ESP8266 Community Forum

- Sat May 09, 2015 11:37 pm #17053 I have a simple-ish program running on the module.
However, it somehow resets itself at times (fairly repeatably.)
I'm either stuck in an infinite loop, or I'm smashing memory somewhere, but I can't find where it is by simple code inspection.
I'm using the ESP8266-01 module, so I'm really low on GPIOs to do something like drive a little LCD display.

What's the best way to debug this problem? Is there some "reset cause" register I can read that tells me why it booted? (Then I can spit it out on the serial port.)

I'm using the Arduino-based download toolchain/IDE.

- Sun May 10, 2015 4:57 am #17070 There has been reports of the ESP-01 (individual modules? Batches?) being very sensitive to electromagnetic noise causing it to reset. Forum user Sweetpants has has a theory that the reason could be that the RES signal is routed along the edge of the module, and proposes to cut the track. I succeeded stabilizing a module that was very sensitive (tre previous that I fried for the same board did not have the same problem) by cutting the RES track (but not adding a chip capa as proposed by Sweetpants).
http://www.esp8266.com/viewtopic.php?f=13&t=1924

Other users have reported having used "hundreds of ESP-01 modules" without seeing the problem. This could indicate that this is a batch issue, and / or that those users have a particularluy well-designed power supply / decoupling circuitry. I have not seen reports of random resets on other modules than the ESP-01, which could indicate (my speculation again) that the designers have seen the issue and designed later modules better.
A carefully installed decoupling capa is recommended, and *could* reduce the problem.
http://www.esp8266.com/viewtopic.php?f=13&t=2839

- Sun May 10, 2015 11:30 am #17097 That's interesting. I was considering power supply as well. I'm using a Xbee Explorer, which uses a small SMT 3.3V regulator that is probably quite marginal for the power draw needs of the ESP when transmitting.

My current plan is to wire in a high-power through-hole power supply (linear regulator, can caps) and see how it goes!

- Tue May 12, 2015 2:47 am #17215

That's interesting. I was considering power supply as well. I'm using a Xbee Explorer, which uses a small SMT 3.3V regulator that is probably quite marginal for the power draw needs of the ESP when transmitting.

My current plan is to wire in a high-power through-hole power supply (linear regulator, can caps) and see how it goes!

I believe having seen 200 mA specced during transmission bursts. Assume a bit more.
If your current power supply is marginal for the peaks consider adding a huge electrolytic capa, which will serve as charge buffer. But a more solid power supply will never be wrong.