SONOFF Zigbee Dongle (CC2652P) power consumption as router


short introduction, because it’s my first post: Up until now I had very little home automation because the tools I used were cloud based and I have privacy conerns. Since home assistant solves that problem perfectly and it seems to be a lot of fun, I’m going to expand my automation efforts as time permits. All devices I have now are mains powered an thus I went with Wifi (Shelly mostly). Now I’m going to start using battery powered sensors and decided to go with Zigbee. Since I’m a technical person and prefer to have more features/possibilites over convenience, I’ll go with zigbee2mqtt instead of ZHA.

I’ll start with a very specific question. The answer would help me determine the feasability of one of my projects.

How much power does the Sonoff ZBDongle-P typically consume when flashed as router? The documentation just says “100mA max”, but I hope it’s much lower on average.

In case you are able to answer the question, please mention if you measured power at the wall (AC, including PSU) or on the USB side (DC) and approximately how many devices connect to the router and how many packets it handles per time unit.

Thank you all so much for this community, I already learned lots of interesting things just by reading this forum.


Welcome to Home Assistant!

I have 3 USB meters but the best two are MIA so I used an older one. The meter was not able to read how much power the Sonoff P Dongle was drawing and since it was displaying 0.00A, it should be < 10mA. I will repeat the test when I find the other two better meters. Also, while I don’t believe it will make a ton of difference, the dongle is on a test system that has no Zigbee devices connected to it so the dongle is likely mostly idle. Anyhow, I’d be surprised if the power usage grows significantly with a busy router… but I guess I can test this next time I bring my production HA down (same dongle there).

1 Like

@tschaboo - Ok, I pulled out the big guns… I made a USB extender cable where I could connect a 5 1/2 digits bench DMM to measure the current draw so I could provide you a histogram of the power usage. Since the device has peaks during TX, looking at a simple meter would possibly miss the maximum current draw. To cause some traffic I added and removed multiple times a motion sensor. Not like a busy network but I’d say close enough as to do those operations it has to transmit so what I expect to be different is the period of time it is at that current draw. Either way, the DMM was stable at 23mA

While setting things up, I measured a draw of 4mA when just power is provided to the stick (likely the case of the basic meter). The power draw went to 12mA when the data lines were connected but the stick was not being used by HA as Z2M was down. The draw went up again and remained stable at 23mA when Z2M was using the stick. The histogram above was captured while HA was using the stick.

Unfortunately, I cannot do this test on my production system as they are in rooms far apart and moving the DMM is quite disruptive to the workbench layout. Hope this helps anyway.