calling all Keppinger tube theremin builders

Posted: 12/7/2009 8:47:29 PM

From: Brooklyn,NY

Joined: 12/1/2009

Hi everyone,
I was planning on building one of the excellent tube theremins designed by Mark Keppinger. After studing the schematics and internet for a bit, I cannot find the value off C11, the panel mount trim cap, and the inductance vaules of the oscillator coils. The parts list only lists what gauge wire over what size form how many times. It would seem helpful to know the value when winding those monsters.
anyway, I know that there are only like five people around who have the knowledge that i seek...I hope they are listening.
thanks in advance.
Posted: 12/7/2009 10:48:58 PM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as . ...................................

Joined: 12/7/2007

I cant help you much, if at all - however..
If you know the frequency the oscillator runs at, and you know the antenna capacitance, you can calculate the aproximate inductance required for the equalizing coil..

The antenna capacitance is the real 'fly in the ointment' here, as it can be really difficult to determine - in fact, the "easiest" way I have found of determining antenna capacitance is by using a modified Theremin which has no equalising coil, and high capacitance sensitivity, and known tank inductance and capacitance - I can then null the beat frequency by varying the reference oscillator, measure the reference oscillator frequency, and from this determine the antenna capacitance.(this approach also has the advantage that, if you put the "test Theremin" in the location that the 'real' board will go, connected to the same leads etc, stray capacitances etc will be "seen" by this test rig - and I can usually get a usable result which is within the 'real' oscillator's tuning range)

The "test Theremin" can be quite basic - I originally used a Cmos 555 timer (TS555C running at about 500kHz with antenna disconnected - see Feedback 555 osc in this spreadsheet ( .. you must use a TS555C not one of the other CMOS 555's, as this one is far more stable and operates to far higher frequency).. In fact, this is all you really need if you have a good frequency counter - just measure the change in frequency when you connect the antenna..

You might find this .xls spreadsheet ( useful if you cant get the information you need elsewhere.. This spreadsheet computes both the oscillator frequency based on capacitance and inductance values of the oscillator, and gives an inductance value for the equalising coil, based on the oscillator frequency and an antenna capacitance you supply... It is far from perfect, but I produced it because I always get my calculations wrong when I use my brain - My PC has a higher IQ than I do - and its not too bright.... 8-(

Posted: 12/7/2009 11:43:15 PM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as . ...................................

Joined: 12/7/2007

One final suggestion - it is something I did a lot when experimenting with big Eq coils..

If the instructions say "500 turns", wind 400 turns, then take a 'tap' off (Twist a few cm of the wire together and leave this sticking out..) then wind another (say) 50 turns, and take another tap.. do this all the way up to 600 turns..

The more 'taps' you take, the more accurately you will be able to select the right value - you can connect to any 'tap' and effectively increase or decrease the number of 'active' turns.

When you have selected the 'right' tap, you can cut and remove any unused windings (you dont really need to do this, but it does reduce inter-winding capacitance).

Also - put the taps at the oscillator end, NOT at the antenna end... I have found that the closer one gets to the antenna (if one has a linearly wound "Honking" coil) the greater the chance that even tiny irregularities in the winding will have strange and unexpected consequences on the performance and linearity.. Irregularities in the first 25% of the coil closest to the oscillator seems to have little effect.. In fact, I have found that the easiest way to tune a EQ coil is by making the coil about 25% less than the expected required inductance, and then using cheap ferrite inductors at the oscillator side to trim the total inductance.
Posted: 12/7/2009 11:58:51 PM
Jeff S

From: N.E. Ohio

Joined: 2/14/2005

I sent Mr. Keppinger an email to see if he can provide some values, assuming Philip doesn't have them already.

Check back in the next day or two.
Posted: 12/8/2009 1:48:54 AM

From: Brooklyn,NY

Joined: 12/1/2009

wow,thanks a lot guys.
Im not sure of the ocillator frequencies...
other than the parts list (, schematic (, and Phillip's construction journal (,
there isnt a whole lot of information out there. If Mr. Keppinger does reply, Id be much obliged. Maybe we could add a construction notes page with the schematics.
Posted: 12/8/2009 2:01:58 AM

From: Brooklyn,NY

Joined: 12/1/2009

I used the "add a variable inductor" trick with my
scratch built etherwave (
the large coil measures around 30mH andd there is a small 10mH variable under the shelf between the big coil and oscillator.

edit: this photo is older and the variable coil actually appears on TOP of the shelf in this photo. I've since moved it to the bottom.
Posted: 12/8/2009 8:07:25 AM
Jeff S

From: N.E. Ohio

Joined: 2/14/2005

Mr. Keppinger was apparently up at 2:30 AM this morning and provided this information.

The front panel trimmer (C11) "is going the be between 7pF and 20pF, although values such as 3-14, 12-20, etc. have also worked satisfactorily". He used an air trimmer and removed metal vanes to achieve the value needed.

You have the the wire size, form size, and turn count of the air coils.

He said he would have to look through his notes to find the coil values he measured on an inductance bridge or measure the coils he has.

He also says, "However, the Q is as important as the value."
Posted: 12/9/2009 7:14:09 AM

From: Canada

Joined: 8/1/2008

The Q can make anything work simply by changing the universal constants and the laws of thermodynamics.
Posted: 12/9/2009 10:13:39 AM

From: Croxley Green, Hertfordshire, UK

Joined: 10/5/2005

Posted: 12/10/2009 2:20:30 AM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as . ...................................

Joined: 12/7/2007

[i]"The Q can make anything work simply by changing the universal constants and the laws of thermodynamics" - StarPort[/i]

LOL! ;-)

In fact, Q can be easily understood without deep understanding of "the universal constants and the laws of thermodynamics".. "The Q" should never be spoken about.. The "Quality Factor" can be talked about, and is Q, but not "The Q" (?

Enough nonsense!

Q is simply and directly a function of the effective series resistance of the inductor - In plain english, what this means is that if the resistance of the wire used to wind the coil is high, Q will be low -

Thicker wire has lower resistance than thin wire - gold plated ECW has even lower resistance, pure gold is best! ;-)

If the wire used is as specified, and the coil is wound as specified, Q should not be a problem.. Unless the wire is thinner (>R =
My limited expierience leads me to believe that the inductance value (in combination with antenna capacitance - ie - they must be carefully matched to achieve resonance close to the oscillator frequency) is by far the most critical factor in achieving linearity.. and that the next most important factor is the (inter-winding) capacitance, and the physical positioning of the coil with respect to other 'grounds'..

Adding small ferrite 'tuning' inductors does reduce over-all Q (these inductors tend to have quite low Q values, and quite high capacitance) but I have found this had no noticable effect.. It was quite interesting when looking at the insides of the Tvox (which has good linearity) to see that quite a small ferrite inductor is used for equalization..

Tapping the coil to allow selection of the optimum inductance, has no disadvantages.. it does not decrease Q (particularly if you 'tinned' the taps so that the extra wire did not add extra resistance.. but this is being absurdly pedantic!) does not increase capacitance, and allows the option of changing the oscillator frequency if one ever needs to in future, and selecting a different 'tap' to retain optimum linearity.... After having hand wound many coils (I have now built myself a winder) I learned that putting a LOT of taps was well worth the extra effort.

Big coils look great - and when accompanied by glowing tubes produce wonderful instruments..

But I do not think there is any substantial advantage in big coils.. Or, put it this way.. You will get better linearity from a low Q ferrite inductor which has the absolutely right inductance value, than you will from a large coil which is slightly out of tune.

I have never built a tube Theremin, and all my observations are based on solid-state... I cannot see how this could make any difference to the above - But, with Theremins... Well, sometimes I think maybe "the Q" uses them as a means of transportation or something....

You must be logged in to post a reply. Please log in or register for a new account.