ESP8266 Community Forum

- Sat Mar 19, 2016 5:31 am #43537

Well that's one way to quickly remove the charge out of any small capacitance on the RST inout

The problem is that the capacitor will be "trying" to push the RST up to 6.6V after pin 16 stays low long enough to charge it - or pull RST down to -3.3V after something else (e.g. reset button) releases RST from 0V. This leaves it entirely up to the input clamp diodes to prevent damage to the pins. This is how clocked voltage doublers, triplers etc. work.

of course You are right, but values 6.6V and -3.3 these are extreme/max theoretical values,
remember that this is not ideal model of charge pump/voltage doubler etc,
RST on wemos have 10k pullup to 3.3 and 100n to GND.
I think that diodes so well protect inputs from peaks as generates double voltage.

MY GITHUB: https://github.com/reaper7/

SDM Energy Meter: https://github.com/reaper7/SDM_Energy_Meter
MLEDScroll (TM1640 led matrix): https://github.com/reaper7/MLEDScroll

- Sat Mar 19, 2016 10:39 am #43561

....RST on wemos have 10k pullup to 3.3 and 100n to GND.
I think that diodes so well protect inputs from peaks as generates double voltage.

Hi reaper7, thanks for confirming those values, I'd been too lazy to look them up and just went by the CR time-constant I was seeing on a scope. 100nF is quite a bit of charge for a cmos output to drive into. I wonder what they were thinking when they used that value on RST given that there's a frequent requirement to drive it low with GPIO16

I still think it's safest to link the two pins with around 100R. This triggers RST within well under 100uS. Here's a scope shot taken with 180R. Notice that the pulse is still at 0.5V after 300uS but the threshold is probably crossed as soon as the pulse begins.

RST.jpg

You do not have the required permissions to view the files attached to this post.

- Sat Mar 19, 2016 1:08 pm #43570 @AdrianM - I like real tests good work!
I have only old crt oscilloscope without "hold"

I'm talking only about my experiences with wemos mini + 10u between GPIO16 and RST, this device works for me without any trouble and hangs more that one week! This is a proposal for people who have problems with different values for resistors and esp hangs.

MY GITHUB: https://github.com/reaper7/

SDM Energy Meter: https://github.com/reaper7/SDM_Energy_Meter
MLEDScroll (TM1640 led matrix): https://github.com/reaper7/MLEDScroll

- Sun Mar 20, 2016 3:58 pm #43650 Hello @reaper7 and @AdrianM , thanks for your suggestions.
My Huzzah breakout worked great for 36 hours on battery until discharge, then restarted when connected to power. No resistors, just wire between GPIO16 and RST.
Feather hangs, always, with wire and tests with various resistors (160, 330,660).

I'm really a beginner in electronics, I'm a programmer, so I understood nothing of your discussion about the capacitor between RST and GPIO16.
I've tried everything, I'll try the capacitor too... 4,7 and 10uF.

Maybe you can, from the schematics, tell the difference between the Feather and the Breakout, and understand why they behave differently. May it be a programming/library issue? After all, in the Arduino IDE you have to select a different board for compile: "Generic 8266" for the Feather and "Huzzah" for the Huzzah.

Funny thing, my project was supposed to be deployed on Huzzah, with Feather for development, because of the connectors. But then I wasted 2 weeks trying to get stable deepSleep on Feather, and Huzzah worked fine the first time.