From breadboard to PCB

Posted: 9/14/2026 10:13:34 AM
ElGreco

Joined: 9/14/2026

Greetings everyone! I registered today but i watch the forum some time now and you guys really helped me to get some ideas and technical theory about theremins. My main goal is to make an easy to produce purely analog theremin with materials found in any electronics shop. So lets get to the topic...

My main headache now is how to go from breadboard to PCB. I have made 3 pcb's so far 2 of them weren't working at all which is pretty reasonable considering it's my first pcb design. The 3rd one i got it to work but the problem is that all my antennae sensitivity was gone, from a pretty good range i had to almost touch the antennas to get a sound and the volume antenna was even worse i had to touch it to work. So lets get to the current design 

https://ibb.co/9mpS87SH

The oscillators i go with are typical colpitts ones powered by current source with lp filter at the front to avoid injection locking as much as possible. Next the outputs are buffered via jfets and mixed with a gilbert cell. After some filtering and amplification the signal is going to the VCA. The volume oscillator is going to a discriminator followed again by a jfet and then to some transistors which their purpose is to invert and scale the dc voltage feeding the VCA.

This is the current layout of the PCB

https://ibb.co/DHN41mrW

So my main question is how to avoid the failure of the previous pcb. You think it might be a circuitry disfunction or purely a matter of pcb design?   

Posted: 9/18/2026 9:53:01 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Welcome ElGreco!  

I only took a brief view of your images, but have these observations / comments:

Schematic:
1. It's not clear to me where your antennas connect?  Are they in parallel with significant capacitance?
2. The schematic is rather low resolution, so it's difficult to tell crossovers from connections.  
3. Use a conventional voltage regulator for the entire circuit, voltage stability is very important for oscillator stability.  You might use a separate regulator for each oscillator to isolate them via VCC coupling.  A ferrite bead before the regulator would help decouple at higher frequencies.

PCB layout:
1. Looks like KiCAD?
2. It looks like both sides are ground, with many connecting vias?
3. The VCC traces are extremely thin and snake all over the board.
4. Connecting traces are too thin as well.  Use a bare minimum of 0.25mm width here, much wider if space allows.
5. Consider making one side VCC and the other side ground.
6. Be very careful with copper fill around capacitance sensitive sections of the circuitry.  Make keep-out sections as necessary, or don't use fill at all.

Posted: 9/18/2026 4:43:44 PM
ElGreco

Joined: 9/14/2026

The antennas connect in collector of each oscillator.
The software is KiCAD 
When you say capacitance sensitive you mean the tank capacitors? Should I make a keep out zone as well?
Thanks for the advice

Posted: 9/18/2026 5:21:13 PM
ElGreco

Joined: 9/14/2026

A more close view of the schematic

https://ibb.co/6Jv0vv4B

https://ibb.co/XMCSWnC

Posted: 9/19/2026 10:13:54 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"When you say capacitance sensitive you mean the tank capacitors? Should I make a keep out zone as well?"  - ElGreco

I mean there is significant capacitance from sensitive nodes to the surrounding copper fill.  The main thing to use copper fill for is to establish very low impedance ground and VCC.  For a 2 layer board, I would probably make one side ground and the other VCC, and would do most of the interconnect on the VCC side.  You have to watch out that the pour doesn't get choked off by the interconnect, though this is easier with through hole components.

I would definitely make the entire areas occupied by the oscillators copper fill keep-out zones.  And I would beef up the interconnect widths.  Also, try to keep the coils physically separate, particularly those that are heterodyning together.

Posted: 9/19/2026 11:42:17 AM
ElGreco

Joined: 9/14/2026

"I mean there is significant capacitance from sensitive nodes to the surrounding copper fill.  

Yes I understand what you mean. The reason to make the top side VCC as i understand is to avoid the long traces and make more room for the signal traces?

Posted: 9/20/2026 12:07:03 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"The reason to make the top side VCC as i understand is to avoid the long traces and make more room for the signal traces?" - ElGreco

Yes, it's mainly to lower the impedance of VCC by making it copper fill.  Another trick is to do horizontal interconnect on one side and vertical interconnect on the other.  This keeps the copper fill on both sides better connected to itself and more contiguous.

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