Help needed with NRF24 reference design

[makerbob]

Hi

I'm relatively new at pcb design, and have been trying to create my own designs fully assembled with SMT as opposed to just creating a soldershield. I'm making a wireless receiver device and have tried a design with a mifa (?) antenna before, which failed miserably. I was adviced to follow the reference design which I think I did in this one, albeit I get even poorer or none working signals with this:

With the mifa antenna I was getting signals, but muddy and with a lot of noise. With this one I am not able to get any so I am assuming there is a big flaw, but I cannot spot it when comparing my nrf24l01+ layout to the reference design? Any help or advice on how to toubleshoot would be appreciated.

Tia,

Bob

[ketiljo]

You need ground planes on both sides, connected with plenty of vias. The area under the matching network should be solid ground. Route as much as you can on the top layer to make the bottom layer a good ground plane. 

The antenna doesn't necessarily needs to look like this. It's just a simple 1/4 wave track. So you can make it a straight track instead. It's more important to keep it away from any metal. 

[makerbob]

OK great. So if I understand correctly, I should add a ground plane on the top layer as well and connect that one with many vias to the bottom ground plane? Apart from that, is there anything else that could hinder the performance of the antenna?

[ketiljo]

It can also be the crystal that's out of spec or incorrectly loaded. What is the spec of the crystal you're using? Also the value of the capacitors on the XC1 and XC2 pins?

[makerbob]

The crystal is 16mhz and the caps are 22pf. This is the specific crystal I'm using https://lcsc.com/product-detail/SMD-Crystal-Resonators_Yangxing-Tech-X322516MLB4SI_C13738.html

Also, only connecting the crystal pins, gnd are not connected.

[ketiljo]

The crystal you are using have a CL value of 9 pF. So 15 pF load caps are correct here. The GND pins of the crystal should be grounded, but this shouldn't affect the operation of the crystal