Own-built Theremin hums

Posted: 11/24/2010 3:50:01 PM
clemfon

From: Hamburg, Germany

Joined: 11/24/2010

Good evening from Germany,

wow, never thought of a Theremin-forum, fantastic!

I'm in the 10th grade and for a competetion a friend and I build a theremin.
We've soldered it successfully, but when we turn it on, it just hums. You can hear sometimes a difference, when you move your hand to the antenna, but something's wrong. It's not a clear noise, it's just an ugly humming.

The amplifier is also made on our own, but he is tested, so this may not be the problem. It must be something in the Theremin itself.
Here is the circuit diagram we used:

Circuit diagram (http://netzspannung.org/learning/iswdh/theremines/images/ET_out.gif)

The ELKO stands for electrolyte capacitor, and all the other capacitors are ceramics except of C3, which ist a film capacitor. Maybe that's the problem, because it should be a ceramic capacitor?


Well, thank you for your help and I hope, you've got some good ideas...
Please write in an English, which is quite easy to understand,

Clemens
Posted: 11/24/2010 5:48:17 PM
Thierry

From: Colmar, France

Joined: 12/31/2007

Hi Clemens,

welcome on the thereminworld forum!

You linked the circuit board layout (Platinenlayout) instead of the circuit diagram (Schaltplan). If we could see the latter, diagnostics were somewhat easier.

In ever case there is no worry for the 1uF capacitor, may it be a polyester or ceramic type, its coupling function in this place remains unchanged. So the problem has to be elsewhere.

BTW: From where did you get this circuit? Although I know lots of (more or less useful) theremin circuits, I never saw this one before. Do you know somebody who realized it successfully? Did you make a (LT- or P-)Spice simulation before?

EDIT (somewhat later):
I figured out how it should work: Parts of the 4069 are used as fixed oscillator with R1, R2 and C1 as frequency determining components, the 4046 internal VCO is used with R3, R4 (variable), C2 and the antenna as variable oscillator while the VCO control voltage (Pin 9) is set to maximum. The internal XOR gate serves as mixer whose output signal goes through C3 and R5 (variable) to the amplifier.
What kind of antenna did you connect?

I hope my English and German is understandable...
Kind regards from France
Thierry
Posted: 11/24/2010 11:05:24 PM
FredM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007

[i]"The internal XOR gate serves as mixer" - Thierry [/i]

LOL! ;-) .. Seen that trick done before! I think sometimes the idea of 'exposing' the fact that one is using an XOR is sometimes a reason .. But it doesnt really make sense (does it?)

Thierry - I could not find that damn schematic! Would you be kind enough to put a link to it in your last (or following) posting ?

Fred.
Posted: 11/25/2010 5:54:26 AM
Thierry

From: Colmar, France

Joined: 12/31/2007

The link is in Clemens' initial posting, but I'll repeat it here: circuit board (http://netzspannung.org/learning/iswdh/theremines/images/ET_out.gif). I also found a superficial project description for beginners in (very faulty) German: website (http://netzspannung.org/learning/iswdh/theremines/)

BTW: Fred, if one is already using logic gates as mixer, the XOR is the most efficient variant, no reason to "LOL"... I object only to the use of logic gates in the signal path of so called "professional" theremins (if you know what I mean...)
;-)
Posted: 11/25/2010 11:26:58 AM
clemfon

From: Hamburg, Germany

Joined: 11/24/2010

Thank you for your replies.

The diagram I've linked is exactly from this German website, which Thierry posted.

As antenna we've used this type of antenna:

Antenna (http://www.m-w-modellbau.de/out/pictures/1/HS1115.jpg)

We've connected it to the board with just one solder point.
I don't know how an antenna exactly works, but doesn't there have to be more than one soldering points, so that the power can go through the antenna?

Would you think, that some solderings aren't ok, or???

Thierry: How can I do this simulation?

Clemens
Posted: 11/25/2010 11:35:58 AM
Thierry

From: Colmar, France

Joined: 12/31/2007

I would connect the antenna with a 15cm wire. I can't see at distance if your solderings are good.The circuit which you are building is very simple (technically), so that slightest tolerances of the components, bad contacts etc. may trouble it. The simpliest and quickest thing in order to get it working would be a detailed diagnose with a frequency counter and an oscilloscope.

So I invite you to send me your circuit board by post in order to get it fixed and working for free (out of the shipping costs). If you are interested contact me by email:
theremin(at)tfrenkel(punkt)com

Concerning the simulation: There are free softwares like LtSpice or PSpice which can be found on the internet. After installing one of them you may draw the whole circuit by picking, placing and wiring the components with the mouse, run a simulation and see what happens at different points of the circuit. But in order to do so you should have understood before how your circuit works and what are the eventual problems of discrepancy between simulation and reality...
Posted: 11/25/2010 1:21:19 PM
FredM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007

(one)Reason it doesnt work..

Fault on layout.. 4069 pin 10 is shorted to +V supply. Only INPUTS should be tied to a supply..

Excessive current being drawn from IC. (an inverter with its input high will have its output low.. but as the output [pin 10] is connected to high (+V) the full current from the supply is available as a heat source! .. This IC should be getting hot (or it may just have bolwn up! - in which case it wont get hot, it just wont work!)

Pins 13,11 and 9 should be connected to +V .. NOT pin 10!

The IC will need replacing. The fault you describe can be fully explained by the layout error.

My advice: Always work from a schematic, not from a layout - Check the schematic against data sheets for the IC's and components used.. If you had drawn the schematic from the layout (I am assuming you only had a layout) you will see this kind of fault instantly. I started at the 4069, and the fault was instantly identifiable - I stopped there.. So there may be more errors.

One other thing I notice.. There is no local decoupling on any IC's.... You NEED a capacitor between the +V and -V pins of EVERY IC to ensure reliable operation.. The layout is an inductive mess! put 100n ceramic decoupling capacitors with leads as short as possible (position the capacitors on top of the IC's if you must).

Other engineers think I am OTT on this matter - but 90% of the really baffling problems I have seen on electronic equipment comes down to the 'simple' matter of inadequate supply decoupling.

[i]"BTW: Fred, if one is already using logic gates as mixer, the XOR is the most efficient variant, no reason to "LOL"... I object only to the use of logic gates in the signal path of so called "professional" theremins (if you know what I mean...)" - Thierry [/i]

I agree absolutely with the above! But why waste a 4046 PLL on an application like this? one 4069 (hex inverter) and one 4077 (quad XNOR) would have done the required job just as well, cheaper, and simpler! .. 3x XNOR forming a reference oscillator, 1x XNOR as mixer, 3x inverters as variable osc.. or whatever way one wished to combine them.

In fact, I suspect that having both reference and variable oscillators built using the same parts will keep drift lower than having one oscillator based on the 4046 VCO and the other based on 4069's - the oscillator circuits being different is likely to mean different temperature dependent behaviour..

If one was to have two 4046 ICs, one for reference and one for variable oscillators, and use one of their XOR phase comparators as a mixer, this might make sense (particularly if one used the CV input to the VCO's for tuning).

Fred.
Posted: 11/25/2010 2:07:46 PM
Thierry

From: Colmar, France

Joined: 12/31/2007

Oooops...

I have checked thoroughly the 4046 section and found it was ok. But I must admit that I had no closer look at the 4069's unused gates since I couldn't imagine that one would make an error there... sorry!
Posted: 11/25/2010 2:19:38 PM
FredM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007

[i]"I have checked thoroughly the 4046 section and found it was ok." - Thierry[/i]

Great! Together we have now checked the whole circuit... You did the more 'difficult' bit, I just peeked at the simple bit, found the fault, and stole the glory! LOL! ;-) Life just isnt fair!

Fred.
Posted: 11/25/2010 2:27:25 PM
clemfon

From: Hamburg, Germany

Joined: 11/24/2010

Cool,
you seem to have found the fault, thank you very much. I will try the solution as early as possible.

Have I understood everything right:

Just Pins 13, 11 and 9 should be connected to V+ and not number 10? What about number 14, which is shown connected in the circuit?

Should I (after solving the V+ problem) replace both IC's?

Where should I put the capacitors for the IC's exactly in?


And thank you again, now I've some hope again, that it'll work! You're great.

Alors, je vais essayer utiliser mon francais :)

Merci beaucoup pour votre aide!
Vous ĂȘtes fantastique!

You must be logged in to post a reply. Please log in or register for a new account.