13 Theremin Myths – Don’t recognize any then add yours?

Two of the TW staff members showed curiosity as to what was on my myth list. I hope the staff doesn’t think I express heresy.

omhoge said: (excellent)

So what are the 13 forbidden truths???
You've really got me curious.

I can't help but guess, are any of them:
- volume control is more important than pitch
- the theremin is not that hard to play
- pitch Linearity should be adjustable
- ear training is useless if you don't know phrasing
- bent pitch antennas work better
- tubes are better than transistors
- loops should be square


Below is RS Theremin's misconception myth list, it is always evolving.

The list expresses concepts in reverse of what I interpret as the truth.

(Words in brackets are truth.)

--start

Tube amplifiers (tube oscillators do) will give a theremin a better sound.

Clara’s voice is due to the way she “shapes” her volume hand.

To capture Clara’s instrument sound (technique-we need her) is not possible without her.

Hand capacitance explains all that is needed to understand the pitch control.

If a theremin design concentrates on the professionalism of the schematic, detail to the board/chassis layout, components faithful to the original design, a beautiful enclosure, be cautious as this has no relationship to a good sound.

A theremin sounds bad because it is cheap, (design is the issue not cost).

All different future theremin models cannot have similar pitch and volume field response.

Perfect pitch field linearity does not add anything to developing your playing ability.

A beautiful sound does not encourage better playability.

Operating your pitch oscillators at the lower original frequencies works better.

Large air coils have more advantages over small high Q ferrite inductors.

Single diode detectors are less effective than more complex configurations.

The signal removed by diode detection from heterodyning is a 1/2 wave shape. (What is detected is one side of the two symmetrical sidebands plus the mixed signals.

--end

Anytime someone talks theremin or sales without demonstrating a “sound byte” RUN!

This sample below is my “personal” sound which endorses my Myth list. It develops from a pitch oscillator operating at 900 kHz, is drift free, uses small high Q ferrite core inductors with simple single diode detection giving me an ideal sine wave. Pitch field linearity is perfect (not almost linear) using a custom antenna.: My Theremin Sound  .mp3 175k Can't play a tune though )-:

OK Theremin World…

I know what you’re thinking: 'Did he fire six shots, or only five?' Well, to tell you the truth, in all this excitement, I’ve kinda lost track myself. But being this is a .44 Magnum, the most powerful handgun in the world, and would blow your head clean off, you’ve got to ask yourself one question: 'Do I feel lucky?'

Well, do ya, punk?

The most powerful handgun in the world is as good as the the skills of the person pressing the trigger...

That also applies to the theremin and its puppet ;)

Here are a few myths that come to mind:

. A theremin was used for "Good Vibrations"
[i](a myth so ubiquitous that I don't bother to refute it anymore)[/i]

. Wikipedia is an accurate source of theremin information

. YouTube is a good source of theremin performances and information [i](yes, there are a some good ones but not in general)[/i]

. One can develop one's technique/approach by combining different methods [i](for example, Pamelia Kurstin's and Carolina Eyck's techniques are incompatable -- one must choose one or the other -- and yes, I've watched both of them perform in person)[/i]

. the theremin is capable of nothing other than spooky music and sound effects

. if the theremin sounds in-tune to the performer, it will sound in-tune to the listeners [i](poor balance or speaker placement may mask the note attacks from the performer while the audience cringes)[/i]

. if the theremin sounds off-tune to the performer, it will sound off-tune to the listeners [i](if the balance is correct, the performer will hear the attacks and make quick micro-adjustments before the notes are heard by the listeners)[/i]

. there is something wrong with use of a pitch-preview

[i](use of a pitch-preview enhances hand-ear coordination and enables one's non-previewed playing.)[/i]

> Large air coils have an advantage over high Q ferrite inductors.

This one caught my attention in particular, because in an uninformed way, I've actually always thought that was true. Can you explain more?

Hello Jason,

It has been a while since we chatted. What I list as myths are my observations and opinions, they may not necessarily be science or fact but they do give me the most desired sound, pitch field response and excellent volume control shading I am after.

[i]> Large air coils have an advantage over high Q ferrite inductors.[/i]

Large air coils which I used religiously in my beginning years of research added to pitch drift because of parasitic capacitance between windings and the effects of the thermal expansion of a larger surface. They were also more susceptible to picking up stray magnetic fields and noise. They have nothing to do with generating a better sound or the method I found that creates a perfectly linear pitch field. They can have a more linear pitch field but with fewer octaves and they do affect the sound, most likely with a less than desired wave shape. They look beautiful but are not cost and time efficient when a $1.00 commercial coil can be better and used instead.

Higher Q coils develop more energy in the theremin antenna, this can be measured. My antenna theory has been dismissed by some because I focus on the behavior of current in a tuned antenna which is then affected by body capacitance. I left this subject off my myth list to be polite.

Just for the record: The sound of the theremin I am after is achieved after generating a perfect sine wave then adding in harmonic overtones which can easily be done using vacuum tube oscillators and a trick that for now will remain undisclosed but will be public eventually. I have one timbre knob that fully left is the “sweetest” sine wave, while turning it clockwise begins to excite harmonics. Simple as that sounds some days the rich sound is slightly off as far as I can tell, I need an assistant to help me listen and maybe he could learn a little construction. $$

Christopher

I allow to jump in, perhaps I may be able to put some light onto these coil myths and make subjective perception and reality to become closer:
The "ideal" coil, the pure inductance does not exist. The wound wire always has also a real resistance which is to be seen in series with the ideal inductor. Out of that, the fact of winding the wire creates parasitic capacitance between the windings. The sum of these parasitic capacitances can be represented as a small capacitor in parallel with the ideal inductance plus the wire resistance in series.
That means that we are in reality never dealing with an "inductance only" but with a lossy (thanks to the resistance) parallel resonant circuit which has a self-resonant frequency (SRF).
Winding the wire around a ferrite core increases the inductance. That means that in order to obtain the same value, less windings are needed which decreases the resistance and the parasitic capacitance. But the ability of a ferrite core to store magnetic energy is limited, at one point, with high currents through the coil, it risks to be saturated and the magnetic field will not longer be proportional to the current which translates into a decreasing inductance.
Last theory paragraph before we have a look onto the practical results of all this: The often cited Q factor is the relationship between the inductive impedance and the resistance of a coil. Since the inductive impedadance is proportional to the frequency ( Zl = 2 * pi * f * L ) the Q factor is also frequency depending. Thus, our real coil behavior will strongly depend from the circuit design. I'll give an exhaustive example in the next posting.

Let's take the example of these wonderful and expensive 3-pi wound 10mH coils which are used in the Etherwave Standard at the right side of the main circuit board and in the Etherwave Pro hidden inside the pitch arm. The data sheet gives us the following information:
Nominal inductance: 10mH
DC resistance: 31 Ohm
Self-resonant frequency (SRF): 700kHz
Max. current before the core goes into saturation: 60mA

Let's check first for current/saturation problems: the Ethervave's (Standard or Pro) oscillators give an amplitude of 24Vpp around 280kHz. This signal flows through three of these inductors in series with the antenna which has a static capacitance of about 8pF + 1.5pF if the player's hand is close to the antenna. That means that the antenna has a capacitive impedance of 1 / ( 2 * pi * 280kHz * 9.5pF) = 59.83kOhm. That limits the current through our circuit to 0.4mA. Thus we are far away from saturation problems which start occurring at 60mA. In general, a theremin designer will not have to care much about saturation due to the weak signal levels while designing RF power transmitters or switched mode power supplies requires to take that into consideration.

Let's attack the RF side: A nominal inductance of 10mH and a SRF of 700kHz allows to determine the parasitic parallel capacitance of our coil C = 1 / ( 4 * pi^2 * f^2 * L ) = 5.17pF. This value allows now to calculate the effective inductance of our real coil at a given frequency: Leff = L / ( 1 - ( f^2 / SRF^2)). You see that our real coil will have an inductance close to its nominal inductance only if the operating frequency is much smaller that its SRF. The closer the operating frequency comes to the SRF, the higher the effective inductance will be and even become negative when the operating frequency is greater than the SRF. A negative inductor is a capacitor, thus a real coil can behave as one under specific conditions. Let's see what this strange behavior makes in our Etherwaves in my next posting.

In our Etherwave Standard or Plus theremins, the pitch oscillators work at around 289kHz (although the Hot rodding manual states 285kHz). At this frequency, our 3 pitch linearization coils will have an effective inductance of 12.05mH instead of the nominal 10mH, giving a total inductance of 36.15mH instead of the 30mH you would normally expect.

One day I thought that if I used a single 10mH coil with a lower SRF, I could achieve the same effective value just by adding some parallel capacitance instead of havin 3 of these expensive coils. I calculated the needed SRF to be 340kHz which meant that my overall parallel capacitance should be 865.1pF. Since the coil already had about 5.1pF, I added 860pF in parallel. My pitch field was as linear as before, but proportionally divided by 3...
The octave spacing which was at about 4.5" before, was only 1.5" now! WTF?!? I thought....

Finally, mathematics helped me to understand this phenomenon.
Approaching the hand to the antenna should detune the variable oscillator by up to 2kHz, lowering its free-run frequency from 289kHz to 287kHz. What I had done before was creating identical circuits at 289kHz, computing without taking the dynamic effects when playing into consideration. So I calculated what would happen in both cases at 287kHz.

First the original configuration with the 3 coils in series: we remember that they had an effective inductance of 36.15mH at 289kHz. Applying the same formulae I found that I had an effective inductance of 36.06mH at 287kHz.

Now my alternative circuit which was designed to have also 36.15mH at 289kHz: I found that it had only 34.8mH at 287kHz.

There was much more variation of the inductance which resulted in compression of my pitch field

Short conclusion of this long story:
A) Ideal coils do not exist.
B) You can't say in general "This coil is better than another" All depends on the "real world" effects and their effect on the surrounding circuitry. What compressed the pitch field of my Etherwave could have stretched it in combination with an oscillator having another L/C ratio in its tank circuit.

This explains also why Christopher's experiments with tuned antenna circuits (Lev antenna etc.) may greatly improve linearity and tone spacing of theremins operating at higher frequencies and a higher L/C ratio while the same circuitry has catastrophic effects with oscillators at lower frequencies and a lower L/C ratio.

You see that "obvious contradictions" may be resolved and understood just by taking enough relevant factors into account. Deep insight and mathematics can replace myths!

... and since we are explaining coil myths, I'll add some more stuff:
Yes, the inductance of every coil depends on its temperature since the wire length and thickness does expand with increasing temperature. Thus the inductance of an air wound coil tends to increase with the temperature. But in a resonant circuit, we have at least one capacitor in conjunction with our coil. Since increasing the temperature will also make the dielectric layer between the capacitor plates get thicker, the capacitance will decrease and, if the circuit designer has done a good job selecting components with opposite temperature coefficients, both effects will compensate resulting in a rock stable frequency.

It's somewhat more difficult with coils having a ferrite core. A ferrite core is made of graphite powder which is glued together. At higher temperatures the glue will expand, reducing the density of the ferrite which results in a decreasing inductance. Now we have a problem since most capacitors will also decrease with increasing temperature. The frequency will go up with the temperature. You may now take special capacitors whose capacitance increases with the temperature, but they are rare and more expensive. Or you may undertake other steps to keep the temperature of the oscillator constant.

But all that is not forcibly interesting for or needed by the theremin designer. Temperature drift should not be an issue!

WHAT???

Temperature drift is an issue in RF transmitter or receiver circuits, yes. But in a theremin, we have 2 oscillators and we use only the difference of both frequencies. Thus as long as both oscillators have an identical temperature drift (this is normally and easily achieved by using the same components in both oscillators) the difference frequency will remain the same, even if both oscillators go up by several tens of kHz.

Finally, once more, we find that "common issues" or myths aren't such. We just have to dive deeply enough into our stuff in order to take enough parameters into account, so that our observations in the real world match to the theory.

Hello Thierry,

[i]“This explains also why Christopher's experiments with tuned antenna circuits (Lev antenna etc.) may greatly improve linearity and tone spacing of theremins operating at higher frequencies and a higher L/C ratio while the same circuitry has catastrophic effects with oscillators at lower frequencies and a lower L/C ratio.”[/i]

What brand of coffee do you drink? I do not believe our thinking is very far apart. We just have a different approach to achieving the same outcome. You use math and then more maths. I use a substitution box and just click through the range making observations. I would imagine our desktops look similar, soldering iron always hot, spaghetti wires running everywhere causing their own self oscillation and a flashlight because our eyesight that could once read color codes in the dark, now are no longer up for the task.

I believe hard ferrite was available in the 1930's but not soft used commonly in coils today.

I think University Students are the prowl this week for theremin projects as I have received an unusual amount of schematic requests. I do not draw schematics; I sketch, breadboard and go direct to a computer circuit board drawing. In RF a clean construction is major, something I don't like about chassis wiring at higher frequencies. I have never used any computer aided circuit design software. Seems to me it would not interpret theremin behavior properly, like what generates a better sound characteristic.

[b]My question for the Thierry Forum:[/b] Would you put your EtherWave Standard method “tuning for ideal linearity” documentation in the new forum? Maybe this would give me new respect for my EW which I will eventually donate, the way it functions now (normal) is an embarrassment. I have the necessary equipment including a forty year old TV IF sweep generator! LOL

Thierry, thank you for your extensive input.

Christopher