Any users of LiFePO4 batteries ?

[Nca78]

Hello, has anyone here used LiFePO4 batteries to power MySensors nodes ?
The look really interesting for low power nodes in 10440 format :

• low capacity compared to li-ion but somewhere between 200 and 280mAh which is ample for over a year
• low self discharge which allows to survive the full year
• maximum voltage below the NRF24 max so no need for a regulator
• minimum voltage around 2V but with a light load if I believe the discharge curves I see over 90% of the capacity is available with a voltage over 2.7V meaning most of the capacity can be used without changing the bootloader and BOD.

As it's a bit rare and expensive (at least for sellers offering delivery in my country ) and it needs a special charger I would like to have some feedback from people who are using some.

Thank you for reading

[gohan]

LiFePO4 batteries are known for the higher safety in respect to LiPo but at the expense of capacity. I read that are also used e-cigarettes because are safer and people don't like holding something that may explode in front of their faces.
Imho CR123 (there are also rechargeable ones) could also be an alternative for small footprint.

[gohan]

Take a look at this video. https://youtu.be/heD1zw3bMhw
Interesting is the connector fake battery to convert 2 AA batteries to a single LiFePo4 cell. @sundberg84 with 2 of these cells in parallel you could more than double the battery life of AA powered nodes and you don't need a regulator

[Nca78]

@gohan said in Any users of LiFePO4 batteries ?:

Interesting is the connector fake battery to convert 2 AA batteries to a single LiFePo4 cell. @sundberg84 with 2 of these cells in parallel you could more than double the battery life of AA powered nodes and you don't need a regulator

I don't think it's great as it's using the same space than 2AA/AAA with 5 times less capacity.
Regulator is not needed with 2 AA/AAA, and you need a booster only if you keep the BOD at 2.7V else you can draw the batteries down to 0.95V each, atmega and nrf24 can run below 2V opposite to the ESP8266.

[gohan]

According to the video it has 82% of the capacity of 3 AA alkaline

[Nca78]

@gohan you cannot compare a test made to power an ESP8266 which accepts only 3 to 3.6V and an atmega328 (1.8 to 5.5V) and NRF24 (1.9 to 3.6V).
Capacity of his AA LiFePO4 is 700mAh. Capacity of AA alkaline battery is somewhere around 2500mAh. With atmega & nrf24 you can suck all the energy from alkalyne, as well as from the LiFePO4. So in this case you have less than 1/3 of the capacity. It's acceptable if you use only one battery and save half the battery volume, not very interesting if you keep the volume of 2 batteries

And with AAA (because I want size to be as small as possible) capacity is max 280mAh for LiFePO4 when normal AAA has around 1000mAh so it's only 1/4th of capacity or half the capacity per volume. But when I see battery life of my first sensor on CR2032 (over 1 year at my entrance door and most of the capacity still left) AAA is overkill for low power sensors and one LiFePO4 in AAA size should be enough is reality matches the promises.

Charging the sensors just once a year would be great !

[gohan]

keep in mind that they have double voltage of a normal battery, so to make a fair comparison at 1.5v you would have to double the capacity (700 x 2 = 1400). Rechargeable NiMh like eneloops that can maintain charge over a long period of time have a capacity between 1900 and 2000mAh (not far off the LiFePo4) while higher capacity AA batteries (2500-2700) tend to self discharge (I had a bunch of Duracell rated 2500 that were self discharging within 10-15 days, after those I bought only eneloops ). The nice thing about the atmega328 and nrf24 is that they can work with low voltages but I still like a lot the discharge curve of the LiFePo4

[Nca78]

@gohan no I compare 2 AA/AAA providing 3V (enough to run my nodes) with 1 LiFePO4 so from there I can compare capacity.
But yes what's interesting is the low self discharge of LiFePO4 that allows long battery life, it seems to be the only rechargeable solution able to provide 1-2 years of battery life. Li-ion cells I have discharge in months.

Seems no one uses them for real so I'll invest and see

[gohan]

For long discharge time I think eneloops are good. For your needs keep in mind that you can use a single LiFePo4 to power a node while you would normally need 2 AA sized batteries.

[Nca78]

Just received LiFePO4 batteries in AA size today, with claimed capacity of 700mAh (Soshine brand).
Bought a "Miboxer C4" charger on AliExpress to charge them and it's really great to charge and to rest capacity, so that's what I'll test first and then next week I will make one or two nodes, but it will take some time to have an idea about the self discharge
With the charger I tested some other batteries and no surprise with the cheap AliExpress batteries: 1000mAh CR123 li-ion is 300mAh and 3800mAh 18650 is only...400mAh so just 1/10th of the claim.

I also tested 18650 cells from a 2011 laptop. I charged them over 6 months ago and left them in the drawer and surprise they were still nearly full, only a few minutes needed to charge. And they can still provide 2400mAh of the claimed 2600. So probably a good li-ion battery could power a node for over a year too, with a ldo like a 3V xc6206.

[gohan]

Chinese capacity and capacitance of supercaps are not to be trusted

[Nca78]

@gohan said in Any users of LiFePO4 batteries ?:

Chinese capacity and capacitance of supercaps are not to be trusted

Well I just tested the Soshine LiFePO4 and my charger measured over 600mAh when rated at 700 so I guess they are honest as it discharged at 400mAh which is quite high, and I'm not sure it went as low in voltage as it could.
So some Chinese batteries can be trusted

[gohan]

The soshine charger is junk, there are some videos about that.

[Nca78]

@gohan said in Any users of LiFePO4 batteries ?:

The soshine charger is junk, there are some videos about that.

I'm using a MiBoxer C4 which is I think not related in any way with Soshine. It gave me valid results on branded li-ion and nimh batteries and measured very poor capacity on Chinese li-ion batteries so I don't think it's cheating when measuring those batteries.
For the Soshine charger, no idea about it but if I remember the prices seen on AliExpress it was less than 10$ added cost to have them with the batteries so for this price it's not a surprise if it's not a good product.

I'm interested in the capacity of batteries and for that Soshine keeps it's promises.

[alexsh1]

@gohan said in Any users of LiFePO4 batteries ?:

The soshine charger is junk, there are some videos about that.

It may well be, but their LiPOs are good, really good. I have several 9V batteries powering Mega 2650 and they last for a few months (on and off). They are rated 9V 600mAh.

[NeverDie]

I read somewhere (can't find it now) tht some LiFePO4 batteries were found to be spiked with Li-MnO2 or similar to give them better mah ratings. Plainly, that somewhat defeats the purpose of using them in the first place. Not saying that is the case here. However, if you have LiFeP04 batteries whose mah seems too good to be true.... In the end, it's such a non-obvious thing to verify, that I wouldn't be surprised if it's happening a lot, as we already know there are a lot of fake batteries out there of one kind or another.

[Nca78]

@NeverDie said in Any users of LiFePO4 batteries ?:

I read somewhere (can't find it now) tht some LiFePO4 batteries were found to be spiked with Li-MnO2 or similar to give them better mah ratings. Obviously, that somewhat defeats the purpose of using them in the first place. Not saying that is the case here, as I have no idea. However, if you have LiFeP04 batteries whose mah seems too good to be true.... In the end, it's such a hard thing to verify, that I wouldn't be surprised if it's happening a lot.

I can't find the disadvantages of LiMnO2 batteries, any link ? Do they have a high self discharge ?

[NeverDie]

@Nca78
No links. Just the usual "burst into flames" safety concerns.

[Nca78]

@NeverDie said in Any users of LiFePO4 batteries ?:

@Nca78
No links. Just the usual "burst into flames" safety concerns.

I see, then I'm not too worried. Soshine seems to sell a lot of batteries and I didn't see any report of explosions or related things.
And when charged and discharged quickly (for capacity testing) they seem to stay much cooler than even my brand name 18650 cells.

The guy from Lygte has tested a bunch of batteries and results are good for all the soshine, and it means capacity measurement on my charger is not bad, I measured 600mAh for the LiFePO4 AA and he has 550.
Every other battery brand that is available on AliExpress and he tested has completely bullshit claims about capacity. What's worrying is they also lie about having protections, and some brands like TrustFire rewrap some old batteries from powerpacks and hide marks/rust/... with new tops and bottoms, this is mainly the case for 18650. Old rusty batteries without protection, and all different despite behind inside the same wraps
Page for small batteries :
http://lygte-info.dk/review/batteries2012/CommonSmallIndividualTest UK.html