Thread Starter

all_about_the_dip

Joined Dec 15, 2021
2
Greetings,
I've been in a design for a while for my university. It's a security system based on sound(a glass break detector) but the point of the question is not the functionality of the circuit but the EMC issue. I've tried to reduce the EMI but I would like to ask to any expert any kind of improvement it could be done in the distribution of components and/or tracks. I'm attaching pdf files so you can properly see the pcb circuit(I'll upload the schematic in case you could find it useful).
P.S : The red tracks of the top layer are wires.
Thanks in advance.
 

Attachments

Papabravo

Joined Feb 24, 2006
21,158
Greetings,
I've been in a design for a while for my university. It's a security system based on sound(a glass break detector) but the point of the question is not the functionality of the circuit but the EMC issue. I've tried to reduce the EMI but I would like to ask to any expert any kind of improvement it could be done in the distribution of components and/or tracks. I'm attaching pdf files so you can properly see the pcb circuit(I'll upload the schematic in case you could find it useful).
P.S : The red tracks of the top layer are wires.
Thanks in advance.
Have you done any testing yet? Do you have any clue about what you are up against?
My point is, that if you can't do the same tests as the compliance agencies, you can't know if this or that mitigation strategy is working. Also you will never know what done means. Setting up compliance testing may take more time and expense than the development itself. Be prepared to open your wallet.

One more thing. It is less about PC Boards and more about power and wiring. Those are the things that boards will turn into antennas.
 

ronsimpson

Joined Oct 7, 2019
2,985
The op-amps will not produce RF noise. You should have a capacitor from power to ground at each op-amp.
The micro will cause RF noise. It looks like the micro is getting its power from C5 over very long traces. Put a small cap (0.1uF or 0.01uf) from pins 7 to 8 at the micro. See red circle.

I would mark an area from the bottom edge up past the micro and up a wase and fill it with copper connected to the ground pin on the micro. This ground plane will act as a shield. I usually make a ground plane all over the board but you could just cover the micro and 1cm around it.

It looks like red traces cross where I put yellow circles. Possible errors.
1639604303240.png
What program are you using to make the layout?

Ron Simpson
 

Lo_volt

Joined Apr 3, 2014
316
I'll 2nd everything ronsimpson said. Good advice.

I'd also suggest a 4 layer board with a ground plane. This will help to reduce interference on any high impedance nodes around your op-amps and analog circuitry.

Move your JP1, JP2 and JP3 closer together. This will allow you to twist the power supply wires together. Twisting them reduces external EM influences. Also, near the entry point for your power, place an electrolytic capacitor. 10 to 100uF should be sufficient. EDIT: one capacitor on the +5 volt rail and one on the -5 volt rail.

IMHO, your most sensitive areas in your circuit are the op-amps, especially any high impedance nodes. Keep all of your traces short, most importantly the op-amp inputs.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,167
Really, you want to reduce the EMI (electro-magnetic interference) by providing EMC (Electro-Magnetic Compatibility). The circuit and board layout don't look like the problem, but the processor module certainly might be a big problem. AND if the power supply is a switchmode type it is also a generator of RFI. and EMI. Enclosing the assembly in an aluminum box should reduce external emissions quite a bit.
 
Top