I’ve Gotenna Mesh permanently connected to the power outlet via USB charger.
It worked for ~24+ hours and got discharged unexpectedly, so i’ve started to investigate.
Apparently i can reliably trip GFCI interrupter by sending pings from this Gotenna.
In ~10 attempts i’ve got 6 or 7 trips. I’ve moved Gotenna further away from outlet, but it’s still within ~2-3 ft.
Has anybody observed this behavior?
Any mitigation recommendations? (can’t move to another outlet).
Longer cable, ferrite filter on cable, better USB wall adapter?
Thats super weird.
goTenna Mesh’s RF transmission is definitely not significant enough to interfere with a GFCI breaker.
Can you see if this still happens with a different USB charger?
I’ve tried swapping USB power supply (Innergie cube for Blackberry, so not complete junk), with same effect.
Also tried replacing USB cable, no change. But to my surprise switching to the other outlet on the same GFCI (it’s a 2-outlet box) removed the effect completely, with either USB power supply.
Could be defective or mis-wired GFCI; would poke more next time.
I can reliably trip a GFCI with 5 watts and I know others that can do the same at their homes with similar power. I have never tried 1 watt. Some wiring will be more susceptible to different frequencies. Try these tests.
Leave the Gotenna where it is but unplug it and unplug the charger. See if you can trip the GFCI. If you can a ferrite bead won’t likely help because the wiring in the wall is likely acting as an antenna… In that case an extension cord or longer USB cable might help. If it doesn’t trip RF is probably leaking out the ground wire of the Gotenna. That is the case with many devices and a ferrite bead will probably help.
Try adjusting the way the Gotenna is placed in relation to the socket.(Vertically / Horizontally / above / below / Left / Right) The RF pattern when the Gotenna is lying flat may be very different from the RF pattern of a Gotenna standing straight up.
A ferrite bead is really cheap so give that a try. It really can hurt. Place it on the USB cable close to the power supply.
USB charger. Leave the cable attached to the charger.put an RF choke on the USB charger cable if it doesn’t have one. I honestly don’t think that will help but it’s a cheap test. $1 on fleabay. It would not surp
Just did my own tests. I can easily trip the GFCI I tested if the Gotenna is close to the outlet. No need for it to be plugged in. Shielding electric wiring or distance from the outlet will likely be your best bet.
I was surprised today when I set my goTenna Mesh on a wall shelf, sent a test message, and heard a loud click. Thinking it was coming from my Bluetooth speaker as just normal RFI, I tried sending more messages, but no more click noises.
Turns out the goTenna had tripped the GFCI sitting about one foot away. 100% repeatable, even with nothing plugged into the outlet. However, turning the goTenna 90-degrees so it was parallel to the outlet instead of horizontal eliminated the nuisance trip.
This outlet is an older 15A GFCI dating back to the late 80’s. The newer 20A GFCI outlets in my kitchen don’t seem to be affected.
Might try a different power supply just to see if the issue is duplicated with it. I suppose the issue could be caused solely by RF, but with only 1 watt maximum ERP, I tend to think maybe not?
GFCIs also age. That yours is an early one makes that a factor, too, hard to say.
Since the goTenna was unplugged, the GFCI trip would have to have been induced entirely by RFI. Whether it was picked up by the wiring in the wall, or the outlet is another matter. The outlet is due for replacing anyway.
I’ve had GFCIs trip from other noisy appliances, such as an electric car charger. I’ve also had GFCIs detect real dangers, such as a cracked heating element in a portable hot tub, and similarity in an espresso machine. They do serve an important function, to be sure.
Yeah, I missed the part where the GTM was unplugged and the GFCI was exhibiting this behavior.
Very doubtful anything is out of line with the GTM to cause this. Yes, probably was RFI that prompted it, but the reason why it reacted to the RF likely has to do with aging or some other fault with the GFCI.
The GoTenna uses a SI446X transceiver IC. I’ve built a couple prototype systems with that chip, and besides the link budget appearing to be less than advertised, the chip had an annoying tendency to mess up any hardware debugger hooked up to a microcontroller when it transmitted at anything over the minimum gain on the transceiver chip. I always thought it was because I had ground loops or in general a bad ground plane, but this whole tripper a GFI suggests that it is a SI446X transeiver IC issue. It’s fascinating that Silicon Labs and Gotenna have both got it past FCC certification when the SI446X seems to have weird behavior when the gain is cranked up.
Makes me wonder ff the GTM is an older one. I haven’t evidence, but have heard the Full metal RFI cover is a recent intro/reduction. Among the reasons for needing it is limiting spurious output.
I’m also curious what it’s doing when something more substantial than a ping goes out? A Shout gives more info, a lengthier, “thicker” message…does this trip the GFCI from further away?
Just as an update, none of my newer GFCIs can be tripped from transmissions from the goTenna Mesh (and, these gTM units were purchased in the last week from REI). On the older outlet that is affected, the gTM needs to be within 12-18" and there seems to be pattern of live and dead zones where the outlet picks up the RFI.
I did find the an article on ARRL website discussing nuisance-tripping GFCI and AFCI from radio transmitters, and that newer breakers/outlets are better shielded.
GFCI outlets are supposed to trip between 4 and 6 mA of sensed current imbalance between Hot and Neutral. This is detected using a single current sense loop that both Hot and Neutral lines go through. Stray RF could certainly induce a tiny current in that loop sufficient enough to trip the GFCI. I imagine the newer GFCI outlets have some sort of shielding to block outside interference.
I just had this happen, today, with a goTenna Mesh unit that was close to the GFCI…about 4 inches away. Others in the same room (kitchen) did not trip. I could replicate this behavior just by the goTenna mesh being in the proximity of that GFCI—even as far as a couple of feet away. So far, it’s only been that one particular GFCI that has reacted, but I didn’t stick the goTenna Mesh close to any of the other GFCIs in the house—girlfriend giving me “the hairy eyeball” about tripping her GFCI with a gTM.
My gTM was purchased from Amazon and arrived about 8 days ago.
Well it is a radio and radio transmissions are known to occasionally trigger older GFCIs.
At 1 watt and using a form of short-acting burst mode transmission (a CB is 5 watts and people tend to gab forever on them) the form of goTenna Mesh interaction is virtually the least power that can potentially trigger this in GFCIs. If you notice this happening, then simply move away a few inches, or better a few feet, and it will most likely “solve” the problem.
There’s nothing “wrong” or “broken” on the goTenna Mesh when this happens.