Mi Flora battery life

I put in a fresh battery a month ago, running 3.2.2 since then and the battery is still reported at being 100%

Due to the change of two variables it’s hard to say whether the improvement was achieved by either:

  1. The new firmware 3.2.2 or
  2. Setting scanning to passive via: active: false
    active: false

In any case, I think now the battery should last longer…

I’m going to give this low-tech glamorous raincoat a go. It’s just on the balcony getting occasional side-rain.

Edit: I have already upgraded to this breathtaking spice jar. I just pushed the jar into the dirt. I figure the bottom of these things should be ok.

1 Like

I stille have ridiculous battery life. Wondering if the ones I got are shit. Which battery did you use?

Why would you cover it? Mine are outside all year, getting rained and snowed on and still work…

Just a normal CR2032 battery, bought at a local drugstore.

I have one in a window box on the balcony of a third floor apartment and it’s fine. I just have the side with the hole aiming downwards.

I’ve had two uncovered on the balcony running pretty well for about two years. In recent months, the battery of one that gets more rain exposure now keeps getting a flat quickly (a month or so). So I thought I’d try this as a hopeful and easy trouble shooting attempt. They’re both on the same firmware/configuration etc.

1 Like

I have stopped paying attention to the battery sensor and just decided to let them fail before I change the battery. So far so good. I guess they could have changed the algorithm for battery level percentage.

Hi @Gymnae How are your mi flora sensor batteries doing?

I’m on the verge of ordering some and trying to see if the 2mo battery life has been solved.

There’s a myriad of options on aliexpress and who knows what i’l actually get. Some say green is the CN version, some dont.

I’m also wondering about using them in larger planters. ~10-12" deep. There was a larger version but it seems to have fallen off the face of the earth, and never worked right from what I can find.


Franken-flora is still going strong.

You guys that are putting this device out in the elements without protection are eventually going to get electronic aids. The thing ain’t protected at all from moisture. Unless things have changed during the factory assembly,


How has your jar-approach hold out? Any moisture issues yet?

No issues yet. Seems to have helped and is going strong…

Does it create a mini-greenhouse and thus report a higher temperature (i.e. do you see condensation inside)?

Probably quite likely that it affects the temp I guess (although I only really monitor the moisture and light sensors myself). Can’t say I’ve noticed condensation, but I guess it would make sense that it would create a greenhouse effect.

Hello everyone, joining the family of the complainers.
I was me too cursed by some sensor getting shorter and shorter life, I have 14, all outside, and now some of them are dying in just around 5 days. So I decided to find a solution.
I share the point that it must be temperature or humidity related, or both, because they always die by night.
I tried to coat them with plastidip but nothing.
So I opened one and powered it by 3V (actually 3.3V at the moment, but seems to work).
What I have noticed is that the idle power consumption is spiky but averaging on 30uA, during polling by the bridge goes at 100uA and time to time does it by its own. This matches with what I read that the chinese market version is actually actively communicating from time to time.
This alone would not be enough to justify the short life of the sensors.

BUT here comes the interesting part I realized that if I power cycle the sensor, it stays in a kind of pairing mode consuming astonishingly high 2-3mA till it gets a bridge connection which in my case is every 30 minutes. This may be the reason for my sensors suicides.
Now I assume that during cold nights the voltage of the batteries drops low enough to power cycle the sensors and then the guys when warm again get powered and jumps into this pairing mode, then wait for their death in there…

Now I’m working on a way to power them externally with a battery each 4 or 6 of them and then cut the power by night. I will keep you updated.


I’m tinkering with powering them from 3.7V 18650 lipos. It works for tests. Do you or others have thoughts on the likelihood of death from lipo voltages as opposed to 3v button cell?

Hi, before doing my modifications I checked this teardown : https://wiki.hackerspace.pl/projects:xiaomi-flora and the main chip is rated 2.35 -3.3 V, so i would not do that…

1 Like

Different voltages seem to radically affect results anyway (values around 20% are on 3.7V, values around 5% are on 3V). I wonder how much this factor affects readings over the voltage range of the cells battery life?

Maybe I’ll look at cheap/efficient ways to step down 3.7V to 3V. If anyone has any pointers (@tom_l ?) please shout out, otherwise I’ll get Googling. Like I found this [TPS63031], but I don’t know what I’m doing yet…

Hi, to try reducing the battery consumption I bought a Xiaomi Gateway v3 and I connected it with HA using the following addon: https://github.com/AlexxIT/XiaomiGateway3
I’m also feeding Influxdb for data analysis
Let’s see how long the batteries work

Replaced battery after 9 months. Not too bad but not amazing…