Interrupts being triggered by AC Load when pull resistor is connected

[Tushar Kanani]

Hi,

I am also facing the same issue.
Does anyone have any solution to overcome from it?

Thanks

[fritsjan]

Same problems here.
I use a button with an external 4k7 pullup, connected to a twisted pair wire (1m long).
I need to use interrupt-falling to detect the button (because in the main loop there are some internet things going on which can take too long to detect a buttonpress in time).

Somehow I get false triggers once in a while, which is strange because I use a pull-up AND twisted wire... So should be pretty stable...
When I used a 10K pullup, I got false triggers when I just touched the metal housing of the button... (touch sensitive it seems)

Also when switching an AC load nearby (or even when switching a big one in my house which is not nearby), I get false triggers...

I use a adafruit huzzah32 dev board.
I put an extra cap over 3.3V and GND to help the 3.3V power supply a bit. This seems to help a little bit, but not all.

Just a thought:
Could it be that the 3.3V LDO (AP2112-3.3) in the huzzah32 dev board just can't deliver enough power (600mA)?
And that when the internal supply drops too much the interrupt gets trigered?
I do not have any other major external items attached to the board, so power draw is only from say 3 pullups and from the esp32 itself.
It connects and checks the web every 20sec so it might be that that is drawing too much power?

[TonVeenhof]

Goodday. I experienced the same problem. An interrupt triggered unwantedly by switching on the AC light on the same AC circuit. The interrupt is attached to a push button, on a rising event.

In my development setup I simulate the button by briefly connecting two cables. One to GND, the other to a data pin. I can see the interrupt fires because I added a Led for debug purposes without being attached to the Serial monitor.

I have defined 10 interrupts. The only interrupt that fires unwantedly when switching on AC light is the one defined on the pin on which I have attached a cable for testing purposes.

The problem of the unwanted interrupt disappears when I disconnect the cable from the data pin I use to simulate the pushbutton. No other changes. I can only assume the source of the problem is static and not electric, or else the problem would persist on all interrupts, and also within the ESP32 without cables attached.

So: I will deliver the ESP32 and its external devices like buttons in a cage of Faraday, metal enclosure.

Hope this helps