Theremax Electronic Problems`

Hi,

We are trying to build a Paia Theremax and are having some troubles. We are implementing this on a breadboard so we can experiment with it later. So, this is for fun, not to make a completed product.

The problem we are having is that we are getting no audio output. Don't worry, we have more information than that! :-)

First, our test environment. We have tied the volumn pot high to bypass the volumn osc. per the tuning instructions in the manual. We have a 'scope, multimeter, and function generator at our disposal. We have disconnected the antennae. And finally, I cannot spell.

We strongly suspect there is a problem with the osc. or the mixers. To verify the operation of the rest of the theremin, we disconnected the pitch oscs. and hooked the function generator to C23 (the input to the filters and amp). The result is the audio output one would expect from the theremax.

The osc. seem to be working to some extent. When we measure the emitter of the osc. transistor (Q1, Q2, Q3, Q4) we get 500 - 600 KHz (which seems low, but OK), ~150 mV RMS, with a ~1.8 VDC offset. This is all pretty close to the manual's spec.

However, when we measure the output of the osc. coils (tap 4 of L1, L2, L3, L4) we get nothing measureable. It maybe that the output is too low for our 'scope, or it may just be flat. (We noticed when using the func.gen. that the theremin works really nice with inputs of under 35 mV RMS).

Likewise, the collector of the pitch amp. (Q8) shows a ~+5VDC with no AC.

We have quintuple checked the wiring, we are 100% confident that it is correct. Similarly, the fact that all four oscs. and both mixers act like this indicate that it is not a faulty component or bad wiring. Out of frustration, we went ahead and metered all of the involved resisters, caps, and diodes. They all check out.

If you are still reading this, then I owe you a beer.

My questions:

1. Is the output of the oscs. really that low?

2. Can you provide any direction on where we should go with our testing or flat-out tell us what we are doing wrong?

3. Can you provide any information on how the mixers work (D3/D5 and D7/D9) for our own understanding?

Thanks,

Ted Young

Your major problem is that you're a want-to-be electrical engineer. Bread boarding a Pala Theremax is insane.

I think you were thinking about modifying it for yourself or a friend.

First of all. without knowing your test equipment, it's suspect. Look at the specifications of your test equipment. They might lead you astray.

Always keep in mind this is a design that works and that you're fucking it up. You could have just done a few hours of soldering and it would have been just fine. No! You have to have to put it in bread board. Now it's all screwed up and you come on this forum looking for help.

You probably got transistors and transformers backwards as you can't or neglegate to read specfications and your brain dead uncle can't help you either.

Whoa....

a little "heavy-handed" for a Theremin person...

The usual problems with Theremax kits stem from bad solder joints. If your breadboard has dirty contacts, or you have breaks in your jumpers, the Theremax would not work. The coils themselves are very fragile, too.

The breadboard introduces more variables to a kit that (compared to other kits) is pretty complex. Your breadboard project would be subject to grounding/electrical interference problems, too.

If the problems with your project go beyond what I've mentioned here, then you will need someone with more knowledge to troubleshoot.

Do I get my beer, now? :)

Thank you for your replies. Yes, this is a tough kit to breadboard. I have been very careful about grounding. My cans and even my co-ax shields are grounded. But, it definately won't be as good as the PCB product.

kkissinger, I think you are on the right track about the coils. And if you ever come up to upstate NY, let me know. We will have those beers!

Before I present an update, I want to mention to anyone who is researching the Paia and stumbles on this post that these problems stem from the fact that I am doing something in left field. They have nothing to do with the kit.

So, with nothing better to do, we did yet another wiring and component check of the board. Everything was fine, but we did notice something.

It turns out the pin positions for the osc. coils as illustrated in there datasheet are a little ambiguous. In fact, I had the coils wired in reverse (pin 1 wired as 3 and 3 as 1). I had to examine the PCB included in the kit to even be sure. Once we changed the wiring around, we actually got pitch response. It was fuzzy (grounding problems probably), but tunable and playable.

The outputs of the oscs. still seem too low. I am basing that on the fact that the Schmidt triggers aren't firing.

Also, I noticed that the lug of one of the volume coils won't turn. It took some trial and error before I was able to figure a way to mount the coils on the breadboard, so I probably hosed the coils. maybe while soldering leads to them.

Since none of our test equipment is sensitive enough to diagnose the problem, my only real course of action is assuming I hosed the coils and replacing them. They are on order!

Even though the Theremin isn't working yet, it was kind of neat to be able to play it, even if only for a few minutes.

Ted Young

Ted,

I had the same problem -- that is a weak output that couldn't fire the schmidt triggers.

In desparation, I finally removed C23 from the PCB, tested it (it was OK), then soldered it back into the PCB... and voila! Turns out that I had a weak solder joint... enough to transmit some signal albeit a weak one.

Don't lose hope with the coils... you can indeed, remove the metal cases. With a magnifying glass, inspect the connections from the coil wires to the posts.. you may be able to repair it with a good magnifying glass and a drop of solder to reattach the wire lead to the post.

The coils are used "backwards" .. that is, the winding with the center tap is used as the primary. Note that the coil without the tap is feeding the diodes D5 and D3 which are designed to pass only the sum and difference frequencies of the oscillators. Note that C22 acts as a low-pass filter (that is, it sends the high frequencies to ground) -- obviously the sum of the frequencies is way above the audio range.

I spent many hours over the first couple of weeks of ownership tweaking and re-tweaking the coil settings however the result is a very clear and beautiful tone.

When I'm in your area, I'll take you up on that beer.

I AM an electrical engineer. Your response was way over the top and vulgar to boot.

Keep in mind that there were probably a few people that thought Professor Theremin was crazy back in his time.

I think the idea of breadboarding a theremin kit is pretty neat.

As long as you realize that it may cause lots of trouble have at it.

Maybe you will come up with something new!

I've even heard of people bread-boarding an entire Apple II computer - "homemade" chips and all. I say go for it, but be aware of the extra noise & hours of debugging it might require.

Check out this pic of one of Lev's original 1922 theremins - it's practically bread-boarded:

[img]http://www.theremin.nl/popov/img12.jpg[/img]

Full photo journal here (http://www.theremin.nl/popov/eng/img1.html)

Man, that's great. Thank you all for your inspiration. I am still waiting for the coils. I will keep you posted.

My main reason for breadboarding was to be able to experiment with different vocal effects. There are several mods available to enhance the voice, but I also want to try things like routing the divided pitches from Fred Nachbar's Dis*Player back in. I didn't know what to call what I wanted to do, until I was reading his history page:

http://www.dogstar.dantimax.dk/theremin/mytherem.htm

He built a Theremax (of course) and added switches to control the input of the Dis*Player back into the Theremin's voice. He called them "stops". That pretty much describes the kinds of things I want to design.

I could do most of what I wanted with the Theremax soldered on the board. So, if I absolutely cannot breadboard it, I will go that route.

Thanks,

Ted

Ted,

This is an intriguing idea...

Perhaps each "stop" could be routed thru seperate volume control circuits that are all controlled from the volume antenna yet each with an independent amount of gain. That way, as the player changes the volume, the mix of the stops changes continuously.

Now... another idea.. (I have done this on my modular synth) -- split the cv output from the pitch antenna, invert one of the paths, and run it through a low pass filter (a lag circuit) such that it is not sensitive to vibrato. Then combine this voltage with the primary path. The result is that the pitch (the DC offset) is cancelled leaving only the vibrato. Then you could use the vibrato CV to modulate your VCA's for the individual "stops".

The problem with panel mounted switches is that they are cumbersome in performance -- would be nice to have such features controllable in real time by the player. And I think that continuously variable output for each stop opens doors that mere on/off switches wouldn't.