EM theremin build

The LM386 would be dandy as a headphone amp. There are much better chip amps if you want to drive a speaker.

A very useful site for solid state amplifier construction is HERE.

After having to put the theremin build on hold, it's time to start it up again. I've started putting the circuits together, I have a few of the dual mini boards from radio shack that I'm using.  I just had a quick question for you guys regarding the boards and the unused holes after the required soldering is done: Would it be worth the time to scrape off any of the copper around the holes that aren't used? Would this help at all with unwanted capacitance? Or just be a waste of time?

Also, is there a minimum distance to keep the antennas apart? I'd like to start constructing the box itself at home (most of the electronic work is done at school in my spare time).  I have a pretty good idea how much room the circuitry will take, but do not want to accidentally make the box too small.

Oh, and as far as to the circuits themselves, I have started putting together the Fixed pitch oscillator, and the Pitch tuning onto one of the squares. I think the volume oscillator and Volume tuning will fit on another square,  then the variable pitch oscillator will go on its own square, and the voltage controlled amplifier and vca processor on one square, and probably lay the power supply out across two squares, depending on how the caps fit. 
From the article, the only spacing concern I've found mentioned is keeping the VPO and the FPO a couple inches apart, which will be easy enough to do since they're on separate boards.

Does this seem like a reasonable approach? 

Also, is there anywhere in the circuit that should have shielded wiring?  Possibly the connections between the circuits? or the connections to the antennas? Or any and all wires? 

Again, thanks for any and all responses. It's great having a forum like this that has active, knowledgeable, friendly folks.

Luc234 asked:

" I just had a quick question for you guys regarding the boards and the unused holes after the required soldering is done: Would it be worth the time to scrape off any of the copper around the holes that aren't used? Would this help at all with unwanted capacitance? Or just be a waste of time?"

IMO, a complete waste of time.

"Also, is there a minimum distance to keep the antennas apart?"

The antennas themselves wont interact (if correctly tuned) But the closer they are, the more hand movements over the volume antenna will affect the pitch (your volume hand, as it moves up/down, doues not direcly affect pitch - but there will be some horizontal movement from the volume hand, and this will affect pitch.. the further the antennas are apart, the less this effect will be) I like my antennas a long way apart - 50cm minimum.

"... they're on separate boards....  Does this seem like a reasonable approach?"

Yes. The only problem with this approach IMO is signal and power interconnection - Add 3  extra decoupling capacitors and connect one between GND (0V) and +V,one between GND (0V) and -V, and one between +V and -V, on each board, close to where the wires enter the board (any good small decoupling capacitor - ceramic, 100nF to 470nF .. One wants to prevent the inductance of the wiring from causing problems).

"Also, is there anywhere in the circuit that should have shielded wiring? "

Yes - Any audio and dynamic LF control signals should, IMO, be screened if taken from board to board - So anything after C23, and I would possibly screen the volume control signal from R14 to R35.

Antenna leads and HF leads should not be screened.

The biggest interconnection problem is the oscillator output signals - Try to keep these short - the connection to the VPO (via C2) is capacitance sensitive (as is the connection to the FPO) - Not extremely capacitance sensitive compared to the antenna, but still a potential source of trouble.. I would build the detector circuit onto one of the oscillator boards, and configure the layout so the boards could be pushed close together and a short wire connects the other oscillators output - one can then take the mixer output (across C23) to the VCA board and use screened cable for this..

I would change R24 to 10k and C23 to 2n2 on the mixer, and have a 10K and 2n2 between pin 3 of U3-A and ground - As in, split the filter components so that the Z on either end of the connection are balanced, and pin 3 is never floating if the connection is removed.

The better way is to buffer each oscillator with an emitter follower before taking the signals to the mixer - but thats a seperate subject I dont want to get into.. 

Fred.

Fred wrote: "I would change R24 to 10k and C23 to 2n2 on the mixer, and have a 10K and 2n2 between pin 3 of U3-A and ground - As in, split the filter components so that the Z on either end of the connection are balanced."

Excellent idea, Fred! And perhaps experiment with 2 x 2n7 or 3n3 instead of 2n2 which makes the sound less sharp or nasal when the waveform knob is in the left half and the brightness knob in the right half.

"... I have a few of the dual mini boards from radio shack that I'm using."  - luc234

If it's the ones I'm thinking of, I've learned not to use those.  They're phenolic and the traces peel off with the application of a little heat / mechanical force. 

I haven't bought these but they look really nice: http://www.adafruit.com/products/591

BusBoard stuff looks nice too: http://www.mouser.com/Search/ProductDetail.aspx?R=BR1virtualkey57130000virtualkey854-BR1

"If it's the ones I'm thinking of, I've learned not to use those.  They're phenolic and the traces peel off with the application of a little heat / mechanical force." - Dewster

Funny that - I use phenolic board all the time, and dont have any problem with it - perhaps its down to the manufacturer.. Sure, FR4 is much more robust - but at least 3x the price - and difficult to get in small panels.

It should also be remembered when soldering pad board (FR4 or SRBP) that you must keep the temperature low and complete the joint as quickly as possible - there is no sinking of the heat applied (no thermally conductive area to dissipate the heat, as one gets with strip board or PCB) - IMO, you MUST use good tin/lead solder (60/40 or better 63/37) - you CANNOT use lead free solder for this type of board IMO.

IMO, spend the money on the best solder - it is far more important than the board material.. Also, you really do need a thermostatic iron - Yeah, you will pay £35 for the cheapest, but this is not a luxury - it is absolutely essential.

The board I use most often is available from several UK suppliers..

http://www.rapidonline.com/Tools-Equipment/Circuit-Prototyping-Boards-300749

http://www.rapidonline.com/pdf/34-0602.pdf 

IMO, the boards you linked to are probably better in terms of quality - but they are not pad boards as far as I can see.. they are designed more for IC's than for point-to-point wiring.

It is always a bad idea to rely on the pads to provide much resistance to mechanical force - component leads should be shaped to give mechanical stability, and tight-fitting pins used to connect to any large components or components subject to stress.

But there is a HUGE variation in quality of boards from different suppliers - This is even true for FR4 boards - I have had chinese manufactured PCB's where the copper lifted off the FR4 with just a little heat - And I have had SRBP boards (phenolic) which happily survived complete re-works without any problems.

Perhaps the Radio Shack SRBP's are just crap.

Fred.

"Perhaps the Radio Shack SRBP's are just crap." Fred.

They are. The copper falls of if you look at it. I've given up on them and usually use plain, copper less perf board for the simpler projects.

Another question regarding the tubing for the antennas: The build guide suggests 3/8 soft copper tubing, but I don't have that handy at the moment. Would there be much of a difference if I used 1/2" or 1/4" instead?

Anyone know if the C24 from this build is supposed to be 3300pf or .01 uf? The parts list says 3300 pf, and I'm guessing that's what it's supposed to be (seems like a random number to accidentally type in) but the schematic itself shows 0.01 uf. 

Also, most of the inductors shown seem to be 3 section RIF chokes.  I've been looking at the data sheets for several and they don't really have much info in them, and haven't been able to find much on google about them.  I was just curious what makes these different from other inductors that are just one section.  I'm assuming the 3 sections are one winding around a ferrite core and just spread out? 

Again, thanks for answering my questions, I'm enjoying putting this together so far. I have most of the circuit built, but haven't soldered anything yet, just waiting until I'm sure I've laid everything out correctly. 

"Another question regarding the tubing for the antennas: The build guide suggests 3/8 soft copper tubing, but I don't have that handy at the moment. Would there be much of a difference if I used 1/2" or 1/4" instead?"

If you use 1/2" (slightly thicker) you would probably need to reduce the length a tad, if you use 1/4" (thinner) you would need to increase the length a tad.

In my opinion, longer, thinner antennas perform better than fatter shorter ones.. and the exact length required depends on a lot of strange factors.. But I would go for 1/4" and multiply the length so that the total area (imagine the antenna is a flat plate - and the original Area is 3/8" multiplied by the length - you want the same area for whatever you change the diameter to)

Alternatively, make a 1/4" antenna the same length (or a little shorter) as specified for the 3/8", and use an adjustable scheme like I do here: http://www.thereminworld.com/Forums/T/28592/coil-tuning?Page=1

- But the simplest is probably to go out and buy some 3/8" tubing!

I havent actually done the sums - you may find that 1/4" and 1/2" are fine, and there is enough tolerance in the antenna eq circuit that it makes little difference - As I understand it, the EM had poorly calculated antenna inductances anyway.

Fred.