PCB-layout for EM Theremin?

Re: Chokes..

The Bourns / JW miller 6300 series chokes are the best off-the-shelf parts I have found for antenna equalizing.. they do have ferrite cores, and are probably somewhat inferior to well constructed air coils - but they are pi wound and have low capacitance - and are FAR better than any other ferrite inductors I have tried.

I think these are probably the ones you are looking at from Digi-Key .. they are a good choice. FEC also do them.

Do not be tempted to go for the larger inductance chokes - for the EW you need 40mH - I would use 8x 5mH in series, but certainly no less than 4x 10mH... Within reason (and depending on the layout) having a larger number of lower inductance coils wired in series - spaced reasonably far apart - improves performance.. Effectively, inter-winding capacitance of each coil ends up in series with the IWC of the next coil - so the total capacitance from antenna to circuit reduces the more series wired inductors one uses ( 8 coils, each with 5pf, wired in series without any additional capacitive coupling between them, gives a total capacitance of 0.625pF, 4 coils each with 5pF gives 1.25pF .. The capacitance for these 6300 choles is about 5pF each... If one put a single 40mH choke with 5pF IWC in circuit, there would be 5pF capacitance)..

One problem with using a lot of coils is that layout needs to be carefully designed - these particular inductors 'leak' quite strongly - so you need to be sure that the poles are not anti-phase to each other, or inductance and Q will reduce.. One also wants to take care that the windings and connecting tracks between the inductors do not couple strongly to any other winding or track.. Also - The closer one gets to the antenna end, the more sensitive the circuit becomes.. It is best to have (at least) the final inductor or two actually mounted onto the antenna connector..

IMHO the best layout is to not put the EQ coils on the circuit board at all.. Simply string the inductors together in series in a straight line with at least 3cm seperating the outer coils of each inductor, then 'encapsulate' them in heat shrink sleeving to fix their relative positions, then put this assembly into a piece of plastic tubing or other non-conductive assembly (plastic trunking works well) and secure the inductor assembly into this so that it is most distant from the cabinet / board / other wires etc... [i]If one uses a lot of inductors, it is quite likely one can simply run these in a straight line from the circuit board to the antenna, which would be optimum.. The pitch circuit is the most important, so position the PCB to get optimum wiring (shortest, most direct connection to pitch antenna)- then do whatever is needed for volume antenna connections.[/i]

Now connect one end to the antenna socket with the shortest lead possible, and connect the other end to the circuit.. The wire to the circuit will be the least sensitive point (each 1pF change in capacitance will alter the oscillator frequency by less than 500Hz, a tiny fraction of the change one would get at the antenna) .. it should be as short as possible, but it is not critical.

The above also makes converting a theremin for left-handed playing an easy matter.. The highly sensitive connections are close to the antennas - move the antennas and their inductors, and everything will work.

Fred.

Hi Fred and thank you for your helpful information.

I have ordered the chokes from Digikey (GB) Alas, the 2.5 mH choke is on a 2 month back order and the whole order is on hold until it can be sent as one shipment.

I am planning to build the EM theremin following the Robert Moog 'words and music' but you suggest fitting the chokes in a long line with a 3 cm gap between the components. In the EM that would seem to place the far end of the chokes about half way along the case. If I followed your idea, but turned the circuit board end for end placing the variable pitch oscillator near the end of the the chain of chokes that should be OK?

Bob

[i]"Alas, the 2.5 mH choke is on a 2 month back order .. "[/i]

Oh dear.. Need to check my supplier on this - I am using a lot of these chokes, and was led to believe there were no supply problems - been holding back on buying them until I had verified my complete unit.. (I have bought so many parts which I expected to use, but then changed the design, that I am being cautious now.. )-:

Regarding the board placement - the actual position of the board (orientation etc) should not pose a problem.. One thing to try to avoid is any interaction between volume and pitch antenna wiring and placement of their respective inductors.. If these must cross each other, make them cross at 90 degrees.. keep them apart so that capacitive and inductive coupling is minimized.. Other than that, how the board is positioned should* not matter.

* I say "should" - because with theremins, nothing is ever certain, and unexpected things can and do happen.

My last posting's details were a bit pedantic - one does not strictly need to have the inductors in a straight line.. they could be angled in a VVVVV type arrangement to reduce their length, for example.. keep the angle no sharper than 90 degrees and you should be ok (go sharper than 90 and there is a possibility of some loss of inductance and Q - this can be useful for fine-tuning the EQ inductance though.. but thats another story! ;-)

I must just add that I have not played much with the EW volume circuit - I see no reason why what I said should not apply to this, but I have little expierience with this part of the circuit (I use a completely different method for my volume circuits)

Fred.

Hi Fred

Thanks for your previous advice regarding coils for the EM theremin.

Comparing the circuits for the Etherwave and EM theremins I note that in the EM Moog suggests a single 100 uH variable inductor in both the variable pitch oscillator and the fixed pitch oscillator and a single 68 uH varible inductor for the volume oscillator.

In the Etherwave circuit smaller (47 uH) variable inductors in series with smaller fixed inductors, 47 uH in the case of the variable pitch and fixed pitch oscillators and 22 uH for the volume oscillator.

Now I can see how this might improve the fine tuning of the oscillators because with the smaller variable inductors the end to end movement of the slug will have about half the effect of the same movement in a 100 uH inductor, but, and this is my real question, is there any other reason for the separation of the inductance into two bits?

Looking at my options, as I have already purchased the 100 uH and 68 uH variable inductors sugested for the EM theremin would you suggest searching out the inductors to follow the Etherwave circuit?

If I did adopt the pattern suggested in the etherwave what type of fixed inductor would you recommend?

Bob

There are several reasons for dividing the inductance up:

a) The variable inductors are more temperature depending than high-quality fixed ones. So dividing the variable inductance by two will also reduce the temperature coefficient by this amount.

b) The parasitic capacitance of the overall inductance will be reduced (that's also while the linearization coils are divided up).

c) A mass production reason: using more common parts reduces costs...

Hi Thierry

I am coming to the conclusion that the way ahead is to get the soldering iron hot and start sticking bits together. The idea of etching a PCB and soldering through the hole components in place to achieve a working theremin seems unlikely to work.

I sometimes read Practical Wireless and one regular contributer the Rev. George Dobbs has championed the Manhattan construction technique for RF circuits (see http://www.cliftonlaboratories.com/Prototyping.htm#Manhattan_Construction_for_RF for example of method).

I would like to try this method to prototype a theremin based on the Etherwave circuit/ EM theremin circuits. I plan to build each oscillator on separate boards so that each circuit can be verified in isolation.

Any comments on this approach?

Bob