Page 1 of 1

Shielding against inductive interference

Posted: Wed Dec 09, 2020 5:09 am
by anttok
Hi, I'm a relative beginner, so please bear with me.

PROBLEM: I am experiencing POWERON_RESET when riding and esp powered electric skate board on street, particularly going over manhole covers and other metal utility covers. I think its very possible that some of these covers have a small current through them from poor grounding, or contain switching devices just below.

SETUP: I have a devC v4 board installed in an abs enclosure under the deck. It's connected via two pairs of twisted wires, about 10cm long (rx/tx and 5v/gnd) to a vesc motor controller, which is in turn powered by a 33v pack and runs a small hub motor.

TRIED: made wires short as possible, installed a 440uf, 100uf, 1uf, and 0.1uf caps between Vin and GND. 5k pullup resistor from EN to 3v. (It appears that after installing the caps the reset reason changed to TG1WDT_SYS_RESET under same circumstances)

What should I try?
What type of shielding is most effective in this situation?
What frequency range am I likely dealing with here?

Re: Shielding against inductive interference

Posted: Wed Dec 09, 2020 5:49 am
by ESP_Sprite
It sounds very unusual for manhole covers and other metal things like that to have a voltage on them, let alone a current going through them. A more likely explanation would be that there is some mechanical connection making intermittent contact when you're driving over them, in my opinion.

Re: Shielding against inductive interference

Posted: Wed Dec 09, 2020 2:30 pm
by anttok
First, thank you for the quick reply.
A mechanical disconnect was my first thought also. At first both components experienced simultaneous power reset, which I tracked down to mechanical switch bounce on the battery pack bms, which I shorted (for now) with a wire. The vesc is now stable (I compare the reported runtimes via usb). I also soldered the rx/tx and power wires directly to the pins of the devboard, and hot glued the jst connection on the vesc side to its header. I can't imagine what could be lose.

On the other hand con ed utility company identified over 300 sites in the city with poor grounds after a woman and several dogs were electrocuted by electrified metal plates on the sidewalks, not to mention the mains wiring and equipment for traffic lights, subway etc that may be under them (I don't actually know what's down there). I agree its unlikely, but I can't think of anything else, and I'm yet to replicate the problem by riding over sidewalk panels, rumble strips, or shit pavement in the park etc. (I'm really trying!)

Re: Shielding against inductive interference

Posted: Thu Dec 10, 2020 4:57 am
by becorey
Can you compare if it happens when you're riding vs if you just slowly move it over the manhole cover? Separate out the vibration from potential EMI, then you can be more sure of the cause.

Shielding against any interference would involve a big grounded metal plate covering the electronics. Then the noise couples into GND instead of your circuit. Cables, use shielded cables. PCBs, use smt shields. Or put the electronics in a metal enclosure and ground it.