When I took apart the SONOFF Zigbee 3.0 USB Dongle Plus I noticed that the metal casing is completely painted which is not ideal for shielding purposes.
Sonoff calls out “Aluminum housing effectively reduces signal interference from peripherals” as a benefit of its case but I would say it falls short given there is no electrical conductivity (at least in the 2 that I have) between the case and the ground of the PCB.
In a preliminary attempt I made to ensure grounding of the case it appears that my highest LQI scores have increased by 10 to 20 points reaching 200 LQI for the closest devices which I had not seen before. I can not be certain that the increase in LQI is due to my grounding improvement but it may become more obvious when I work on it some more. I purchased some thin copper tin foil tape to see if I can improve the shielding further. The case doesn’t open up except for the tine side openings so it is really hard to work with.
For now all I did was remove at least some of the paint in the rail where the PCB slides it but the board is so loose I doubt it makes much contact. Notice the board has grounding rails on each side top and bottom (if only the case were better designed…). I also removed the paint where the SMA and USB connectors may make contact with the side panels. I removed paint on the back of those panels on the edges where the screws go, and I removed some paint where the screw head comes into contact.
I have a few ideas on how I could possibly improve the shielding with the copper tin foil tape but I will only share details if it turns out to be feasible.
RF Engineers out there? What are your thoughts?
This dongle is an RTX radio so shielding it from all other noise / RF in the environment should help at the very least with its ability to receive a clean signal (better SNR). The shielding should also help with limiting the RFI spewed by the dongle.