How can I make a 3.3v Solar power supply that uses an 18650 battery?

I have a project which I would like to make self sustaining solar powered that runs 24hrs a day. the electronics themselves use between 60ma and 220ma of power. the 220ma is for about 1 second every 2 minutes when the device transmits it's sensor's data, the rest of the time it is 60ma to 80ma. I want to make sure that the circuit has overcharge and over discharge protection for the battery as well as voltage regulation to between 3v and 3.3v for the load. I have been searching and I cannot seem to find a premade solar controller board that will do what I need. If there is one <$20 USD please point me in that direction, but I would still like to learn how to build one.

sort by: active | newest | oldest
homerruma1 year ago

Waterproof Solar Power Bank+lm2596

Ada fruit do a solar charger for lithium which looks suitable to me.

funinalaska (author)  steveastrouk2 years ago

I just thought of a potentially very simple solution to this problem. If it will work. Would I be able to plug the output of a 5v solar panel directly into the USB input (charging port) of a small USB power bank? Is there any reason the solar panel would not operate the charger on the power bank?

all depends on the USB power bank. It may try to draw too much current. It may not be able to higher or lower voltages.

It's worth a try, but bear in mind 5v is the output only in full sun conditions.

What are you actually doing, that is pulling a constant 60mA, can you drop it down further ? What processor are you using ?

funinalaska (author)  steveastrouk2 years ago

using an ESP module.

when it is in standby it actually draws about 30ma but when it goes into transmit it is drawing about 127ma but can draw up to 250ma.

so I have had a module running on a cheap quality 18650 battery for about 36 hours now and the battery is only down to 3.6v I have a protection board on it that will shut down at 3.0v. so I think I will have no problem with the battery capacity or voltage drop issues just need a charge circuit that will protect the battery from overcharging from solar

What panel are you intending to use ?

funinalaska (author)  steveastrouk2 years ago

I am also open to other battery options the 18650 option seemed to be a good option but as I research I a beginning to wonder if other options would be better.

You're finding the problem with using lithium cells - they are a royal PITA to nursemaid. I'd look at a nice fat SLA, and make sure its capacity is used to less than 20%. You're only running, say 80mAH, which even a small SLA should handle very cheaply.

funinalaska (author)  steveastrouk2 years ago

Unfortunately I am not as familiar with SLA batteries as I am with Li. What are the voltages (max, minimum) for say a 4v SLA battery. I found a 4AH very cheap locally. Would I still want to use a 5v solar panel? Would I still get away with the diode voltage reduction or would I need something ore like a switch mode regulator?

SLAs usually come in 2, 6, 12 or 24 V. For an application like this, a Cyclon SLA, 2V cell would be ideal. Cyclon's are almost impossible to damage, and are very small. Then you'd need to boost it up to 3.3V, which is pretty straight forward as you know, and very efficient.

funinalaska (author)  steveastrouk2 years ago

I have not decided on a panel yet but I would prefer something in a 8 inch square or smaller.

Have you checked your electronics won't work on 3V ?

funinalaska (author)  steveastrouk2 years ago

Yes My electronics will work reliably fro 3.3v to 2.8v

I'd put a diode in series with the battery to your load, and drop 0.7V in that then to "regulate" the power to the load.

I calculate your mean power consumption as about 0.2 W, so you can probably run this off a very small panel, even allowing for night operations.

The next issue is a charging circuit, and your nominal panel voltage will decide what's feasible.

funinalaska (author)  steveastrouk2 years ago

I am doing an experiment right now. I got out a bunch of diodes I have and I began testing them with my Fluke. most of them were testing between .5v drop and .7v drop so I started testing them in pairs in series and I found a pair that has a .9v drop. I hooked them up to a fully charged 18650 cell (4.2v) and they gave me 3.29v. I have that running I want to see how long it takes before the electronics get to the unreliable state. I am hoping for 24 hours.

funinalaska (author)  steveastrouk2 years ago

I was thinking about a diode as a voltage drop regulator. What I am not sure of is how to make the power draw be shut off when the battery voltage reached say 3.2v as the electronics will continue to attempt to overate just not reliably and continue to drain the battery all the way down to 1.8v which at a 0.7v drop takes the battery down to 2.5v which is well below where I'd be comfortable allowing the battery to drain.

As far as the solar panel I have seen a couple 5v or 5.5v panels that size wise fit the bill for me both in size as well as in cost. One I keep going back to is a 1w 5v panel that is 9cm X 7cm

a diode is not a voltage drop regulator. it does have a voltage drop but there is no regulation.

You could put a mosfet in series, such that it is on, but when the voltage drops too low, it turns off and so stops the drain on the battery.

funinalaska (author)  funinalaska2 years ago

OK, I figured out the over discharge issue. I'll just use a battery protection board. I have one laying around from a protected battery I had that dies a while ago. I'll test it and see if it is working but that should solve the over discharge issue. Since most of the batteries I have are unprotected cells I will wire it in with the project rather than using it on the battery then it will be there no matter what battery I use.

Have you checked the Ibles on the right, like this one?

funinalaska (author)  Downunder35m2 years ago

Yes I looked at that and many others. They neglect two major components of this project. First the regulates 3v to 3.3v output and the over discharge protection on the battery. Also how to take solar panel output directly into the charge controller for the Lithium battery. Not planning to go from solar panel to 12V charger and battery bank, for 5V buck to 3.7v charge controller. That seems very inefficient for my purposes. and will make the size prohibitive.

A "proper" charger for a single cell usually runs off 5 or 12V for the input anyway.
A solar panel will never give a constant voltage or current on its own.
So wether you like it or not you have to combine different things to make it work.
Even if think of a single circuit it will, in the most basic form, still consist of:
Charge controller, regulated output circuit and a fuse.
So grab yourself a 12V panel, charger, a voltage regulator if the charger does not accept a wide input range and for the actual output a 3.3V regulator.
Add some protection diodes between panel and charger as well as between battery and output.
Sorry but size is a limiting factor you might want to create your circuit or integrate what's missing into the battery charger.


Bun in any case you will have a hard time finding all this for under 20 bucks - unless you go Chinaware and actually ignore the safety features you seek.

funinalaska (author)  Downunder35m2 years ago

That is why in my original question I stated that I have been unable to find a off the shelf board that does everything I need so I am asking for help with building one.

kmossman2 years ago

Doesn't sound so difficult. Use a specialized IC for the battery charger than has the capabilities you want. There are quite a few on the market.

It will be very hard to regulate to 3V. Maybe use a shunt regulator but even with a LDO regulator there is little headroom, especially as the battery discharges.