Layout and Copper pours

Hi Guys,
we’ve been working on our project for a few months now with our prototypes using a custom board designed & laid out according to the specification in the documentation. Because we were working with small numbers and the potential for changes and tweaks we have been mounting our Photons in sockets to make them easy to remove/swap to another board, however for production the intention was always to use the surface mount module and then at a later date the P1.

What concerns me however is a recent thread that had feedback from several people suggesting that copper pour anywhere near the Photon antenna interferes with its operation significantly when surface mounted and that the specified keepout defined in the Docs is potentially inadequate.

Is there a wider issue here that needs addressing or were those who reported this problem in their design simply unlucky and had some other issue that was somehow causing the WiFi problems?

Hi @Viscacha,
I have done several boards based on both P0 and P1 lately. Most of my designs have required an external antenna via a u.fl connector, but this one is using the builtin P1 antenna

I’ve excluded copper pours under the antenna itself and the radio has good range. I’m also keen on hearing more on this as I’m no expert in terms of RF. I’ve had no problems thus far though.

1 Like

I would say that the key is to have it (the antenna) positioned at the side where a ground loop is not formed.

Was asking an experienced hardware mentor about the general guidelines on using a wireless module populated on another board and that’s the advice given.

Hi @Viscacha,

Good question! Typically with anything that has an antenna or RF circuits on a PCB, there’s generally a keepout / keep-away / copper exclusion zone. You don’t want your PCB to accidentally become part of the antenna since that’ll change the transmission characteristics. There’s a note about this in the datasheets here and here:

I hope that helps!


Hi @Dave,
naturally we follow design guidelines. My concern was raised by this thread here Using Electron / Photon in high volume commercial product where @sparkly, @dfarny & @Suprazz who presumably also followed those guidelines suggested they had to do a little tweaking to get the performance they were expecting.

Hi @Viscacha,

Apologies, I should have clarified a bit more. :slight_smile: I brought this up with our hardware folks internally, and they’re going to see if we can make the recommendations in the datasheet more specific.

Although it’s normal to do some PCB tuning (ground loops, accidental harmonics / emissions, power supply, etc) before going to scale, we have a service partner who can do testing on an actual prototype in an anechoic chamber. A review like that by RF experts can take a lot of guesswork out of the process, and can help make sure you’re getting the best possible performance.

I hope that helps, sorry if I missed your point earlier. Please definitely post any issues or bugs you find in the datasheet / recommendations if you can, I’m sure your experiences could benefit others. :slight_smile:



I am sending design of our board with surface mount Photon, this design shows no problems with DHCP/WiFi, but there is only a small batch of <20 boards so far - I cant say it is 100% OK.
You can see the copper pour keepout zone, interrupted side plating…

I use L7805ACD2T voltage regulator dedicated just for Photon with 12V@2A power adapter. The track from voltage regulator to Vin on Photon is 85mm long, 1.27mm wide, 18um copper. There are 100n MLCC and 47uF tantalum caps on the 5V line close to voltage regulator.


But these recommendations have been proven to be less strict than needed by several users, and are in need of a revised recommendation.

The issue can be reduced by using the header version that spaces the antenna a bit from the pcb.

1 Like