Custom NRF52 No Antenna Signal

[ricewnr]

We created our own PCB for the NRF52 with a Johanson chip antenna but no matter what we do, we can't get a signal from the Antenna, not even a weak one to show up on the spectrum analyzer. We verified that the NRF52 chip itself works by running a peripheral example.

Has anyone else run into this problem, can you take a look at our design? We're missing a pi network but we should still see something right?

We've consulted the following questions: devzone.nordicsemi.com/.../ devzone.nordicsemi.com/.../

Gerber files for pcb layout nrf52_pcb_layout.zip

Bill of Materials

nrf52_pcb_layout.sch

[bjorn-spockeli]

@ricewnr: Can you upload your layout? (Preferably Gerbers)

[ricewnr]

@Bjørn Spockeli: Thanks for the reply! I uploaded the gerber files as edit to the original post. So far we've tried:

1. Verify that LFCLK sources off synthesize RC clock, HFCLK sources off external XTAL.
2. Verify that pin XC2 is getting a 32MHz input from the crystal (suggesting a clock rate of 64 MHz?)
3. Verify that the SPI pins work on the chip.

However, we ran into problem when we tried to program the Radio Test example on to the chip. We modified the example such that the antenna should be constantly outputting signal at exactly 2.4GHz. We tested this code on DK and it worked fine, but on our PCB it is just dead silent.

What could be causing the problem? Thank you!

[bjorn-spockeli]

Ok, but the modules are produced by a third-party? Are the two lots of modules identical in terms of hardware? Have you verified that the 32MHz crystal is correctly mounted and that the load capacitors have the correct value? Could you upload the schematic and BOM as well?

[ricewnr]

We used the official NRF52 DKs to test and prototype. We then created our own PCBs with parts we sourced from digikey and mouser. The materials are not exactly the same from the same vendors but they are of the same specifications and footprints.

I believe we did verify that the crystal is mounted correctly but we can verify it again. We'll also test the values of all of our capacitors to ensure they are mounted in the correct place.

I uploaded the BOM and schematic.

[bjorn-spockeli]

The load caps for the crystal have the right values corresponding with the CL value of the crystal, i.e. 12pF. Have you verified that the crystal is properly mounted? Have you tried running an example that uses the external clock instead of the internal? That would be an easy way of checking that the crystal is working properly.