-->
Page 1 of 1

What could have damaged a pin on the ESP8266 ?

PostPosted: Tue Feb 16, 2016 6:15 pm
by Zawyer1
This is my setup

Power source: 220VAC to 5VDC transformer with diode bridge and regulator (from USB cellphone charger)
Custom board: contains the ESp8266 12E, 3.3v regulator (LD1117 SOT-223) and a 10k resistor to connect CH_PD to 3.3v, GPIO15 and GND are tied together
Relay board: 5V - 250VAC 10A (SRD-05VDC-SL-C) with diode and transistor
Switch: Normal house light switch to connect/disconnect GND to GPIO13
Everything else is disconnected

Got several of these 'nodes'. Left 2 of them plugged to the mains for about a month on a remote area and another 2 plugged at my house for the same time.
When I went to check on the remote ones, found that they weren't working and after further inspecting found the flash was corrupted (one of the nodes wouldn't boot, just send garbage to the serial port, checked all common baud rates, and the other node was stuck on a reset loop by the watchdog) and the input pins weren't working anymore as inputs, output works correctly
Relays were OK, power source OK, regulator OK
Reflashed the software and everything went back to normal (except the input pins that now only serve as outputs)

What could've gone wrong?
The nodes I left home are still working perfectly
I'm aware that the relays needs 5V and I'm giving them 3.3, I think I did the math a while back and the regulator was able to supply the extra current
Whats going oooonnn :(

Re: What could have damaged a pin on the ESP8266 ?

PostPosted: Wed Feb 24, 2016 3:58 am
by scropion86
i'am interested tin similar deign as my PCB design still not complite i wan't to know if any thing i can manage to avoid before complete my design.

Re: What could have damaged a pin on the ESP8266 ?

PostPosted: Wed Feb 24, 2016 9:47 am
by Zawyer1
Well, one thing I've learned since I posted is my math was wrong, relay won't be taking more current if I decrease voltage, it will actually take less and may not activate (all my relays work well tho)

Current theory is: multiple power outages (or resets) while the switch was in close position (GPIO13 connected to GND) may have damaged the pin because it may start in OUTPUT HIGH mode before my program runs and set it to INPUT mode, thus creating a short to GND for a little while.

I don't have the necessary equipment to test except a logic analyzer arriving before the weekend :P