Home
  Features
  Theory
  BOM
  Links
  Coils/Caps
  Chassis
  Layout
  Calibration
  Sounds
  Contact

 

This is the Mr. Theremin web site! This site is about a home-built tube theremin similar in design to theremins of the early past.

This site is merely a detailed account of what I had to do to build my theremin, and this author does NOT in any way recommend building a tube project because of the real risk of fatal electric shock. This unit produces 330 volts DC! Be warned!

After all, what fun is getting killed building something you picked up off the Internet when you did not understand how dangerous the technology was?

 

MR. THEREMIN FEATURES
 

All tube circuitry (the only silicone is diodes).

Uses four NOS 12AU7s and one 6X4 rectifier.

Uses four P-C70-OSC standard oscillator internal tunable coils.

Uses copper plates as the volume and pitch antennas.

Has volume, volume sensitivity, and pitch beat frequency controls.

Outputs to a common guitar jack to be played in a standard amplifier (and effect petals).

Uses a Hartley oscillator.

The design is from the Forbes theremin.

Both A/C and B+ have fuses.

Power and standby switches.

Attractive wood case (from an old Canadian Whiskey gift box).

 

 

THEORY OF THE THEREMIN
 

The 'laypersons' description.

The theremin dates back to around the very beginning of the birth of the thermoionic valve (tube) triode in 1906. In 1917, Lev Sergeivitch Terman (name commonly referred to as Leon Theremin) created the theremin while trying to create a better radio.

The basic principle of the theremin is to take two harmonic radio signals, mix them, ignore the sum of the two frequencies and use the difference between the two to generate a tone. The pitch of the tone is altered by moving your hand back and forth close to an antenna (or a metal plate).

The radio wave disruption caused by the hand in the air moving back and fourth is harnessed to produce a musical tone.

The first theremin had two antennas: one controlled pitch and the other controlled the volume. Traditionally, the pitch antenna is sticking straight up on the players right hand, and the volume antenna is horizontally located at the players left. The volume antenna is bent into a oval coil. Most theremins now use a rod antenna (like the original), but use a metal plate mounted horizontally as the volume antenna.

Some theremin designs use only pitch antennas (the volume is fixed). All of the best modern commercial theremins are transistor-based. The hands-down best theremin you can get (the holy grail for theremin enthusiasts) was the RCA theremin. They were made around the 1930's and have a unique sound that has yet to be reproduced. The RCA theremin was the first large-scale model for the general public.

The best theremins ever produced were from Lev himself for such great players as Clara Rockmore ('The Art of the Theremin' was her album). The cello-like tone was unique to the theremins Lev produced. Theremins are compared to either the Clara Rockmore theremin or the RCA for their tone and quality. Modern theremins can produce excellent sounds, and high-end theremins can emulate classic theremin sounds (such as the Big Briar by Ethervox). There is no modern theremin that is considered as good as the RCA (depending on what you consider good).

The theremin uses oscillators that are capacitor and coil based. There is momentum to remove the dependency of coils to produce the same effect. The RCA and Lev theremins used big hand-wound coils. The use of these inductors (coils) seems to not be necessary to produce the effect, but rather the sensitivity and tone quality. Coils seem to be the biggest problem with the theremin general design (in designs that use coils).

The oscillators (volume and pitch) are configured to operate at different frequencies. The operational range of pitch and volume is different. Each oscillator (pitch and volume) is comprised of two oscillators (capacitor and coil) that are harnessed to do different tasks.

The pitch section has one oscillator wired to the antenna, and the other oscillator is not. The frequency produced by the antenna-wired oscillator varies with the hand changing the capacitance of the antenna minutely (enough to matter). The frequency of the non-antenna wired oscillator does not change with the movement of the hand. It is the difference of these two oscillators within the pitch section that produces the musical note. The technical name for what it is doing is called a beat frequency oscillator.

The volume section works the same way as the pitch section, but its two oscillators are harnessed into a voltage controlled amplifier for the output. Traditionally, closer is loudest and diminishes to completely off farther away.

The first theremins were tube-based. There is a movement (small but present) within the theremin world for tube-only theremins. This is no doubt a subtle quest for a theremin that sounds like the original (or RCA). The obvious way to do this is with tubes, but also copying the original construction techniques. It is possible to create a not-fantastic sounding theremin with tubes (as the famous Uncle Howie calls 'an electronic whistle').

Note: I am not hear to say that tube sound is better than transistor sound, but I am a vacuum tube enthusiast. I also believe that the best sounding/playing tones come from tubes. This may or not be true with theremins, but is true with guitar amplifiers.

The original theremins used all tubes and large hand-wound coils. The volume and pitch operational frequencies were less than 500 kHz. They had low sensitivity (meaning more hand distance was required). Most of the components that were used in the original theremins are not available today.

It is not clear why a modern tube copy of the original has not been created: It is probably due to the fact that there are many transistor electronic engineers today and few tube electronic engineers. Simpler tube design versions can be found, but none of them is in the same league as the original or RCA theremin.

Better understanding links (in no particular order).

The Therminvox site technical description of the theremin

Art's theremin site

 

BILL OF MATERIALS
 
This is the spreadsheet (click here for it) containing the part list, prices, part numbers and vendor contact information. The part list is about 99% accurate. I tried to break the vendors into individual bulk purchases to keep shipping charges as low as possible.
This is the schematic I used (click here for it).
This is the information that Doug Forbes provides with the schematic (click here for it).
LINKS
 

Theses were links I used to provide information needed to build the theremin:

The Doug Forbes theremin site

This is the site put up by Doug Forbes. The only complaints are no finished product pictures and no way to contact him for questions.

Art's theremin site

This is perhaps one of the best theremin sites in the world! It is with Arts help that I was able to build my theremin. He is a theremin electronics expert and also has kits available for the digital builder.

Art does not condone building tube theremin projects (as I do not) because of the real danger of electrocution if something is wired incorrectly.

The Therminvox site

This site is a must have, as it contains some good technical information about theremins. The author DID build a Forbes Tube Theremin, but was not able to get the volume to function as he thought it should (he thought the volume response lacked some sensitivity). I have noted in my build that the volume DOES have far less sensitivity than the pitch (so I can confirm what Giorigio is reporting). Some elements of the design may need to be rethought to change this.

The author is a theremin expert credited with designing a first class sounding and performing analogic (not digital) solid state theremin.

The theremin web ring

I found some good sites and information here.

A non-Forbes tube theremin description

This is a site is a description of a tube theremin (not a Forbes theremin).

Theremins

I found some good information here.

Maxies pages

I found some good sites and information here.

AX84 tube amp project

This is the world famous tube amplifier project web site. In this site, Chris Hurley offers the tube DIY'er a selection of tube amplifier projects (guitar). I gained tube technology with this site. He also has a very active forum to support these projects (as well as other technical issues).

Hoffman Amplifiers

This is the famous Doug Hoffman amplifier web site. Doug is a guitar tube amplifier expert, and his discussion forum is second to none when it comes to solving your tube amplifier issues. He also is a great vendor of tube project parts.

Ampage tube projects

This site is the tube projects site that is know for its popularity with tube enthusiasts. This site is a must!

Levnet

This site is a collection of users and builders. It is here I first met the (theremin) world famous Uncle Howie. He is an expert is design and playing theremins, and is one of many helpful individuals on this site.

The LEVNET site is the only real meeting place for the worlds thermin-people. It is a mistake to not be involved with this site if you are serious about the theremin!

COILS AND TUNING CAPACITORS
 

The variable capacitors I am using are NOT 150pf. I could not find any in that range, so I bought 365pf, as this allowed me to bump my fixed pitch oscillator tank capacitance up to around 360 or so if I wanted to.

I did not know anything about these air capacitors, so I wired them wrong. They work by the case being grounded (GND) and you can pick one of the 4 places to connect the other connection (typically to the grid of your valve).

 

You can see that the darn thing is all but impossible to mount, so I used a strap retainer out of some 1/2" wide brass stock and used two nuts and bolts through the chassis to hold each one.

I used a fender washer to eliminate the wobble and you only need to tighten just enough to hold to be effective.

 

The coils were quiet unexpected from what I expected: I thought a tube theremin needed big coils, and I was prepared to wind my own. I decided against this, as I don't know anything about coil construction.

Besides that, the design was around the Antique Electronics p-C70-OSC adjustable oscillator coil available and inexpensive. I even made my own coil adjuster (I carved it out of some real firm plastic).

I included the literature that came with the coil (in case anyone was interested).

I find it interesting that these are actually transformers (they have a primary and a secondary).

 

Note you read these coils looking at the bottom (opposite of the little mounting 'crown' they have).

Note: The use of the center tap may not be correct (but is working in this design).

CHASSIS
 

 

I used an aluminum chassis, as I knew I would put it into a cabinet. While a steel chassis would have been better, I was able to cut all of the needed holes in one day by hand with the aluminum.

I put some little feet on it, because I thought it might work without a case (it turns out that it needs the wooden case).

Note: If you want to have something not hidden in a box (like mine), use a steel case. My aluminum case would flex with the weight of the mounted components (stressing the chassis).

The big drawback to steel is that it does not nearly cut as easily as aluminum. I spent three days cutting holes for my guitar amplifier (with its heavy steel case).

LAYOUT
 

 

The layout mentality was to keep distance between the A/C and the coils, and keep the coils as far apart as possible (from each other).

I used point-to-point wiring, and only used PCB for the power supply and star ground isolation points.

I broke the tube plane (a no-no for guitar amps) to bring the tubes close to where they were needed to eliminate inevitable wire crossing. I kept them 3 inches (about 8 cm) on center. The heater wires were drill twisted.

POWER SUPPLY

I did not use the specified power supply (it did NOT work for me at all). I am using an unregulated power supply and trim the voltage with a 50 watt variable resistor (sounds hillbilly, but it works). A regulated power supply is the way to go if you can. I put 2 extra stages of RC filtration AFTER my B+ fuse, and grounded the filters with the ground point used by V2 and V3.

My operational voltage is 130 volts currently, and this does not change with different hand positions (except by +/-15 volts in extreme).

I did burn up a 5 watt 1K resistor testing the power supply (with no load) as to not risk my 12AU7s. Later, I installed the specified power regulation system (The Forbes zener-shunt system), and weird things happened (nothing worked). I noticed that this damaged one end of my 50 watt resistor!

I had to use the other end and the wiper (it is as if the other end was burned up). That was it for me, as the big resistor power supply is serviceable (so my power source is unregulated).

Note: An unregulated power supply with less than half the voltage specified seems to work optimal with no ill effects.

Note: An unregulated power supply does appear to work with this design. My power supply is nothing more than a standard guitar amplifier type. I also do not see any advantage in a very high working voltage for this design. I am using 130 volts and I get serviceable results.

"THE CICADA SYNDROME"

When I first assembled the theremin, it seemed to work, except there was a buzzing characteristic to the sound. I thought that was normal, until it was pointed out that this is NOT normal.

It would seem that there are small design flaws in the Forbes theremin, but it appears they are NOT insurmountable. It appears it is up to the builder to adjust some values of components to compensate on the layout they selected.

Here is a few suggestions to make the cicada sound go away:

(1) - Clean B+. Use lots of RC filtration.

(2) - Use a plate resistor on V1, V2a, and V4 that is 470K. This worked for someone else, but not me. I restored the plates to their original specification (no resistor on V1, V2a, and V4).

(3) - Change the 220pf condenser on each coil to 47pf or 56pf. This worked for me. I used 47pf.

(4) - Use a high voltage ceramic disk capacitor (0.01uf or 0.1uf) on each plate as a high frequency by-pass filter. I did not try this (it might work for you).

(5) - Try grounding the bottom of the 1M grid resistor INSTEAD of connecting it to center tap with the 220pf capacitor value (there is a possibility that using the tap like this is a misuse of the coil). If this cures the problem, that means there is unwanted feedback through the 1M resistor (let me know if this is the case! I did not try this).

This is the sound if it is NOT working properly. Art Harrison calls this "the cicada syndrome"

STAR GROUNDING

I used a total of 5 ground points.

(1) - A/C has its own ground point, right as it enters the chassis.

(2) - The power and rectifier has its own point. The stock filter capacitors are grounded here.

(3) - The volume has its own. Both oscillators are grounded at this point. I also ran the 1N4148 diode to the volume ground to further isolate noise (or so I believe).

(4) - The pitch has its own. Both its oscillators are grounded here.

(5) - The mixer, amplifier, and input are grounded at one point.

All of these points are grounded to one point.

Even for a guitar amplifier, this theremin is quiet.

ANTENNAS

I used two copper plates (8 inch long, 4 inch wide, 1/16 inch thick).

I could NOT get any form of a rod antenna to work (I almost had to touch the antenna to get it to work).

I could not get the theremin to work when the antennas were close to the chassis. I instead had to install them about 10 inches (about 23 cm) above the chassis mounted to the wooden case. Even using tested insulated mounting brackets to the chassis did not work (in case you were wondering if I tried it).

The plates I am using give good sensitivity and I recommended for that reason using plates. I think it would have looked cooler to have the rod-type antennas.

The antennas I tried to use were standard off the shelf Radio Shack replacements for a cell phone base. This reason alone may be why I could not get them to work (perhaps the wrong type of antenna for what I needed).

CALIBRATION
 

This is how I calibrate it:

(1) - Set the coil slugs at about half way (You don't have to do this, however).

(2) - Set the pitch beat frequency and volume sensitivity knobs around the center of their travel (half way).

(3) - Set the voltage (ala 50 watt variable resistor).

(4) - Take all of the tubes out (except the rectifier tube).

(5) - Set the fixed oscillators base frequency:

(5.1) - PITCH: With just V1 in, power on, warm up, and then turn theremin on. I set this to 575 kHz on the AM radio, and find where the oscillator is at by tuning the coil (with a non metallic tuner). I have found this sound is the sound of the coil being tuned at its loudest point (it sounds like rubbing violin strings with bow rosin on your fingers). I then repeat this process with V2 as the only valve in.

(5.2) - VOLUME: With just V4 in, power on, warm up, and turn on. Notice this is the same procedure as 5.1 (that's because it is, except with V4 and V3). I tune it to about 700 kHz.

I think its best to start with the fixed oscillator coils first, because the range is more limited and is thereby the limiting factor in frequency determination for the pitch (or volume) oscillators.

Note that with an unregulated power supply, the current draw is different with just 1 oscillator tube in verses all of them. I started out at 250 volts nominal with all tubes in, and at the end of tuning, my operational (plate) voltage is just 130 volts! (The Art Harrison 125 theremin uses a just 50 volts DC).

This means I have to adjust the variable capacitors with the new settings of the variable capacitors (the volume one can change the voltage from 190 volts to 310 volts in its entire range). And after I set the voltage, the oscillators need retuning. Fortunately, this cat and mouse system seems to be quite forgiving, and I am getting great results.

(6) - Now comes the fine tuning.

I thought that was it, and it appears that there is a degree of fine tuning by ear to set the volume sensitivity and correct operation of the pitch. This is done by tuning the fixed volume coil to get the working range dialed in, and testing different settings of the volume sensitivity control. The pitch is so sensitive, I fine tune it by bumping it lightly until I hear it making a low tone so low I do not hear it (you have to activate the volume when you do this).

It turns out that it is possible for this theremin to be slightly out of tune and it will not function as you would expect. In fact, it also seems possible to get different types of theremin response with different tunings.

Update: I find that I don't have to fine tune it so much as when I had originally written this. All I have to do is to adjust the variable volume control to activate with hand proximity, and adjust the pitch the same way (why I made it so complicated in the beginning, I don't know why!).

I have discovered it is possible to have two operational modes:

PITCH ONLY

This is where I only have a pitch control (volume becomes nonfunctional). The sound is great, but REAL quiet. The working range starts about 10 inches to almost 1/4 to the plate. The sound is like an electronic tone.

Note: This is NOT a correct operational mode of this theremin. It happens to be one way it operates, and is not correct.

VOLUME AND PITCH

This is where the volume and pitch both work, and the sound is more like a voice than an electronic tone. Also, the volume can get REAL loud. I do get little strange sounds with different volumes, and I can find a working volume with minimal effects (for a clean pitch).

Note: I find that the volume is best set at 50% and left there and never moved. I play it through a 100 Watt Marshall amplifier through its clean channel and get great sound and power with little (or none) internal noise from wither amp or theremin.

Note that these two modes were discovered early, and I am sure that many more interesting combinations exist with factors such as correct working voltage, different antennas, and setting pitch and volume at other working frequencies.

I set my volume frequencies HIGH (around 735-775 kHz) and my pitch frequencies LOW (around 550-600 kHz). This way, a lot of tuning will not cause heterodyning between the volume and pitch sections because of overlaps.

I do find that the volume and pitch mode is the way the theremin wants to be used in. Finding the spot with the pitch adjustment control, as the point where it works is the loudest spot I can find activity on (many places on the dial seem to work, but are not as loud as the 'one main' spot on the dial). My only drawback is that fine tuning is difficult (the gradients are very small fractions of a millimeter).

 

SOUNDS
 

This is the CICADA SOUND.

This is the sound if it is NOT working properly. Art Harrison calls this "the cicada syndrome". The theremin appeared to be working normally (volume and pitch control), but the sound is not right.

In my case, this was caused by the 220pf condenser capacitors were too high of a value (possibly causing coupling and unwanted oscillation). I corrected this by changing the value to a 400 volt 47pf ceramic disk capacitor.

Pitch only .

When I move the calibrated pitch control LEFT about 3 mm, the volume control does not operate, and overall volume is quieter. The working range is about 8 inches total (about 15 cm). No setting on the volume knob does anything. The sound is like a transistor theremin I once heard (cool SCI FI sounding).

Volume and pitch.

When the pitch control knob is set where it was calibrated at, both volume and pitch operate. This is one quick example I did. Notice it can get louder. The working range of the pitch with this tuning is about 2 feet (about 66 cm). The volume is about 8 inches (about 15 cm).

Please note that this is one of the first times I ever made any sound with this theremin and recorded most of the shrill end of what this theremin can play (in error). In fact, this theremin sounds great at LOW lows and has an excellent operational range. I can make mine sound voice-like, and not whistle-like. In the beginning, I liked to make the shrill whistle scream sound with it, and it shows on this recording.

Build notes  

 

I do not notice any problems with crosstalk: the grid wires do NOT seem to be sensitive to plate and A/C wiring (unlike a guitar amplifier).

The grids on each oscillator tube MUST have a negative voltage (this means the oscillator is working). My variable volume oscillator grid was shorted to ground for a long time and the problem was that it was not floating ground with its 1 Meg resistor.

No matter how I insulated the plates from the chassis, the theremin stopped working when the plates were too close to the chassis. That is why they are mounted about 10 inches (about 25 cm) and mounted to the wooden case.

My copper plates were about $5.00 each at Menards (an American DIY home improvement chain of stores). Plates offer the BEST response (I have been told).

DC filtration seems to be the number one problem with this design. I do not seem to have this problem (as I have added 2 extra filter stages within the B+ line). I am using 3 20uF/500 volt filter capacitors, and have added 2 22uf/400 volt filters in the B+.

The coils were real easy to obtain (I have heard from many people they could not find them).

The 10nf value on the by-pass filters is 0.01uF (this is a real common value).

Notice that the schematic shows the variable capacitors rotor wired to the ground of the coil. I did not run a wire from the variable capacitor to the coil ground. Rather, I relied on the chassis grounding of the capacitor and the separate ground of the coil to somehow meet. They must, as it works. I did isolate them and run the wiring as it showed, but there was no difference (so I changed it back to chassis mounting ground).

The cathodes of V3a and V3b are coupled with a 10K resistor. This one changes the voice of the theremin. I thought it might be interesting to replace this with a 25K variable resistor and call it the 'voice control'.

I added a B+ fuse (1/2 amp fast blow) after the stock filtration and before my two extra stages I added. This saved me when I accidentally touched the cathode and grid of a tube together. A new fuse later , I was back in business.

I used 400-600 volt rated capacitors, and originally used 1 watt or better resistors. I did use some 1/2 watt values (the 10K cathode resistor for V3a and V3b). I also have one for my V3a plate I plan to replace as I only ordered 1 470K/1watt resistor.

Uncle Howie recommends using a 1000-1500 Ohm resistor for the V3a and V3b cathodes (instead of 10K). Then use a 20 mfd/150 volt capacitor at the junction of pins 3 and 8 and put the negative end to ground. This is supposed to richen up the tone and help flatten the response curve.

Art Harrison has (well) designed a pitch and volume theremin. It is called the 126 theremin. This may be the theremin to build unless you want to build the Forbes theremin. His theremins are based upon the Colpitts oscillator (different from the Hartley of the Forbes).

I do have an example of this theremin being used in a recording at my MP3.COM site. See "The Succubus" at:

http://artists.mp3s.com/artists/278/michael_r_rogers.html

(Updated on 02/09/03)

 

 
Michael Rogers Web Information 2002