Here are some sexy picture on my modular from 1978.
Here is the dual VCO with SSM 2030 and a Q81 glued on top:
And here is the power supply. You can see the heater voltage ret - it’s the chip in the lower left with a white plastic “hat” to keep in the heat:
minimoog switched to 3046 in 1979 (third version)
I have a bunch of high precision Q81 (and equivalent) from my previous life of unsuccessful builder 
They tended to be a main point of failure, I remember.
I have it on good authority that a “fix” for the early MiniMoog drifting was to move the VCO components above the power supply to stabilise the temperature.
Pretty hacky!
On the “reversing oscillator principle”. Sadly I can’t find a pdf of Electronotes 115, referred to as giving more information by the author.
Bernie Hutchins, ‘A Voltage Controlled Oscillator with Through-Zero FM Capability’; Electronotes 129, 1981.
Last year bought the “box set” of every single issue. There had been one I really wanted. When I get around to it I’ll see if I can find 115. (and btw, I used to be a subscriber back then, but I didn’t keep all of them 
The lists early subscribers is a real “who’s who” of boffins from back then.
I recently updated my LFM + PM test patches with the addition of Neóni as well as several other changes. One of those changes was the addition of a short melody being played into the oscillators’ V/OCT inputs.
I bring this up because this led me to find something interesting: there are 3 TZLFM oscillators included in the FM patch (Chebyshev is missing because it lacks a sync input), but Terrorform (with hard-sync) stands apart in that it is the only one that holds the melody rock-steady while using hard sync.
(However Neóni-with-hard-sync and Terrorform-with-soft-sync do something else interesting and musical; they do hold the melody but also rhythmically oscillate down a major third. The rhythm comes as part of the algorithm sequence that I’m using, and the pitch comes from one of the 4 main oscillators involved in the configuration I’m using.)
I’m really curious about what Valley is doing differently here to make this happen. Looking at your investigation here, both Fundamental WT VCO and Terrorform use a constant Hz/volt so that’s not it. Given that Terrorform-with-soft-sync acts more like Neóni-with-hard-sync, I’d guess that it’s something related to the implementation of sync but that’s just a guess. It’s odd that the similarities seen between Neóni and Terrorform are reversed in terms of sync.
I need to look at your patch again. I have forgotten which true TZFM modules you used.
I would expect Neoni to sound different than most, given that at 0V FM input, the Neoni stalls, whereas all/most others stall (or nearly stall) at some negative FM input. Being through zero, they all resume oscillating as you go further negative below the stall point.
So the Neoni has symmetry around the 0 crossover point, assuming the modulating wave has symmetry between the negative and positive portions.
What kind of difference to expect? - I haven’t a clue. I am still a complete neophyte when it comes to understanding the sound of FM. So far I have just been trying to understand what the oscillators do from a voltage perspective to implement the FM.
I listened to your most recent vids, but did not look at the patches. I’ll try to find time to look over them and see if it sparks any ideas.
Many thanks. I wish I was better at maths The URL gives the game away for following further links (eg the reference to 113). Very useful. Cheers!
This is the same old site, yes? At least back when I ordered all of them they were not all available here. That’s why I had to buy all three boxes full. Has this changed and they are all here now?
Only a few of them are available online. Here is the overview:
http://electronotes.netfirms.com/free.htm
Most of the FM-newsletters are available for free. EN206 is available too.
Ok, sounds the same. For decades I had been looking for the issue with the 4 pole state variable filter, as I had made one when I first got the issue, but lost the issue. Unfortunately I tried to turn it into a VCV module (F4), but I could never get it to behave, so I made the much simpler F2 instead.
Edit: This is of course the analog one I made around 1978 from the (then) new issue of Electronotes. I haven’t turned it on since maybe 1998, but I did pull it out for some photos a couple of years ago.
Thanks for the heads up; I’ve added the information about the modules used (as well as the plugin versions) to the linked post as well as the video descriptions. I also added some info I forgot to mention about the manual editing which I performed on Neóni’s preset to try and create a “standardized testing unit”.
Anyways, thanks for your time. I’ve also reached out to Valley to see if they have any ideas about what is differentiating Terrorform from Neóni and WT VCO here, as well as what’s causing soft-sync-Terrorform to sound reminiscent of hard-sync-Neóni.
Oh, I forgot. SFZ Player can play samples back and apply “digital” though zero PM.
Jakub Ciupinski has an interesting FM video where he demonstrates a use of a 0 Hz FM carrier wave (VCV WT VCO set to ~0 Hz).
After watching it, it suddenly dawned on me - the Neoni through zero FM is just FM with a 0 Hz carrier wave. The “frequency” setting of the Neoni simply controls the scaling of the FM modulation. Another way to say it is the frequency can be used to set the depth of the FM.
To get the slow beating that the VCV WT VCO produces, you need to set the Neoni FM to DC coupled, and then add a few (maybe somewhere between 2-7) mV DC bias to the FM modulating signal.
The Neoni carrier frequency is so close to zero, that there is no discernable beating. Adding the bias effectively offsets the base frequency of the carrier to something other than 0.
