Need advice on solar powering with LiFePo4 battery
I need to power some outdoor sensors (built with Pro Mini + RFM95) with solar panels. I plan to use a single LiFePo4 battery and a TP5000 type charger and a small 6V solar panel.
I think I should not need any further regulation and can just connect the battery to VCC of the Pro Mini and the RM95.
Or should I be worried about the voltage maybe becoming too high in bright sunshine? Or what would happen if I took out the battery while the charger is still connected?
Does someone have any experience with this kind of setup?
It depends on how much safety redundancy you want. The cheapest solution might be to find a battery which already has overvoltage and undervoltage protection already built into it.
skywatch last edited by
SOlar power is a bit of a minefield in many ways.
How much sun do you get? How often is the node reporting? These are just 2 questions to get you thinking.
First I suggest you run the node as you intend to and see what the average daily power usage is and work backwards from there.
Note that batteries don't like hot conditions and solar panels never produce their stated wattage in real life situations due to haze, polution etc. They are tested in a clean air test room with lights very close to the panels. So if you need 5W for your setup then choose a 10W panel just to be safe.
I suggest you research all aspects more fully before buying components and then only make one to see if it all is as you want. Then purchase more items and build away!
@willemx I have a MySensors weather station powered by a 5.5v solar panel, TP5000, and a AA LiFePO4 that has been running for 2 years.
I also have a couple of soil sensors with a similar set up. It's been very reliable.
If you take the battery out while the TP5000 + solar panel is still connected, I believe it will pulse 3.2 - 3.6v into your board. I would recommend using a HT7333 or XC6206 voltage regulator and a large capacitor, just in case.
You don't need to be super concerned with power efficiency. I find my small solar panel tops off the battery within the first hour of indirect sunlight.
Thank you all for your recommendations!
@NeverDie : the TP5000 also has built-in battery protection, so I think I could just use a standard LiFePo4
@skywatch : my nodes only send a few dozen messages per day and are in deep sleep 99.9% of the time, so I do not need much power. I don't expect battery temperature to be an issue in my setup (I live in the Netherlands, very moderate climate)
@ncollins : do you know if a HT7333 will still work if most of the time there is only 3.3V at the input? I find the datasheets not very clear at this point. I should do some experimenting... And I wil apply your suggested large capacitor.
Did you use a regulator yourself ?
@willemx I believe below 3.3v it will basically pass through the input voltage with a minor drop, less than 0.2v. I also just tested this at 1.8v in the output is 1.787v (no load)...so not bad.
I use the XC6206 more often.
This is a great breakdown of the TP5000 charger.
https://lygte-info.dk/review/Review Charger TP5000 4.2-3.6V module UK.html
my nodes only send a few dozen messages per day and are in deep sleep 99.9% of the time, so I do not need much power. I don't expect battery temperature to be an issue in my setup (I live in the Netherlands, very moderate climate)
Those requirements could be met quite simply with an LDO, diode, and a supercap. For a cheap (but not the best) solution, you could consider just an LDO for charge termination plus a diode on your Lifep04. It would not guard against overdischarge though. A supercap doesn't need to worry about overdischarge.
I just now looked at a lifep04 battery on amazon.com with built in high and low voltage cut-off, but the low cutoff was so low (2v) that I wouldn't recommend it. Nonetheless, the entire thing (battery with built-in protection) costs only around $1.80, pro rata.
I don't see that the TP5000 has an undervoltage cutoff.... Not sure if it does or not. Depending on what kind of longevity/cycle-life you need to guarantee, you may or may not care about that.
SuperNinja last edited by
here is an example of realization which works for me (I have an RFM69).
I transmit temperature, humidity and pressure atmosphere all 30s and sleep. (BMP280 SHT30)
the TP4056: charges and protects the LI-ION against overvoltage and undervoltage (approx 2.7v)
There is surely to criticize or optimize the scheme: power of the solar panel, the battery, the buck Boost etc.
But in the state, it can transmit for 4 days of rain without sunshine (after a full charge in the sun).
@SuperNinja May not work on the OP's LifeP04, so I guess it depends on whether the OP wants to stick with a LifeP04 or change battery chemistry.
skywatch last edited by
Another thing is that most lithium batteries require a constant current charge. Every time a cloud passes by your solar panel output will drop considerably.
I get around this by charging lead acid from solar and lithium from the lead acid. Not the most efficient way, but the safest and kindest to the batteries.
@SuperNinja : Nice 'schematic' !
I had not thought of a buck/boost converter, but this might also be a good idea in my setup (we just had a full week of almost non-stop rain and drizzle).
I think I will stick with the LiFePo4 / TP5000. I guess the switch mode TP5000 should be more efficient than the linear TP4056.
My next target will be designing a 3d printable weather proof case and than I will do some actual field testing. I am a bit worried about the stamina of LiFePo4's in cold weather conditions. Winter is coming...
SuperNinja last edited by
@NeverDie yes the TP4056 is only valid for li-ion. The 18650 you can find so everywhere in new or "second hand" (laptop battery, hand vacuum cleaner ...) At first it was a cheap weather test, it has been in the garden for 4 years lol
I just took it apart to integrate it into MYSENSORS with RFM69.
@willemx , thanks i wanted to share something that works , Indeed, lifepo4 is more attractive with its 3.2 -3.4v and the TP5000 with a negative point: a resumption of charge at 3.5v which seems a little high according to the testers.
Remembering now, the undervoltage and overvoltage protection could both be put under the MCU's control, in which case you could get by with just a diode, or equivalent. It wouldn't be as failsafe in the sense that an unforeseen bug in your code might defeat it, but it could be done. In the end, I'm not sure that the chance of encountering a failure in software is necessarily any higher than that of encountering a failure due to defective hardware.
Alternatively, you could put two cells in series, and in principle that would eliminate any chance of overcharging. With that configuration, if you used a 5v solar panel instead of a 6v, then you'd also eliminate any chance of burning out the atmega328p and wouldn't need extra hardware to ensure against that.
For that matter, you could stick with one cell and just use a 3.5v solar panel. With just that one stroke you wouldn't need anything extra to guard against overcharging the battery or over-volting the atmega328p.
That's what makes embedding solar such a fun problem--depending on the use case, there are such a large number of different ways to solve it. So many that the real game is in comparing the trade-offs.