Regen does now approach that "verge of feedback" sound at maximum. Reduce R9 a little (91k if you have it, else 82k) if you really want it to feedback.