Battery requirements for ESP8266


I’m going around in circles with power requirements for my current project, and I’m hoping someone can point me in the right direction.

I’ve set up a basic DHT10 shield on and ESP8266. The purpose of the project is to monitor temperature and humidity inside an entomology specimen preparation environment, so it communicates via wireless to a ThingsBoard server I have installed on a local PC. But I have trouble with power. I need between 3.3V & 6V battery that’s small enough and light enough to be anchored within the specimen box without risking it moving in transit and crushing the delicate specimens. I quickly discovered that the setup could drain 4 brand new AAAs overnight before I set up some power management and switched the system to deep sleep for 15 minutes between readings. I thought 2x CR2032s would do the job, but the start up requirements of the ESP8266 are about 250mA, so it just wouldn’t turn on.

I’m coming from the coding side and am newish to microcontrollers, so I’m only learning about the electronics and might be missing something obvious. Hoping to springboard from this project into a more complex environmental monitor, so I see that I’m going to be needing similar power setups in future.

Can anyone recommend something that would do the trick?




Hi Heather.
I think you will find this is a physical thing.
Each different battery chemistry has its own power / size ratio and there does not seem to be anything you can do about that. The more power you need the larger the battery. Having said that over the last years great improvements have been made which has involved battery chemistry and I think to a lesser extent manufacturing techniques.
I may stand corrected but at present I think one of the Lithium types holds the flag for the best power / size or density ratio. As far as alkaline goes I believe the current AAA size has much the same power capability as it did years ago. There is nothing you can do to improve that. Just the nature of the beast.
You need to determine firstly exactly what voltage is required. Then the Max current and work up (NOT DOWN) from there. You have quoted 4 X AAA batteries (6V) and 250mA so start from there.
Cheers Bob


This is quite a large range.
If you can provide details of the DHT10 & ESP8266, ie pics would help.

If the devices you have are similar to what is linked below then you need something that generates 5V.
4xAAA’s would do that 2xCR2032s would not.

Knowing the voltage requirements of the devices would help greatly in providing advice.

Something like this would provide enough power and is 5V output, but it is quite large.
L=95.0mm W=26.5mm H=22.0

Generally power banks will not output at low current levels. They are designed to recharge a phone not act as a power supply.

Another solution would be to use a LiPO (link below), but it would need a step up circuit to provide 5V.
They provide a good output for their size.

This would provide the 3.7V to 5V boost and allows recharging when the LiPo is depleted.



Hey Heather,

As Robert has said, different chemistries have different energy densities and properties, diagrams like this come up a lot:

It’s also worth thinking about self-discharge, as this differs based on chemistry, and can actually be a decent chunk of your power use if you manage your micro right:

If you plan on doing a ton of low-power electronics, an Otii Arc is a combination power supply/current meter that works well for determining how long your low-power projects will last.

My first thought is to try storing a day’s values to flash, then only turn on your Wi-Fi/BLE once a day to transmit values back to base, as WiFi is a good chunk of that power draw.

What is your required measurement frequency/duty cycle?



The SparkFun ESP8266 is functionally similar to what I have - I grabbed mine at Jaycar when they didn’t have a seperate wireless module, but it seems to be a pretty standard design (WEMOS D1 I think?).

I am confused as to why 5V is the minimum voltage. Is it because running the DHT10 on top of the ESP8266 requires additional power? That sounds like a total n00b question…which is fitting, really.

I’m not sure how the step up circuit works, but that means I’m learning something!

Thanks for the tips - I’ve got a starting point!


Hi @James46717 ,

You’ve actually found a bit of a hole in our info on the site! That DHT11 is 3.3V compatible, so I’ll get that updated now.

As for @Heather161189

The ESP8266 runs at 3.3V logic, running all your sensors and things at that voltage isn’t a bad idea.

You’ll just need to find out whether your board is happy being powered around 3V, as some would want to be regulating 5V down, others have a ton of range. We’ll need to know for sure what your dev board is.



Many devices today will be specified as needing between 3 and 5V.

I was unsure of the requirements of the DHT10 & ESP8266 you have. I have used a DHT10 before and it needed 5V to operate. The newer DHT11, as @James mentioned, will run from 3 to 5V.

The SparkFun ESP8266 Thing SKU: WRL-13231 is designed to run from a USB port (5V) or from a LiPO battery (3.7V). The board has a regulator and LiPO charger built in. It is quite a good little board.

The only reference I can find to WEMOS D1 is an Arduino UNO type board. It is specified as 5V input but has an onboard 3.3V regulator.
A pic of your board would help identify it.

The different voltages are what the manufacturer has decided to use.
Specifications for a product will list what voltage range it needs to operate correctly. You should always stay within this range.

No, the DHT10 would use very little power. 2.5mA from the Core Electronics page.

The Jaycar web site lists only the following module for a ESP8266 search. Is this the one you have ??

Anyway, happy to help, just need to know exactly the board you have.


Thanks for all this Jim, I do appreciate it.

Yes that’s the board I have. And checking up, it’s the DHT11 and not the DHT10 sensor I’ve been using (though I’m quite sure I’ve got a few 10s as well)

Here’s the set up, which is probably not great for ID.

Only difference is I’ve now got a jumper from RST to D0 to allow the ESP8266 to wake itself up from deep sleep mode.


Jaycar, as usual, don’t have much on the specifications of the ESP8266.
Something that has frustrated me with respect to the ESP products, lack of proper datasheets.

The Jaycar ESP8266 is a ESP-12N which has a 5V input to allow connection to a USB port. While the chips inside the module run at 3.3V. The ESP8266 processor definitely runs at 3.3V.
The power pins are labelled on the board. You should be able to measure 3.3V on this pin when plugged into a USB port with a multimeter.

BUT … It does NOT mean you can connect a power source to the 3.3V pin and it will work.
It would need to be regulated to 3.3V.
Datasheets for the ESP8266 state: VCC Power Pin (+3.3 V; can handle up to 3.6 V)

It is not clear whether the ESP-12N would work with just a 3.3V connected to the 3.3V pin.
There are a number of chips inside the module and some may need 5V.
Providing 3.3V on an external pin is for logic and data communication at that voltage level.

It may work, it may not.
At $24.95 I am not prepared to test it when I may end up with a dead device.

To be safe and ensure correct device operation you will need a 5V source connected to the USB port or to the 5V & GND pins. Probably this is what you did when using 4xAAA cells.

Wall of text here, but it is the design process you go through to be sure of what you need.
Suggestion in next post.



Previous post established the need to provide a 5V source to the ESP module to ensure correct operation.

Previously I linked the device below.
It provides 5V, allows use of rechargeable LiPO’s, and had a built in recharge circuit.
Depending on how long you want the device to run the LiPO could be from 1000mAh up. (400mAh may not provide enough current to start up or power WiFi)

The problem now would be to house the Boost device, ESP8266 & battery.
Jaycar provide many small cheap Jiffy boxes
Or you could 3D print your own.



It’s would be unusual that an ESP8266 module needs 5v in order to operate. The 8266 modules usually only include no more than an optoisolator for some of the external lines, if that. The regulator is only included if there is a USB port, and is only there to allow powering from 5v. There are several variations of the D1, but they all operate just fine with 3.3v input. But it would be nice to confirm the exact module.


Agree, data sheet specifies 3.3V to 3.6V.
Most likely the device would work with a regulated 3.3V power source connected.
Have not found a circuit schematic for the ESP-12N.

You cannot connect a 3.3V battery to the pin, the voltage varies too much and if greater than 3.6V, the device would be damaged.

I wanted to find something Heather could use for portable operation that would not requite too much electronics work. The Adafruit PowerBoost Charger provides the ideal solution. If this was my project that is what I would use. Then mount it securely in the Jaycar Jiffy box with flanges and bolt that somewhere inside the specimen box.

Anyway, if anyone has a better idea happy to hear it.


I had a thought - you’ve made suggestions about using lower voltage batteries and stepping them up (which my admittedly limited knowledge suggests would means the batteries run out faster). I wonder if there’s any value to using a larger battery (i.e. 9V) and stepping the voltage down, and, based on aforementioned conclusion, getting a longer runtime out of that battery?

I’m new to cases…I thought a case for an environmental sensor would need strategically placed holes, so I put it in my “cross that bridge when I get to it” basket, but is it really just as simple as a jiffy box? I mean, I can work a Dremel, no probs (or a die grinder if things get crazy), but I expected cases to be hard.


Yes, the same amount of power would be used, but with a larger capacity the battery would last longer. LiPOs are light and small for their power output and are readily available at 3.7V, so step up needed. They can be made to larger voltages for special circumstances. 3.7V is ideal for mobile devices; phones, tablets, fit bits, etc

Safety is an issue with all Lithium type batteries. Used them as the manufacturer recommends and they will be fine. Never run them in series and always use the manufactures charger. Never leave them on charge unattended, and never in direct sunlight. Too many fires have been started by incorrect use.

No its not. It would need a bit of Dremel work and the sensor should be mounted outside case. Which is a problem with the design of the one you are using. The DHT11 and ESP8266 could be mounted with a hole cut out the size of the sensor. My own experience with sensors is they need to well away from the processor if it gives off any heat.

The Jaycar plastic Jiffy boxes are easy to drill and cut, if you are good with a Dremel tool.

Recently I purchased the sensor below to test with an environment system I am building. It looks like it would be easy to mount outside with some protection and connects via a small cable.



The ESP8266 on board regulator can work up to 6.5V according to the data sheet and will still regulate down to 4.3V. (as far as I can tell from the schematics I can find)

So, 4xAA NiMH cells (2000mAH) in series would give 5.6V to 4.4V over the discharge cycle (approx).
You could use AAAs (900mAH) for smaller size. Most likely they would last long enough. If you don’t have a NiMH charger, buy a good quality one. In my experience NiMH can fast charge in a couple of hours.

Connecting these to the 5V or VBUS pin and GND should work ok. No need for step up or down.
The sensor I linked in my previous post, needs 5V but has a on board regulator to produce 3.3V, so it would be ok with the battey input too.

All the best


Hi Heather ad James.
If you decide on MiMH AA’s do yourself a favour and go for the “eneloop” type. Good capacity and they take years to self discharge. I once had 4 fitted to a Fujifilm camera and used to get something like 300 pics (mostly without flash) before recharge.
Cheers Bob


Okay, I’m back at this, and confusing myself over charging. In order to be able to charge the battery, I need to keep the JST PH2 connector attached, so gotta plug it into something to run the board. If I’m gonna plug it into something, it might as well be a charger, so I don’t have to remove the battery from the unit to charge it.If the battery is 3.7, can I run something like Adafruit Micro-Lipo Charger for LiPoly Batt with USB Type C Jack | ADA4410 | Core Electronics Australia ?

But then I’m not confident about connecting the charger to the board. The above has the USB breakout, so I thought I could connect the data parts to update the board from the same connection, then I remembered that the board kinda has that handy USB connection onboard already, and then, why can’t I hook that up for charging? So now this is looking excellent Adafruit Feather HUZZAH with ESP8266 WiFi | Espressif ESP32 & ESP8266 ADA2821 | Core Electronics Australia, but switching boards may be overkill for a problem that’s probably not as big as I’m making it out to be.

So how would I connect either the above LiPo charger or the PowerBoost 500 Charger previously linked by @James46717 to my project (because I’ve looked at the schematics and learned that I need to learn to read schematics)? If I can comfortably run 3.7V, would the simple LiPo charger work with the aforementioned 1100mAh battery (Polymer Lithium Ion Battery (LiPo) 3.7V 1100mAh | Core Electronics Australia)?

1 Like

This is a simple charger, it takes 5V from any USB port and uses that to charge the LiPo.
To use the charger, remove the LiPo from the project and connect it to the charger, when charged remove the LiPo from the charger and connect it to your project.
So, no this board is not what you want.

D1 Mini
The allows the battery to remain connected all the time and provides the 5V needed for the D1 Mini through the USB cable that connects to a phone in the pic. (white cable)
The other USB port (black cable) is for connecting any USB source to charge the LiPo.

So, the LiPo is always connected.
The white cable connects to the D1 Mini when in use to provide power.
The black cable is only connected to charge the LiPo when it is flat.

Easiest solution if you stay with this board.

Adafruit HUZZAH

This is an excellent board and for what it does pretty cheap. I have used one of these and highly recommend it. It would work for your project if you decided to switch boards.


1 Like

ROTFL dumbest question ever! Sorry, never even occurred to me that you’d use the USB to power the board - I was looking for something to solder.

I’ll order the parts and wait in my hole of shame till they arrive :stuck_out_tongue_winking_eye:

Thanks for all your help!

1 Like