====================================
PLEASE NOTE:
I am not "servicing" or responding to any comments on this thread - I am NOT willing to give any help or advice or assistance on ANY Moog Theremini related matter, and believe this thread was a waste of time and is not worth reading other than for "education".
IMO, the Theremini is not worth modifying or correcting, the CV circuits given here wont fix the core problems even if I got the theory right (and there's no certainty that I did).
Fred Mundell. Aug 2014.
====================================
WARNING! MASSIVE DOUBT AND UNCERTAINTY APPLIES TO EVERYTHING PRESENTED IN THIS POST !!
In the Theremini Review thread, some heavy discussion about CV's (for pitch control) has occurred - I presented a brief "CV tutorial" there which I am now moving here (see next posting - Oh ell! I did CTL-X to copy it, then went and copied a link - So I lost tat post.. Heres a link to Wiki instead.. )-8 .. I think this subject is of extremely limited interest to most thereminists, so wanted to move that discussion so that the "review" thread doesnt get hijacked even more.
I have a hypothetically possible "explanation" for some of the weird stuff people seem to be seeing and Moog seem to be saying (see posts on Theremini Review thread from here) - but this is pure hypothesis based on observations I have not witnessed, and which might be misguided.
The clues upon which this hypothesis is based have been gleaned from somewhat abstract (and frankly unbelievable) postings by synthguy and a quoted email from Moog..
From Moog: "It runs an output of 0-10V, but it is not necessarily following the internal pitch of the Theremini itself."
It was the "not necessarily" which really bugged me about this - this hypothesis is the only one I can find where, under one condition (and one only) the theremini MIGHT track the pitch at 1V/Octave.
then synthguys "It should have been obvious from the first crude measurements I made, showing a two octave spread giving 2.5 per octave @5V. That means each octave increases 2.5 volts."
Well, thanks synthguy - Your "It should have been obvious" is actually really funny, LOL ;-) but the clincher was the "Changing the low and high note range will "tune" the scaling. Setting the low note to the lowest setting and the high note to A#8 will get very close to 1V/Oct (but not quite perfect)."... Hmmm .. I wouldnt even call 1.1V "close to perfect" let alone something 'round 1.4V, but I suppose we have different ideas about "perfection" ;-) LOL
Whatever, without your observations, I would never have dreamed that this bizarre scheme would have been implemented by anyone, let alone Moog... If this really is whats going on, then I am between manic laughter post-traumatic-shock! ;-)
----------------------------
The hypothesis is this:
The CV output from the theremini tracks the theremini pitch, but does so in a manner completely (and IMO bizarrely) different to any CV implementation ever seen before - I suspect that if the hypothesis is true, the CV was implemented by an utter incompetent.. But anyway -
Imagine that the number of semitones set in the playing field (note Range, high note to low note) is divided into the selected CV output voltage - so lets say one set the low note at A1 and the high note at A5, and one selected 5V output.. If my (unlikely) hypothesis is correct, then when A1 was playing, the CV would be 0V, and when A5 was playing, the CV would be 5V.. If all the semitones and octaves were equally spaced within this 5V span, then, under these conditions (5 octaves in 5V) and ONLY under these conditions, one would get 1V/Octave output.
It gets more complicated ;-)
Lets say we set low note to D2 and high note to D6, and got 0V for D2 and 5V for D6 - We would still be getting 1V/Octave, but would need to re-tune our synth to get it to play the same pitch as our Theremini.
The generally accepted (1V/Octave) keyboard CVs are as follows:
Key: A0 A1 A2 A3 A4 A5
Voltage: 0 1 2 3 4 5
And one adds / subtracts voltages into 1V/Octave synths or modules (independently or globally) to tune them or shift their frequencies etc - Adding 1V increases the frequency (pitch) by one octave, subtracting 1V decreases pitch by 1 octave, Adding / subtracting 83.333mV raises / lowers pitch by 1 semitone, Adding / subtracting 0.83mV raises / lowers pitch by 1 cent.
So one can re-tune quite simply by having a box of some kind that adds/subtracts 1V increments for octaves and 83.333mV increments for semitones.. Most VCO's have tuning controls to allow at least +/- 1 octave of adjustment, so getting a 5V 1V/Octave CV from the theremini one should be able to at least track the note, even if one cannot shift the VCO to the right octave.
Less than 5 octaves:
What happens if one wants to play less than 5 octaves ? Say you set low note to A1 and high note to A4 (3 octaves) .. A1 would be 0V, A3 would be 5V, so each octave would output 1.6666V .. To get 1V/Octave you would need to reduce (divide) the 1.6666 by 1, giving a multiplicand of 0.6
This is fairly easy with a couple of resistors and could be added to our box.. But this box is starting to get a bit hairy.. but we will get back to that...
More than 5 octaves:
Lets say you had 6 octaves in 5 volts - each octave would be 0.8333V, and the output would need to be amplified to bring it up to 1V/Octave.
Using the 10V output:
Its not possible to ever get 1V/Octave directly from the 10V output, the closest one could get is 1.4285V/Octave when 7 octaves were selected.. so this would need to be attenuated..
IF (and its an EXTREMELY BIG IF!) the above guess as to whats going on is correct, then a reasonably simple box could be constructed to interface the theremini to 1V/Octave equipment - here is a brief overview:
1:) A 12 way rotary selector switch to set the key (offset) - this would require precision resistors at least.
2.) A 12 way rotary switch to add 6 octaves in 1/2 octave steps.
The above would allow one to tune the synth to frequency of the lowest theremini note without extensive twiddling with the synths tuning controls, but wouldnt deal with correcting the theremini output and making it 1V/Octave.
To get 1V/Octave, one would need to limit your 'span' to 1/2 octave steps (as in, select only 3, 3.5, 4, 4.5.... 7 octaves between lowest and highest notes -- You could have finer increments, even down to semitones, but would need a much more complicated, probably digital* box to do this)
Drive the output at 10V, and you would need another 12 way rotary switch to select the number of octaves you are playing. This selector switch would switch in the required attenuation to convert the output voltage to 1V/Octave.
Below is a table of required multipliers (all <1 so therefore attenuators or potential dividers - probably just a resistor and trimmer resistor for each):
Listed for 10V output - Octave 'span', output voltage per octave, multiplier required to give 1V/Octave
7 | 1.428571 | 0.700 |
6.5 | 1.538462 | 0.650 |
6 | 1.666667 | 0.600 |
5.5 | 1.818182 | 0.550 |
5 | 2 | 0.500 |
4.5 | 2.222222 | 0.450 |
4 | 2.5 | 0.400 |
3.5 | 2.857143 | 0.350 |
3 | 3.333333 | 0.300 |
* An all-in-one digital box with a display corresponding to the theremini display (low note / high note) would enable the CV to be correctly converted to 1V/Octave for any selected values, and provide the required attenuation and offset - It wouldn't require that much work, as one could use DAC's and MDACs to do the job - It could probably be done with a PSoC .
But the above fact is what makes me doubt this hypothesis - because EVERYTHING required to do this simple conversion MUST already be inside the theremini - So Why did (if they did) Moog not implement it ??
However, if this hypothesis is correct, I can fix the problem!
Fred.
I tidied up the tables..
CV Span 10V | Required | |||||
Multiplier | ||||||
V out | For | |||||
Octaves | / Octave | 1V/Octave | ||||
7 | 1.428571 | 0.7 | ||||
6.5 | 1.538462 | 0.65 | ||||
6 | 1.666667 | 0.6 | ||||
5.5 | 1.818182 | 0.55 | ||||
5 | 2 | 0.5 | ||||
4.5 | 2.222222 | 0.45 | ||||
4 | 2.5 | 0.4 | ||||
3.5 | 2.857143 | 0.35 | ||||
3 | 3.333333 | 0.3 | ||||
( 0.08333 Volts/Semitone = 1V/Octave.. 84.167mV/st to 82.5mV/st for +/- 1 cent accuracy)
With 10V Full-Scale Output | |||||||
Multiplier | |||||||
# | # | V/ | V/ | for | |||
Semitones | octaves | Semitone | Octave | 1V/Octave | |||
24 | 2.00 | 0.4167 | 5.0000 | 0.2000 | |||
25 | 2.08 | 0.4000 | 4.8000 | 0.2083 | |||
26 | 2.17 | 0.3846 | 4.6154 | 0.2167 | |||
27 | 2.25 | 0.3704 | 4.4444 | 0.2250 | |||
28 | 2.33 | 0.3571 | 4.2857 | 0.2333 | |||
29 | 2.42 | 0.3448 | 4.1379 | 0.2417 | |||
30 | 2.50 | 0.3333 | 4.0000 | 0.2500 | |||
31 | 2.58 | 0.3226 | 3.8710 | 0.2583 | |||
32 | 2.67 | 0.3125 | 3.7500 | 0.2667 | |||
33 | 2.75 | 0.3030 | 3.6364 | 0.2750 | |||
34 | 2.83 | 0.2941 | 3.5294 | 0.2833 | |||
35 | 2.92 | 0.2857 | 3.4286 | 0.2917 | |||
36 | 3.00 | 0.2778 | 3.3333 | 0.3000 | |||
37 | 3.08 | 0.2703 | 3.2432 | 0.3083 | |||
38 | 3.17 | 0.2632 | 3.1579 | 0.3167 | |||
39 | 3.25 | 0.2564 | 3.0769 | 0.3250 | |||
40 | 3.33 | 0.2500 | 3.0000 | 0.3333 | |||
41 | 3.42 | 0.2439 | 2.9268 | 0.3417 | |||
42 | 3.50 | 0.2381 | 2.8571 | 0.3500 | |||
43 | 3.58 | 0.2326 | 2.7907 | 0.3583 | |||
44 | 3.67 | 0.2273 | 2.7273 | 0.3667 | |||
45 | 3.75 | 0.2222 | 2.6667 | 0.3750 | |||
46 | 3.83 | 0.2174 | 2.6087 | 0.3833 | |||
47 | 3.92 | 0.2128 | 2.5532 | 0.3917 | |||
48 | 4.00 | 0.2083 | 2.5000 | 0.4000 | |||
49 | 4.08 | 0.2041 | 2.4490 | 0.4083 | |||
50 | 4.17 | 0.2000 | 2.4000 | 0.4167 | |||
51 | 4.25 | 0.1961 | 2.3529 | 0.4250 | |||
52 | 4.33 | 0.1923 | 2.3077 | 0.4333 | |||
53 | 4.42 | 0.1887 | 2.2642 | 0.4417 | |||
54 | 4.50 | 0.1852 | 2.2222 | 0.4500 | |||
55 | 4.58 | 0.1818 | 2.1818 | 0.4583 | |||
56 | 4.67 | 0.1786 | 2.1429 | 0.4667 | |||
57 | 4.75 | 0.1754 | 2.1053 | 0.4750 | |||
58 | 4.83 | 0.1724 | 2.0690 | 0.4833 | |||
59 | 4.92 | 0.1695 | 2.0339 | 0.4917 | |||
60 | 5.00 | 0.1667 | 2.0000 | 0.5000 | |||
61 | 5.08 | 0.1639 | 1.9672 | 0.5083 | |||
62 | 5.17 | 0.1613 | 1.9355 | 0.5167 | |||
63 | 5.25 | 0.1587 | 1.9048 | 0.5250 | |||
64 | 5.33 | 0.1563 | 1.8750 | 0.5333 | |||
65 | 5.42 | 0.1538 | 1.8462 | 0.5417 | |||
66 | 5.50 | 0.1515 | 1.8182 | 0.5500 | |||
67 | 5.58 | 0.1493 | 1.7910 | 0.5583 | |||
68 | 5.67 | 0.1471 | 1.7647 | 0.5667 | |||
69 | 5.75 | 0.1449 | 1.7391 | 0.5750 | |||
70 | 5.83 | 0.1429 | 1.7143 | 0.5833 | |||
71 | 5.92 | 0.1408 | 1.6901 | 0.5917 | |||
72 | 6.00 | 0.1389 | 1.6667 | 0.6000 | |||
73 | 6.08 | 0.1370 | 1.6438 | 0.6083 | |||
74 | 6.17 | 0.1351 | 1.6216 | 0.6167 | |||
75 | 6.25 | 0.1333 | 1.6000 | 0.6250 | |||
76 | 6.33 | 0.1316 | 1.5789 | 0.6333 | |||
77 | 6.42 | 0.1299 | 1.5584 | 0.6417 | |||
78 | 6.50 | 0.1282 | 1.5385 | 0.6500 | |||
79 | 6.58 | 0.1266 | 1.5190 | 0.6583 | |||
80 | 6.67 | 0.1250 | 1.5000 | 0.6667 | |||
81 | 6.75 | 0.1235 | 1.4815 | 0.6750 | |||
82 | 6.83 | 0.1220 | 1.4634 | 0.6833 | |||
83 | 6.92 | 0.1205 | 1.4458 | 0.6917 | |||
84 | 7.00 | 0.1190 | 1.4286 | 0.7000 | |||
<EDIT> Looking at the above figures, even making a external box to allow entering the low + high note and using these to rescale the CV isn't going to be easy - I think it will require better than an 8 bit MDAC to perform the analogue division accurately enough - probably needs 10 bits (theres only 8 bit MDACs on PSoC, and other MCU's dont even have this) - Using PSoC I would probably use the DAC's for the offset, and drive off-chip digital potentiometers for the division.... Ugh...!
One really does need to limit the span to increments of 1/2 octaves to keep things "simple".. A rotary switch with resistors/trimmers or a 8 bit MDAC would achieve this.
The easiest would be to limit the span to one octaves steps, that way you could set both low and high notes at the same key on the theremini, say C1 to C4, have one switch on the box to set this key (C), one switch to set the octave span (3), one switch to set the low note octave (1).
- or set the key and octave to whatever you want, they will still track musically at the interval you set - the only thing that must be set correctly in order to get 1V/Octave is the octave span.
ALL the above is based ENTIRELY on a hypothesis about the theremini's CV output which HAS NOT BEEN CONFIRMED!!!