What can cause a very short max communication range? Like 10 inches?
So I rolled my own nRF board, but I followed the datasheet layout fairly closely. The only deviation is the decoupling is done farther away in the circuit.
I then built 3 examples of this board. Two of them worked great, had good range, etc. The third board only works if it’s chip antenna is about 10 inches or less from the receiver antenna. (The receiver antenna is a sparkfun nRF breakout board, only the transmitters are custom). I’ve replaced almost every RF part on the 3rd board, the nRF chip, the inductors, caps, etc. Same issue.
It’s fairly important I get this 3rd one working, building a 4th isn’t an option in the short term.
So what can cause extreme short range in RF communications? I have a 150mhz scope at my disposal, and can drive 40 minutes to get access to much better equipment.
Any idea what points I should be comparing between the working and non-working boards to track down the issue? Thanks!
Here are the images. Note there is a cut trace and an airwire there under the board. I accidently ran MOSI through a ground via. No RF comm is going through that air wire, althrough the air wire is below the RF circuit. SPI communications are working fine, so that air wire is doing its job.
They are the component closes to the bottom of the green PCBs, they don’t have a ground plane. Also, one of these boards works great, the other is the only one with the short range. Even though I’ve replaced all the RF circuit parts one by one.
It’s hard to see details in these photos, but it looks as if there’s a surprising amount of solder on the trace that feeds the chip antenna. With one board working and one not, I’d suspect rework issues in that area. I also see that the rework on the opposite side of the board differs between the two boards, both in where the trace was cut and in the extra wire’s precise placement and length.
The airwire itself is unimportant at this point. I removed it and confirmed the issue persists. The trace cut is slightly different (same position, different levels of ‘overcut’) I’m positive it’s not leading to a logical difference in the circuit layout…
…but could it really be effecting the RF performance so drastically? The only thing touched that matters it the ground plane, Essentially the ground plane has a bunch of the copper exposed and is slightly thinner near the cut.
The more likely culprit is the component-side antenna trace. I can’t make out the details, but it looks like there’s globby solder on the trace, and I’d give that the evil eye. The thickness of a trace or plane isn’t going to matter to the electrons, but spacing or partial shorts from the solder globs would.
tecoist, that glob of solder is only structural, the trace doesn’t go anywhere. Nonetheless I tried removing it and it didn’t help. I also redid the solder joint that does connect without any differences.
Fun fact, the circuit works to about 4 inches without the chip antenna at all. With the chip antenna it goes to at least a foot. (Both cases having the antennas pointed at each other)
