My New Year gift to TW: A new theremin circuit

Posted: 9/8/2022 3:20:17 PM
ptlabs

Joined: 8/12/2020

Thanks Thierry for such a simple design.
I built one of these exactly as described and it worked well.

This year, i decided to build a much smaller version using (hand-soldered) SMD and i'm starting to get some results.

I'm currently using the sot23 version of the 2N5484 which is the MMBF5484. 
However, these are out of production - becomming hard / expensive to find.
I tried some MMBFU310 which worked, but would die randomly - maybe needed some extra protection ?
Either way, those are also obsolete.

Can somebody help me find a suitable robust JFET that is currently in production ?
I think MMBFJ310 (J instead of U) is a current product, but a different pinout, so not tested yet.

Also, I'd like to drop the supply voltage from 6V to 5V - if possible, but i'm struggling to find clear worked examples on JFET biasing online.

Thanks in advance,
Jonathan
 

Posted: 9/8/2022 8:02:21 PM
Thierry

From: Colmar, France

Joined: 12/31/2007

Lowering the voltage from 6 to 5V is not a good idea. First, there is some voltage loss in the decoupling filters r11/c11 and r41/c41. Second, with a lower voltage, it’s not longer possible to bias the 5484 FETs correctly for optimal operation. R12/r42 would need to be lowered in value which would represent a higher load on the RF voltage dividers c12/c13 and c42/c43 thus killing oscillation. To compensate this, c12 and c42 would need bigger values which would then reduce the pitch range. 

You see, there is a fine balance with respect to different considerations… Stay with 6V.

The j310 is not good as a replacement for the 5484. It has a higher cut-off voltage. Thus, with the actual biasing, it never switches fully off which increases the internal power dissipation. I’d rather try a j309 which comes closer to the 5484.

Posted: 9/8/2022 11:53:32 PM
ptlabs

Joined: 8/12/2020

Hi Thierry, thanks for giving such a clear answer.

I'll use a little buck-boost module to get the 9V supply - I might as well leave the passive 6V reg in as its not doing any harm.

The j309 will be handy, as according to the data sheet, its symmetric so 'drain and source may be interchanged' - if this is true in practice it means i can use the j309 as a drop-in replacement for the MMBF5484 without etching a new PCB.

Today, on my SMD prototytpe, I lost another 3 or 4 MMBFU310 devices, i'm wondering if some protection diodes could be added around the oscillator fets to protect from inductor overshoot, static, or pasta attacks ? 


I'll try the J309 - maybe they'll be happier than the U310.

thanks again,
jonathan

Posted: 10/26/2022 4:46:59 PM
Bumblebee

Joined: 10/25/2022

Hey, so me and my friend decided to make this theremin for a school project due next week, the circuit looked good but once we turned it on there was no variation when we put our hands close to the antena, there was however a strange oscillating noise coming out of the speaker. Any ideas as to what we did wrong? Here is the link to a video I made of the circuit.

Posted: 10/26/2022 4:58:05 PM
ptlabs

Joined: 8/12/2020

Hi Bumblebee, 
can you describe what steps you took to test the circuit ?
step 1 would be verify 6V is present at the output of the voltage reg...

do you have access to a basic oscilloscope ?
did you try adjusting the variable cap ?

jonathan

Posted: 10/26/2022 6:39:17 PM
oldtemecula

From: 60 Miles North of San Diego, CA

Joined: 10/1/2014


Bumblebee,

All those lose wires are scary and might doom you. Use an analog AM Radio near each oscillator and you will hear two blank spots on the radio dial around 900 kHz if they are both oscillating. Then tune the blank spots so they are on top of one another to hear the heterodyned sound over the radio.

Get rid of the complete antenna & wire for this test.

Christopher

Posted: 10/31/2022 10:58:38 AM
Bumblebee

Joined: 10/25/2022

Hi Bumblebee, can you describe what steps you took to test the circuit ?step 1 would be verify 6V is present at the output of the voltage reg...do you have access to a basic oscilloscope ?did you try adjusting the variable cap ?jonathan

Hey Jonathan,
We did try to adjust the variable capacitor but nothing happened. Yes 6V is at the output, we verified that at the beginning. We are absolute beginners but from what we saw in the oscilloscope the oscillators seem to be working. 

Bumblebee 

Posted: 10/31/2022 11:03:49 AM
Bumblebee

Joined: 10/25/2022

Bumblebee,All those lose wires are scary and might doom you. Use an analog AM Radio near each oscillator and you will hear two blank spots on the radio dial around 900 kHz if they are both oscillating. Then tune the blank spots so they are on top of one another to hear the heterodyned sound over the radio.Get rid of the complete antenna & wire for this test.Christopher

Thanks Christopher,

But as strange as it sounds we have access to an oscilloscope  but not an AM radio. Both oscillators are presenting a frequency, so we are not sure what is wrong, this is our first project like this so I'm sorry if we got some basic things wrong, we are total newbies. 

Bumblebee 

Posted: 10/31/2022 1:48:21 PM
ptlabs

Joined: 8/12/2020

hi Bumblebee,
great that you have a scope, you should be able to see a sine wave at the junction of each inductor and the osc fet drain.
i get around 410kHz with an amplitude of around 4-6V.  your probes should be set to x10 mode so that they have less of an effect on the circuit under test.
-
to hear audio, the differecne between the 2 oscillators should be as small as 10Khz --> 100Hz, so one oscillator is tuned using the trimmer cap.
however, its very hard to measure this accurately from the scope's display - its around a 2% difference.
The osc freq is affected by the capacitance of the components and wiring, so making the circuit as small and neat as possible will improve your chances of success.
-
for me, this design is working well, but another design i'm trying is not.
so i've just bought a frequency meter kit off ebay, to help me see what's going on.
-
one final note about some of these designs - if the supply to each oscillator isn't decoupled well, the 2 oscs tend to sync with each other resulting in no audio.
the 100Ω res in series with each inductor, along with the 10uF from the top of each inductor to 0V is supposed to keep the 2 oscillators independant, BUT if the wiring is too long and messy then the power lines can oscillate.

I strongly suggest you give this another build and try to stick to the PCB layout in the article, i made a variant with a different PCB and found i had to add an extra 20pf across the trimmer to get it to work because the stray capacitances on my board weren't symmetrical.
good luck !

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