GPIO output signal degraded over time

tsctrl
Posts: 30
Joined: Tue Dec 08, 2020 1:37 pm

GPIO output signal degraded over time

Postby tsctrl » Sun Jan 16, 2022 12:19 am

i have a esp32 wroom that working properly for few month but the signal on rx2-gpio16 degraded now. the gpio only used to a data signal, just with simple voltage check on gpio. the signal suppose to read 0.9v but it now only at 0.6v (depends, but it drops half the amount). i have try to reheat the chip and it works for few day then the issue comes again.

what could be the cause and how to mitigate the issue? is it this gpio have internal pull up or down for me to configure or any setup required on my circuit to make the gipo not breaking?

edit: is this is potentially because of excessive current draw attempt to the gpio pins? if that might be the case, how do i protect the gpios from it. maybe resistor in series? i have 1a (still wondering what value suitable) fuse protecting the 3.3v circuit from LDO which powering the esp.

thanks!

ESP_Sprite
Posts: 8926
Joined: Thu Nov 26, 2015 4:08 am

Re: GPIO output signal degraded over time

Postby ESP_Sprite » Sun Jan 16, 2022 1:23 am

tsctrl wrote:
Sun Jan 16, 2022 12:19 am
the signal suppose to read 0.9v but it now only at 0.6v (depends, but it drops half the amount). i have try to reheat the chip and it works for few day then the issue comes again.
Wait, what? How is it OK for the voltage at a (presumably output) GPIO to be at 0.9V? What are you doing to that poor ESP32? (As in: I'd really like to see your schematic.)

tsctrl
Posts: 30
Joined: Tue Dec 08, 2020 1:37 pm

Re: GPIO output signal degraded over time

Postby tsctrl » Sun Jan 16, 2022 7:04 am

it is data signal and not general io, the voltage is depend on the how much the data out, i guess it is leverage since it pumps lots of details, the 0.9v is the time i measure what is current active state the gpio is outputting doesnt matter if state is high or low.

from the working other gpio pins that outputing the same value, it was measured as 0.9v. but the failing is only 0.4v. i did not hook up osci to be exact the max signal voltage, is just what multimeter measure. i not sure if my understanding is correct though, or else i not sure why the gpio only outputting less than 1.5v max during the transfer. the vcc is 3.3v and no other circuit connected to the gpio pin during measurement.

if gpio output voltage is not high enough during high without any connected components what could be a solution or thing to look out? ignore the reading values from the problem, to take as it is outputting half of what it should. what could be the problem either configuration, codes or circuit wise?

i guess is due to long excessive current draw from the connected components that i did not controls. but still just a thought. i did not have the datasheet to measure what exactly the draw on the components but to be save what and how do i protect the gpios?

and to simplify, do a PTC on an 3.3v let say 0.5a is enough to power the esp with wifi and stop over current draw from gpio to protect the mcu and the gpio? this might be back on what is my requirement, but what could be a common value for me to put in to be at max usage and provide protection on a long run.

ESP_Sprite
Posts: 8926
Joined: Thu Nov 26, 2015 4:08 am

Re: GPIO output signal degraded over time

Postby ESP_Sprite » Sun Jan 16, 2022 11:29 am

Yeah, with a multimeter it's more-or-less meaningless to measure AC data signals and get a conclusion from that. Could just as well be that the data has changed, or something with parasitic capacitance. Also, given the fact that the ESP itself can use more power than a single GPIO can supply, fusing the ESP power line is useless here. You would normally limit the current of a data line by inserting a resistor in series with it, but in your specific case we'd have to see a schematic to see if that makes sense here.

tsctrl
Posts: 30
Joined: Tue Dec 08, 2020 1:37 pm

Re: GPIO output signal degraded over time

Postby tsctrl » Sun Jan 16, 2022 4:31 pm

hi, thanks for the reply,

i could not relates parasitic capacitance with why gpio was malfunctioning. and how esp could consume more power from its own gpio pin?

to measure, i will add resistor in series as suggested

thanks ESP_Sprite!

Who is online

Users browsing this forum: markkuk and 126 guests