ESP8266 Community Forum

Hello

I'm using 600mA single cell Lipo battery for my project and I want to keep things as light as possible.

I need to power ESP01/ESP12 from this single cell Lipo battery , as well as 16 Neopixel RGB Leds (already powered directly from LiPo without any issue)

I'm searching for the best way to power the ESP from a LiPo battery, I already "found" 3 methods and I wanted to which one would you use.

Important note : the LEDs, powered directly from LiPo, are consuming about 400mA (average load during usage, with peaks to 1A max)

First solution :
The main issue is that I want to keep as much autonomy as possible, that's why I firstly eliminated the step-up only voltage regulators (1v-3.3v to 3.3v) because they cannot be used when the Lipo battery is fully charged (about 3,7v-4,2v).
But then, I thought about using a diode to drop lipo voltage off (about 0.8v).
Then I can use the step-up for what's he's done ==> raise the voltage to 3.3v.

I know it's probably not the most efficient solution in term of "power saving" but it's fairly inexpensive because of the very low price of such regulators (about 1€)

Second solution :
Use a lipo voltage booster to obtain 5v from LiPo, then use à 5v to 3.3v step-down regulator to power the ESP.
It's quite inefficient when considering dual power converting but there could be a major advantage, powering Leds with 5v will give even more brightness to LEDs (and that should be a good thing according to the purpose for which I use them)
lipo booster and 5v to 3.3v are fairly inexpensive too.

Third solution :
Using a step-up/down regulator which is able to convert any source from 1v-5v, to 3.3v.

I found one, lightweight : https://www.pololu.com/product/2122/resources
But it's a little expensive when considering that I have to buy a module for each ESP (20-25 modules)

I found another one : http://fr.aliexpress.com/item/mini-2-in ... eb201560_1
But I don't know if the output current would be sufficient (100/150mA) considering I will mainly receive data on ESP (mainly Rx and not Tx) and that I can force 802.11g to benefit from a lower power consumption compared to 802.11b.

The last solution seems to be the most efficient, but the most expensive one too (if pololu regulators are needed).

Considering the overall power consumption of the LEDs, do you think the first and second solutions are still "correct" ?

Last thing ==> may I need to consider that from 3.3v to 3v there's not much capacity left ? (if the case, maybe a very low drop voltage regulator should be sufficient ?)

Any other suggestion is welcome !

Thank you

well, a linear regulator is going to dissipate some of your battery power as heat (because that is how they work!). A switch-mode dc-dc converter is going to be more efficient (probably).

I don't see why the LEDs would need to be on a regulated supply? Why not just worry about the esp by itself?

In a high drain scenario, it is likely that a protected lipo pack will shut off at around 3.7v, so a buck-only dc-dc converter could work, but you would want to check the datasheet to be sure.

I could swear I'd seen cheaper versions of the pololu board on aliexpress or ebay.

Timpanogos Slim wrote:I could swear I'd seen cheaper versions of the pololu board on aliexpress or ebay.

Yes, I've seen a lot (and purchased some) at $0.99 ...

martinayotte wrote:

Timpanogos Slim wrote:I could swear I'd seen cheaper versions of the pololu board on aliexpress or ebay.

Yes, I've seen a lot (and purchased some) at $0.99 ...

Not seeing them now. I must be searching wrong, or there simply aren't any for sale at the moment.

The "tiny" buck/boost regulator he linked to is only rated for 150ma surge / 100ma continuous. May or may not be iffy for a mostly receive-only esp board.

It'd be worth experimenting with at least one, though.

You'd want a storage capacitor for safety, but, if that regulator lacks overcurrent protection, you'd have to do some math to figure out how much capacitor you can get away with. Inrush on some of the low-ESR caps can be brutal.