ESP8266 Community Forum

- Fri Aug 05, 2016 9:52 am #52191 Hi,
with primary cells you would power the ESP via lowdrop/lowpower linear regulator. That makes max. 40µA (esp, flash, ldo) in standby. When the ESP powers up it powers the sensot, takes a reading, collects and/or transmits it and goes back to deep sleep. Either run until it doesn't power up any longer or just take a vcc reading every wake-up and send an alarm and set "sleep forewer" if under threshold.

with optimum LiIon (3.2V LiFePO4) you power the esp directly from the cell. Charging can be done via e.g. USB (smartphone charger), solar cell, etc and LM317 configured for 3.5V. LiFePO's are very robust (charging wise) and have almost constant voltage decharging. Battery monitoring as above. Standby is a little lower since ldo is not involved.

with standard LiIon (3.6V Mn,Co, etc) you go for one of the USB charge/protect boards (0.45$US, could also be charged via solar) and the rest like primary cell. The standby is a little higher, caused by the protection circuit.

It may be that the solution with the ATtiny has lower standby but the partcount is signifficantly higher and parts are more costly. Also from my point of view, it is less elegant ...

It is possible to further reduce the standby, but only if the "trigger" is external.

- Fri Aug 05, 2016 10:45 am #52193 Ok thanks first I am attempting with AA my next steps in the projects are Lilon but haven't worked which would be best I will check out the ones you mentioned above it seems like optimum LiIon (3.2V LiFePO4) would be better for me if so would these be correct:

http://www.ebay.es/itm/IFR-18500-1200mA ... Sw9eVXVvFF

And you say power directly from these, do you mean they are permenant part of the circuit and you build in the ability to charge via USB ?

My plans now are first to do with AA, one version using the Attiny/step up and esp 1/esp 12f, and and one version using the esp 12f and deep sleep and one version using esp 12f deep sleep and step up. After these and I have found out how much I can get out of them I will be doing the same device but with Lilon

One issue I have come up to is suddenly my existing version which is still breadboard is now failing to read the temperature sensor after wake up from light sleep, I constantly get NAN although the first read when the device is initially powered up is fine, this is with 12v ac adapter and 5v ac adapter and also the power from my programmer. I have changed the delay to see if it needed more time to power up but still not working which is annoying so I am going to try the 12f now with just deep sleep. I will eventually have different version and see which is the best then from that stage this will be my standard method for developing future devices similar to this. I am also implementing OTA,I had no success with ESP 1 so hoping with the 12f this will work as expected.

One last question, what would you suggest when the ESP has to be awake for receiving MQTT messages to power sensors etc, thanks for the help again.

Amazing selection of ESP8266 and ESP32 modules can be found here.

- Fri Aug 05, 2016 4:34 pm #52232 Hi,
yes cells like the one you found on ebay (maybe you find cheaper supply in rc-model or e-cigaret shops, indication usually is 3.2V). For this type charging is dead simple. 3.33-3.45V gives you full charge if limited to 4C and done until current drops to 0 (or 1C with 3.6V) and almost no risk of overcharging (they even manage 4.2V w/o exploding). With common smartphone USB charger you probably won't get more than 1.5A, even less on cheap solar cell; so not too much concern.
Stated 2V lowest discharge is not recommended. With 3.0V (safe for esp) you allready have more than 93% discharge. The recommended cc (read: limited current)/cv charge is easily done with simple linear regulators. Either fixed 3.3V or variable (LM317/117) with just using internal current limit or external limiting circuitry (see datasheets) according to a.m. cases. Allthough this is not the recommended nor the most sophisticated charging (but easy and cheap to implement), LiFePO4 have much more (3-4x) cycles than standard LiIon but slightly lower capacity and are much safer to operate!

Being able to save power and receive data should be possible, but seems to be tricky. See: viewtopic.php?f=29&t=11132
But power reduction is far from deep sleep or power down.

- Fri Aug 05, 2016 5:12 pm #52235 Ok thanks will look into it, I have just set up the 12f with the temp sensor and receiving the same nan on wake, following this code (although I have additional code for ap for wifi details and the temp sensor is powered up by gpio4 and read by 5) it is giving nan after the first two readings, any ideas.

http://homecircuits.eu/blog/esp8266-tem ... ot-logger/

EDIT solved this, now trying out deep sleep

Amazing selection of ESP8266 and ESP32 modules can be found here.