This is the PCB I'm almost ready to put into production. In the picture, the parts are laid on a 1.27mm grid. Every component and track has at least 10mil clearance. The smallest track size is 10mil. Ground is 13mil with 11mil clearance but I'll turn that large space on the board into a ground plane so that each ground track is substantially greater than 13mil. The thick tracks are between 0.7mm and 0.9mm in size and both have a 12mil clearance.

The parts in boxes are as follows:

The circuit in red box is the transmitter The circuit in purple box is the receiver detector The circuit in turquoise box is the receiver amplifier

In order for me to effectively use the circuit board space (instead of creating a big gap of nothing in the middle), I arranged the components so that the opamp is located more in the middle instead of to the right.

Now here is where I might have a problem which I'm hoping for an easy solution for without having to redo my layout.

If you look at the light green wire (shown in all three boxes), you'll see the connection starts at the opamp output and runs to an IC via a jumper block which is an xor input (I didn't show the part number). I also added a capacitor with a test point so that I can test my radio setup with a speaker.

In the end, the speaker will be disconnected and the jumper block will be closed so that the light green track will become live. This will be the main wire for all incoming wireless data. The speed of the data is adjustable and is estimated to be roughly 11kbps, but not more than 115200kbps.

The frequencies of the transmitters and receivers are expected to be between 300Mhz and 500Mhz.

Because the wire is within 2cm of the transmitter and receiver oscillators, would my setup be a candidate for EMI (to the point where one signal from an oscillator will jump to another wire)? If so, I would consider increasing the clearance for the wires but by how much? and if there is any other way I can reduce EMI without requiring substantially more circuit board space, let me know.

  • \$\begingroup\$ Plan on adding SMT shields around Rx and Tx separately. Otherwise LO leakage etc problems. make or buy... \$\endgroup\$ – Sunnyskyguy EE75 Apr 14 '17 at 16:16
  • \$\begingroup\$ You've said nothing about the speed or layer stackup for this board. If your at 300Mhz, have you impedance controlled the trace? \$\endgroup\$ – Voltage Spike Apr 19 '17 at 19:27
  • \$\begingroup\$ The board is catalog #630 from mgchemicals at mgchemicals.com/products/prototyping-and-circuit-repair/…. It's a single-sided presensitized board. I'm not sure what you mean by impedance-controlling the trace but I tried to make all tracks the shortest length possible to lower impedance. \$\endgroup\$ – user143136 Apr 19 '17 at 23:53
  • \$\begingroup\$ Make sure the impedance of the trace matches the impedance of the source and load. \$\endgroup\$ – Voltage Spike Apr 25 '17 at 21:38
  • \$\begingroup\$ In the answer I provided, I showed the OP how to add 2 tiny capacitors and 2 resistors as LPF, to exclude EMI from the OpAmp. The very last sentence of OP requested ideas. \$\endgroup\$ – analogsystemsrf Apr 28 '17 at 13:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy