Completely newbie planning to build EM theremin

Hi, Fred and Dewster!

To be honest, I'm rather surprised that Miller still makes those types. The only market is antique restoration, ham radio, and theremins as far as I know. The broadcast industry might also use a handful.

Hamfests, which are basically flea markets for radio geeks, are sometimes a source for decent parts, but often hit and miss.

To find one in the UK, RSGB. In the US, ARRL. Other country's, just Google ham radio. (BTW, see what happens when you type my screen name into the call sign search window at the ARRL site.)

For new parts and NOS....

Fair-rite

Antique Electronic Supply

Rf parts

Amidon

I've used products from the above company's with good results. Amidon and Fair rite specialize in core material, and if they don't stock what you need, they can custom make it.

Antique Electronics is just amazing. They have some of the most obscure parts I've seen. They even have RCA theremin tubes. And sockets. And AF transformers. Very nice.

RF parts is a big name in the ham radio parts world, and always high quality.

"At times I wonder if buying an old RF coil winder might be a good investment" Fred.

My father has a rather nice machine shop and I'm contemplating building a coil winder.

I'm also thinking about something I call "little big coils". Concentric forms with about a half inch of space between them, starting with about .75 inches diameter in the center, and ending with about 3 inches on the outside. The center form might have a cut down AM loopstick core. I think I can get quite a few millihenry's out of just a few inches of height, with an SRF higher than a full sized coil.

I'm going to be a bit scarce for a while. I'm working a ton of extra hours during the holiday season. It's 2:30 AM local as I write this, and sleep is calling, with a bull horn. The good news is I'll have a bit of loot to toss into the TW theremin.

Yawn....

I'm about to finish my order in mouser. But I'm not sure of the quality of the capacitors I should choose for C19 & C20. The PDF just says it must be made of aluminum. I'm going to buy two EPCOS B41827A7228M000 but i wonder if i should choose Nichicon UKL1V222KHD instead.

EPCOS B41827A7228M000
- Price: 0,522 €
- Tolerance: 20 %
- Ripple Current: 1810 mAmps

Nichicon UKL1V222KHD
- Price: 3,32 €
- Tolerance: 10 %
- Ripple Current: 1580 mAmps
- Low Leakage

Such differences matter here? Worth paying more?

Thank you.

The Epcos should be fine - The main really important aspect of these capacitors is their voltage rating - never apply more than 70% of the rated voltage to a capacitor is my rule of thumb - And you should be well below this.

These capacitors are pre-regulator smoothing - no need for anything expensive.. I think it is the low leakage which ups the price of the other - not something worth paying for in this application.

Fred

"To be honest, I'm rather surprised that Miller still makes those types."  - w0ttm

And for $1.85 a pop at mouser!  I'd charge at least $10.  Thanks for the link to Amidon!

"I'm also thinking about something I call "little big coils". Concentric forms with about a half inch of space between them, starting with about .75 inches diameter in the center, and ending with about 3 inches on the outside. The center form might have a cut down AM loopstick core. I think I can get quite a few millihenry's out of just a few inches of height, with an SRF higher than a full sized coil."

Funny, I've thought of something like that too.  You could also control capacitance by sectioning the windings - jumble wind a doughnut, then jumble wind another next to it, and so on, kind of like Miller does. The loopstick core is inherently an "air gap" type, so there shouldn't be much dependence on temperature, correct?

For my applications it seems an SRF of about 3x higher than resonance is OK, but going below that negatively impacts sensitivity.  And the presence of a flux concentrating core minimizes any "metal detector" behavior.

I'm using inexpensive Xicon 1kV NP0 5% ceramic disk caps in my AFE tank, should I consider something different?

"The loopstick core is inherently an "air gap" type, so there shouldn't be much dependence on temperature, correct?" Dewster.

I think so. Fred would know more about this than me.

I'm just looking for a home built compromise that will work.

I love the big coils, but they are, well, BIG.

I think NP0 is a good choice for your tank. They are found in almost every tuned circuit I have, but I stand ready to have my sleep deprived opinions challenged :)

"I think so. Fred would know more about this than me." - w0ttm

I think you are wrong about this! ;-)

The dark art of inductors is not my forte - Even matching the temperature coefficients of capacitors to inductors for thermal stability is on the edge of my comprehension, LOL ;-) (I do have a set of N750 capacitors which I sometimes use to trim drift, but thats about it)

My main scheme of thermal compensation has been brute force - stick a temperature sensor on the board, derive a voltage from this, use this voltage (via a trimmer) to vary the inductance of a saturable reactor in the antenna tuning circuit (I sum this correction voltage with the tuning voltage and any other modulation signals)..

Effectively, the above gives an adjustable compensation, and I dont need to worry too much about the thermal charactaristics of the ferrites - particularly as my reference oscillator is locked to a crystal.

Provided that the temperature sensed is quite close to the temperature of the ferrites (as in, small currents and low dissipation in these) its ok - As soon as one starts having ferrites warm up disproportionately as a result of current through them, it all goes to hell.. My biggest problem was with the saturable reactor itself, as the control current through this heated the damn thing.. All the above R+D led to a neater solution - I found a low current way to implement the saturable reactor.

I also found that using a 6300 series inductor (doing surgery on it, and seperating the 3 coils) I was able to make a superb saturable reactor - Alas, this operation takes time and is fiddly -

Fred.

Hi

Comparing EM and EW schematics i see that C10 and C11 were supposed to be polarized (tantalum) in EM but they are shown as non-polarized in EW. That leaves the question: which material should i choose?

I also found that C29 and C30 are new in EW, and since there is no part list i don't know their voltage and material. Fred, you told me about the importance of the voltage rate of C19 and C20 and that i should not use anything expensive there. So, because i guess it's the same here i think 50V and ceramic should be enough. But i'm far from sure, so please correct me if i'm wrong.

EDITED: I'm seeing that the polypropylene film capacitors might be a better choice... :-?

Thanks!

Hell - My longwinded posting seems to have dissapeared!

As far as I can see, 470nF ceramic capacitors should be fine for all those capacitors - they are effectively HF decoupling caps, so require low series inductance and resistance. Tantilums were probably changed to ceramic due to cost / availability.

Fred.

Hi again!

Some days ago i was choosing a suitable box for my transformer when i had an idea: moving the power supply circuit to the same box so the heat generated by the voltage regulators won't directly affect the rest of the circuits. Do you know if there is any counter-effect by doing this?

I'm choosing a MC7812CTG and a MC7912ACTG as they are 1A voltage regulators (following what Fred says here). Also they have less load and line regulation. <THINKING_OUT_LOUD> Those figures appear measured in mV in mouser but reading about load and line regulation i found they should be measured in a percentage of some kind... i guess the figures shown in mouser are taken from the data sheets and there they are measured at specific conditions (voltage, temperature, etc)</THINKING_OUT_LOUD>

Anyway, forget the last paragraph.

BTW: Happy new year to all the theremin-lovers out there! :D

I think you want the regulators near the circuitry they are supplying.  Long lines can pick up noise and the inductance can possibly oscillate.  Unless you need lots of current, I've read that the smaller (~100mA max) three terminal regulators are often less noisy.