Volume antenna spacing

CHILDISHGAMBINO,

maybe the vertical design will be look better?

Id put the volume coils on a smaller board and mount it vertically, maybe even on the side of the box, then (cant tell if this is what you already had in mind, but )put the pitch antenna inductors, end to end,in the horizontal rod extension.

Wonderful job you did, Ilya! .. Photoshop?

"put the pitch antenna inductors, end to end,in the horizontal rod extension." - Chobbs

There seems to be a concerted attempt to turn this theremin into an E-Pro.. ;-)

Fred.

(none of the above changes the fact that I dont like the creepy image.. But I can see its appropriateness for a theremin... It looks like its some unhappy child trying to snuggle deep into its bear suit to cover its ears! ) 

This is a project for an electronics course I'm taking.  My teacher sourced the components and designed the layout.  I designed a very similar layout but he wanted to go with a design that he has tested so I'm not too worried about the transformer.  I am having some problems with some manufacturing defects in the board.  The through hole plating didn't work properly in at least one location and I'm only getting 8.5 volts out of the positive vcc voltage regulator - something is seriously loading it down.  I have to start cutting traces to isolated the problem.  I've spent at least 12 hrs troubleshooting to get to this point.  I hope I can get it to work I'm pretty disappointed in the defective board and the fact that I can't get any sound yet.  

That vertical pic looks awesome however it is already constructed horizontally.  Nice photoshop work.

" I'm only getting 8.5 volts out of the positive vcc voltage regulator - something is seriously loading it down. " CG

If you have access to a heat sensing "gun" - well, lets just say that its one of the best tools for locating problems like this... Whatever is pulling down the 12V rail must be getting a bit warmer than the rest of the circuitry.. 3.5V drop may well be causing some component to be dissipating more than its rated for - and often even a low cost gun like this can locate such a fault.. The more expensive Flukes and the like are easier and more pin-point.. but its worth trying a cheap gun ..

Better is a thermal camera - depends whats available to you.. Cutting tracks is something you should avoid even thinking about! ... Take the regulator out and connect a heavy duty +12V source directly to the board - Whatever is pulling the supply down will cook harder, get hotter, and be easier to detect..

Horrible SMD parts!! Enough space to do everything through-hole, put a socket in for the 13700, and actually do the layout as single sided with a few wire links - but instead they go for SMD! - Prototyping with SMD is just a nightmare.

Fred.

ps.. Last debugging session here where supply voltage got dragged down - well, it turned out to be a bit sad.... LOL ;-) The PCB had been "mirrored" so the supplies to the ICs etc were all going to the wrong pins..

But the similarity prompts me to ask.. The IC is in the right way round? Its not upside down or upside down, if you know what I mean.. ;-) Damn difficult to determine the orientation on these stupid tiny IC's IMO... Fine for inside mobile phones and the like (and even these bloody phones are so small I find them a pain!) But does EVERYTHING need to go to microscope scale ?? Its getting so they dont even make new ICs in anything except SMD! Yeah - Im having a rant!! Years of learning how to use these damn parts, and just when I feel competent, they go and make the damn thins so small I cannot use them!

I finally figured out the voltage problem.  It was bad through-hole plating on the board on the positive vcc 1000 uF cap.  After thinking I fried two diode bridges with too much current and replacing the transformer(my instructor thought it got damaged)I was still having the voltage problem.  I had already replaced every component of the positive power section I decided it had to be the cap somehow.  I did a continuity check between the positive pin of the cap and the positive rail and got nothing.  All that troubleshooting and it never occurred to me to check the bottom of the hole.  I'm still having some weird voltages on the emitters of Q1 and Q2 but I think that that has something to do with the faulting plating on this board as well.  I'm not happy about this at all I've spent 12 hrs a day at school everyday this week trying to fix this thing and it comes down to board defects that I had nothing to do with manufacturing and I have to pay 80 bucks for the parts and board or I don't get my grades released or my certificate.  

SMD components are terrible.  It took days to place all those tiny parts when it would have taken me about 3 or 4 hours to build this thing with through hole components and there wouldn't have been all this bs with bad plating.  To make things worse the school didn't even have the positive regulator, several of the caps, diodes and resistors in stock when they had a year to make sure it would be available to us.  I am to be frank very pissed.  I wouldn't pay for a board that was defective from a pcb manufacturer and I shouldn't have to pay for this one.  If this board doesn't end up working I think I will try and build the whole thing on perfboard with no microscopic components and not have to worry about a crappy board manufactured by my school.

The IC is oriented correctly although I'm not sure if it is getting proper voltage and current due to the bad through-hole plating.  I ran out of time to get to measuring the inductor voltages but there is definitely something wrong in the variable pitch section as I'm only getting 0.27 volts on the emitters of Q1 and 2 where there should be 0.6v.  Unfortunately I have final exams starting on Monday and I don't really have any more time to trouble shoot this thing and I haven't seen my son for more than 5 minutes all week.  Rant over. :)

"If this board doesn't end up working I think I will try and build the whole thing on perfboard with no microscopic components and not have to worry about a crappy board manufactured by my school."  - CHILDISHGAMBINO

That's what I would recommend for a one-off project like this.  And particularly so for an analog Theremin, where heavy petting of oscillator and mixer topologies and fondling of component values is the order of the day.  (Some never get past foreplay to hard layout, the reverse of your situation).

" (Some never get past foreplay to hard layout, the reverse of your situation)." - Dewster

There is a real issue regarding prototyping on "breadboard" vs going straight to PCB - Anyone going into electronics (particularly analogue or mixed) may benefit from understanding the dynamics .. My opinion follows, based on 40 years working in electronics - from technician to consultant.. What I describe is only getting worse as SMD invades..

Management dont like EEs to spend their time building / playing with prototypes that get thrown away.. The idea that they must pay for time and parts for the designer to play with their idea, and then pay for the whole thing to be re-designed (and fully re-tested / debugged / changed) for production in SMD format - well - it galls them.

They (or the majority) think it best to produce a near production-ready SMD board, and do the "prototyping" and debugging using this..

Alas - the reality is that the costs are FAR higher by doing it their way.. but they are too inexpierienced generally (or too bloody-minded) to recognise this fact even when its staring them in the face.

I cannot count the number of SMD prototype boards I have seen binned before any useful data has been extracted.. The design manager or technical director (who generally cant even understand a simple trace on a 'scope) gets an elaborate SMD board laid out as if its going straight to production - hours spent getting it right - often in collaboration with mechanical designers so a 3d presentation of the finished product is available - real nice - real expensive.

They then, often, sell the product before the board is even populated - they ignore the EE's predictions of possible (likely) re-design.. Boards come in, are populated, and dont work.. And the impossible job of debugging / modifying on the impossibly tight SMD layout begins.

I remember designing an entirely new concept in capacitive joystics.. it was to be a "proof of concept", and I laid out a large through-hole board sprinkled with test points and links and prototyping pads so that it could be messed about with..

The technical director took my design, gave it to a different team, all test points etc were removed, and a 6 layer SMD board was produced - I got given a populated board which was unusable - About £30k was written off eventually, and when I got the project back (which was rare - usually the project is scrapped.. but in this case they had no option as they needed the product and no-one else had a clue) and did it my way, it took me a week to build my thru-hole version and have it running.. THEN the real boards were made, and they worked.. My board was 220mm * 220mm and completely unusable as a joystick - but usable to prototype a joystick.. It was compressed onto a couple of 25mm diameter circular SMD boards.

But even after the above proof that my way worked more efficiently and was far more cost effective, I still had to fight on EVERY new design, to be allowed to actually prototype before going to a pre-production prototype.

And the last job interview I went to I was presented with a question regarding my design philosophy - The technical director interviewing me was dumbfounded that I believed in prototyping - let alone prototyping with TH components.. Oh, I never got the job! ;-)

Most modern EE's come from university and dont go through real prototyping - they dont know how to, and dont realise just how easy it is to get a design operating if one can play with the running design and do modifications / optimizations.. They have been "programmed" into this "straight to production" philosophy.. And it only works because mostly they use standard circuit blocks and avoid anything innovative... But even doing this, many designs fail and many hours arev wasted due to this, IMO, idiotic aversion to prototyping.

Fred.

Troubleshooting surface mount components is horrible.  I don't understand why anyone would want to prototype with them.  I understand why my instructor got us to use them(closer to the real world as you pointed out).  I got the variable pitch oscillator working but both fixed pitch and volume are not oscillating and the only thing I can think of that could be the problem is some of the resistors were miss-labeled or in the wrong container and are the wrong values.  I'm going to start picking them off and measuring the values next week or I might just build the schematic in p-spice and see what the expected voltages across each resistor in the oscillator sections should be.  I really want to get this thing working.  I play bass and guitar and I'm really into effects and I want to see what kind of weird and wonderful noises I can make with a Theremin and my pedalboard.

"I'm going to start picking them off and measuring the values next week or I might just build the schematic in p-spice and see what the expected voltages across each resistor in the oscillator sections should be." - CG

I have a full simulation for LT-Spice you can download from Element-14 theremin group: Etherwave_FrontendSimulation_zip (all the required files - schematic, netlist, simulation card etc are in this zip.. Just unpack it to a directory and run LT-Spice on it).. This simulates two oscillators and mixer, giving the audio output.. but simulation is not instant.

I also have a simpler simulation which is just the EW oscillator simulation.. This is a lot faster and simpler.

LT Spice is completely free, and fully implemented / user supported - there is an active support group with dozens of new postings daily and new models being created and tested - it completely whacks the sh*t out of all others for pure analogue / power simulation.... and now it interfaces with Eagle PCB layout :

http://www.linear.com/designtools/software/ (this link takes you directly to the download page)

LT-Spice is really nice as an aid to debugging.. unlike most other simulators, you do not need to define test points before running the simulation - you can run the simulation and then probe anywhere on the schematic to see the expected waveform..

I dont know if P-Spice supports this (I dont own P-Spice) but all my other simulators (including Proteus which is my main tool) require one to re-run the simulation if you want to change / add test points.

Fred