ESP8266 Community Forum

- Sun Sep 20, 2015 12:09 pm #29507 Hello, I have 2 ESP8266-03 modules, I built a small garage door control setup on a bread board and had it fully operating being powered by an ATX PSU. 5V is supplied to the circuit and a DC-DC converter (I had it laying aroud and only have 7805 and LM317 regulators) drops the 5V to 3.3V for the ESP, the rest of the circuit is as follows: https://onedrive.live.com/redir?resid=7064904358374137!3156&authkey=!AMU5c2a2j21b_5w&v=3&ithint=photo%2cpng

Attempt 1: I had a USB wall charger from an old digital camera (5V 850mA), I removed the USB connector and soldered the output to the input of the DC-DC converter. I plugged the charger in and there was a loud buzzing coming from the ESP, I quickly unplugged it and noticed the chip on the ESP was uncomfortably warm, I cut the wires off the and measured the output at 5.29V DC with about 20mV AC ripple (DMM set to AC) and 100 Hz (this seems low but I'm not a SMPS expert) I hooked the circuit back up to my ATX benchtop supply and the circuit pulls about 130mA on the 5V, Doesn't operate and gets hot. Assuming the chip was fried I tossed it and soldered in my other ESP8266-03 and threw out the Camera charger thinking it was damaged/unusable for a project supply.

Attempt 2: I flashed my 2nd ESP and had it operating as desired attached to my bench supply. I got another USB charger (this one for a phone 5V 450mA) and connected it to the DC-DC converter. The circuit powers up and the ESP spews apparently garbage data to the serial port. I reflashed it and got

Code: Select allFatal exception (0):
epc1=0x40212764, epc2=0x00000000, epc3=0x00000000, excvaddr=0x00000000, depc=0x00000000

endlessly. I tried adding a cap to the output of the DC-DC conveter and swapping the DC-DC for a LM317 set at 3.3V and I continue to have the issue of garbage data on power up and the fatal exception after flashing, I switched back to the benchtop supply and the issue continues. I tried Reflashing with the AT firmware and get the ame results except a different code

Code: Select allFatal exception (0):
epc1=0x40211558, epc2=0x00000000, epc3=0x00000000, excvaddr=0x00000000, depc=0x00000000

Is there any possibility of saving this module and what happened to both why did a SMPS fry the 1st one through a DC-DC converter and what happened to the 2nd one to cause it to throw these exceptions? Has anybody successfully powered one from a USB charger?
Thankyou

Last edited by Shadow351 on Mon Dec 07, 2015 11:26 pm, edited 1 time in total.

- Mon Sep 21, 2015 8:57 am #29580 Inexpensive chargers are rarely well-regulated; also many are designed to deliver the rated voltage only when loaded by the intended battery load. So they are not to be trusted as general PSU until testing them out thoroughly, or providing known-good regulation. Note also that many common regulators won't regulate nicely from 5v to 3.3v; the 1.7v differential isn't enough. I bought some surface-mount LM1117 regulators {10 for like US$1 on ebay); these do regulate 5 to 3.3 correctly.

USB power supplies should theoretically be a dependable source of 5v, but again they're cheap and should be tested before use. Currently in my dev setup I use an old transformer-type surplus 10v 1 A wallwart, a LM2596 switching reg to deliver 5v, and LM1117 right by the ESP for 3.3v. No power problems in 9 months of use.

Your attempt #2 problems sound like 100% firmware. All I can suggest is to clear the flash (there's a 512k blank bin file available somewhere) and try reloading. Maybe try something basic like the blink example. I recommend that people buy their ESP boards 5 at a time Stuff happens...

- Sun Oct 11, 2015 9:25 pm #31132 OK thanks for the input. I received my shipment of new modules (I ordered 10 this time). I programmed it while powered by my atx power supply, then used a 6 volt x-former power brick and an lm317 set at 3.28 v (as close i could get with the resistors i had) and it is operating properly. I will let it sit powered up on the bench for a few days to make sure it doesn't spaz out before attaching it to my garage door.

as for the second module i could never get it to reflash (flashing it blank didn't help) I did some reading and found some other posts about the same issues i was having. I guess the flash memory these modules use has a fairly limited write cycle and being that module was my tinkering module before it may have just been reflashed so many times it just has a worn out flash memory chip.
Thanks
-Brad

- Fri Apr 08, 2016 11:38 am #45114 Dear Shadow351

looks like i am facing the similar issue here and got no response from anyone regarding the problem.I was successful to upload data to thingspeak by powering the ESP8266 and atmega8 uC from my 5v/1A mobile charger for 1-2 days.But soon after 2days or so,the ESP8266 either stops responding or misbehaves.Also flashing with new firmware doesn't help.I already fried 2 modules and third one is running and powered from a linear PSU for long test for last 7/8 hours.Were you confirmed that the problem was regarding the smps type power supply??Please reply as no one is answering/didn't face this type of problem....