The FM Tube Modulator Jimi Loved

His guitar slung across his back, his dusty boots is his cadillac.

Moderators: VelvetGeorge, BUG

Post Reply
User avatar
Senior Member
Posts: 2470
Joined: Sun Apr 19, 2009 5:27 pm
Just the numbers in order: 7

Re: The FM Tube Modulator Jimi Loved

Post by Xplorer » Sun Jun 14, 2020 9:07 am

You never cease to amaze me JC ...

New Member
Posts: 3
Joined: Sun Feb 25, 2018 9:41 pm
Just the numbers in order: 13492

Re: The FM Tube Modulator Jimi Loved

Post by Eb7-9 » Sat Dec 04, 2021 9:02 pm

Hi gang, it's been a while ...

so, last time I was here I'd just about exhausted a short experimentation phase using AVR's (Arduino) in audio circuit designs - a thing that happened through a fortunate collaboration with Tek456 who supplied all the code ...

but, as I think I pointed out before, I was curious (obsessed) about finding an analogue solution to a speed variable 3-phase oscillator ... hopefully an elegant and simple one

just to recap, two of these AVR-LFO based designs were single-channel embodiements of William Wayne's Choratone circuit using matched opto-couplers, and later using 4066 analogue gates ... both of these being linear-modulation circuits, somewhat unlike the original Triode (Baldwin) and jFET hardware emulation (mine) versions ...

I will say the two single-channel AVR based circuits sounded very good, but not necessarily better than the jFET-based 3-phase LFO versions ... so, having triangular-shaped or ring-oscillator shaped output waveforms is not a make-or-break thing here // it's just that the original Baldwin unit was of the second variety (something worth pointing out)


here's a chronicle of my efforts in this area ...


my first guestimate was based on emulating the signal shapes produced by the AVR solution which consists of perfectly linear 3-phase waveforms ... here's my original diagram:

the problem, of course, is the missing constraints setting the upper and lower limits of the waveform - something I ended up wrestling with considerably at times ...
I tried all kinds of strategies, external servo and the like ...

this was the result of considerable effort using a parallel servo circuit involving 6 OTA's in all ... (too complicated again)

looks all fine and dandy at first glance - except when we start looking at things more long-term
the following simulation shows what happens after 3000 seconds of run time with the LFO set to very low rate ...

unable to tell if I was dealing with accumulated simulator error or some ther aspect of my OTA models (I was using my own discrete macro) I decided to let this OTA approach go //
maybe I missed something obvious ... one day I'll go back and gloss over that work to see if I did ... dropped the idea and moved to other design work


some time later I turned my attention towards Barry Gilbert's translinear quadrature oscillator technique :

which yielded very promising results on the simulator when configured as a ring ... operating strictly in current mode I found it a little challenging to integrate a DEPTH control into the greater scheme things in things ...

certainly not a complete write-off, I might go back to it later // and to be honest, it's probably the most elegant looking circuit so far ... no op-amps or OTA's, just bipolar devices and diodes


then later again, the idea came to try a finite state machine (FSM) approach which would produce square wave outputs and then applying current-controlled first order filtering that would give me a constant amplitude quasi-triangular output ... I also wanted to know if square pulses - as we had discussed with Dave in the beginning - would produce any effect ...

again, a 3-phase FSM circuit is something that isn't revealed anywhere on the net ... so, I had to design it from scratch // not a hard thing to do, but I remember hating the stuff in school ... so had to swallow some goat here a little bit ... running on four off-the-shelf CMOS IC's and wired on a single layer prototype layout using 23 bridges: ... date-2020/

the conclusion: no dice on producing any shifting effect, but it provided the opportunity to focus on the balancing of the three CTP Lattice legs and see that I'd been slightly off in my previous CTP efforts ... (!)

again, I realised that in order to produce the full enchilada (ie., FSM with current-controlled clock and filters) I would be skirting with an overly complex design in the end ...
a fun idea, not especially pretty but, likely, way too much work ... next,


fast fwd to Dec/2019 where I put my study cap on and immersed myself (again) in compressor and limiter designs - thanks to some conversations I had with a dude who goes by mictester ...

well, one thing lead to another over 18 months of running sims and I actually manage to come up with a bunch of new ways for producing bone-fide linear compression, bone-fide linear limiting, as well as current-controlled high headroom filters, phasors, and more ...

which are all about get built finally


from this side-work I used this newly discovered circuit block inside a three-stage ring oscillator and finally found what I set out to find four years ago ...(!)

with perfect wave-forms, duplicating almost exactly the original Triode version of William Wayne's circuit and the jFET emulation that I first started playing with in the beginning ...

here's a glimpse of what the output looks like:

with good simulated SPEED range the circuit is very amenable to coupling to a variable mixer via three OTA's and a triple-ganged pot providing DEPTH control ... from my other builds it seems OTA's are the best tool for the mixing stage in this context


OOPS! ... did I say triple-ganged pot ??!

yes, something I originally tried avoiding // which means I'm coming back full circle to where I started on page one of this thread with the op-amp based ring oscillator I posted there ...

of course, if we're going to end up using a triple-ganged pot to provide DEPTH control ... we may as well use a second one to set the oscillator RATE

that settles the argument ... (*!*)


in the mean time I've also figured out potential way of making the jFET and Triode (ie., original Baldwin) based circuits speed variable ...
and just another side caveat here, there are other alternatives in the design with regards to FET's ... so the direct hardware emulator approach is not completely dead either ..(!)


obviously at this point the next thing to do is to order some triple-ganged pots and see how an op-amp based ring oscillator works using the pot as speed controller ...

probably someone here can beat me to it ... :wink:


that's it for now, I hope y'all have good holidays ...

thanks for reading ...
best regards, peace for now


New Member
Posts: 1
Joined: Mon Mar 13, 2023 10:44 am
Just the numbers in order: 13492

Re: The FM Tube Modulator Jimi Loved

Post by KirkSutton23 » Mon Mar 13, 2023 10:45 am

if you're interested in cloning the vibe there's one chap you should know about (n case you don't already) who asked if he could use my work as a starting point for a definitive treatise on the Univibe / his name is Brad Burt and operates under the name Classic Amplification ... he's the author of the "Forum Vibe", a design essay that he composed w some consultation on my part (it's based on my original work) / and was meant as a definitive work on the vibe ... I'd ask him about cells.

Post Reply