Solar charging lithium batteries, EEs have a sec?

[TT_Vert]

I had a ton of spare 18650s, spare LEDs, a 3D printer, a passion for designing stuff and some time on my hands. I have a solar panel that puts out 5V/500mA that I am charging a 3S2P config. I went with a boost convertor that uses an XL6009. I am using two BMS' that do 3S balancing. They have over/under voltage as well as short circuit protection. Basically I am going from the 5V/500mA solar panel to the boost convertor which is boosting the voltage to 14V which is then feeding the two BMS' I have for each set of 3S 18650 packs. Is this an acceptable/safe way to do this? While bench testing I set my power supply to a current limit of 300mA and 5V which was about the average output I was seeing out of my solar panel. With this setup I'm drawing about 300MA from the bench power supply with a voltage of somewhere under 4V as it was current limiting. Based on my calculations using this solar panel (5V/500mA) , with boosting from 5-14V and about 90% efficiency I'm seeing best case 160mA to the pair of BMS's I'm using. Given the capacity of the 2P3S config and the fact the LEDS are only used for a few minutes a week or so and draw 1A I don't think this charge rate will ever be an issue. Is my math accurate here?

Here is my finished product, so far it seems to be functioning as design but I'm just paranoid about charging Li batteries unattended. Pay no attention to the bodge wire

 

[jkeyser14]

Lithium Ion cells want to be charged at constant current (normally 1C or less) until they hit a voltage of 4.15V. Then the charger needs to switch to constant voltage at 4.2V and for an 18650 sized cell should have the charging cut off when the current draw (at 4.2V) tapers down to 150 mA.

Failure to do this can degrade or damage cells. Charging at less than 1C while following the above charge profile should be perfectly fine.

 

[TT_Vert]

I will be charging WELL below 1C but will always be CC never CV. With that, CC gets you 80%+ of the way charged and I believe there are even some chargers that still do charge CV only and no CC. From what I've read you prolong the life of a lithium cell by keeping it below 4.2V max. I hear 4.1V is the sweet spot for capacity and longevity and I think a CC at a low current like I have will get me there. Appreciate the info and any other info anyone has.

Dave

 

[gpiggaz]

In my experience- I built a solar address light with a ready made charge controller and solar panel. Indeed the sweet spot is around 4.1V and if you size the pack so that it only discharges partially, I think mine is only cycling to about 3.9V due to my sizing the pack and solar panel/controller to keep it in that bucket, the cells will last a long time. I have my Solar address light running for I think about 6 years now and interestingly, I have only had to change the actual LED light in that time. It turns on at dusk and I have it set to turn off 6 hours later- all built into the charge controller. I think I have a lot more $ invested in it than off the shelf solutions, but it was a fun project non-the-less.

The hardest part of this project was finding a solar charge controller that had a Lithium Ion setting- most of them are SLA charge controllers.

Lithium controllers are more available now though- here's one on eBay https://www.ebay.com/itm/203529783389?epid=21036903286&hash=item2f6352085d:g:tBcAAOSwC6NhAkbO

 

[TT_Vert]

In my experience- I built a solar address light with a ready made charge controller and solar panel. Indeed the sweet spot is around 4.1V and if you size the pack so that it only discharges partially, I think mine is only cycling to about 3.9V due to my sizing the pack and solar panel/controller to keep it in that bucket, the cells will last a long time. I have my Solar address light running for I think about 6 years now and interestingly, I have only had to change the actual LED light in that time. It turns on at dusk and I have it set to turn off 6 hours later- all built into the charge controller. I think I have a lot more $ invested in it than off the shelf solutions, but it was a fun project non-the-less.

The hardest part of this project was finding a solar charge controller that had a Lithium Ion setting- most of them are SLA charge controllers.

Lithium controllers are more available now though- here's one on eBay https://www.ebay.com/itm/203529783389?epid=21036903286&hash=item2f6352085d:g:tBcAAOSwC6NhAkbO

Interesting. So your charge controller also had an integral timer? I definitely needed something in a much smaller form factor to fit into the footprint of my solar panel which was about 135*95mm but batteries have to fit in there also. It seems this needs 12V input but what is the minimum current it can use? What panel did you use and what was its output?

Dave

 

[gpiggaz]

Interesting. So your charge controller also had an integral timer? I definitely needed something in a much smaller form factor to fit into the footprint of my solar panel which was about 135*95mm but batteries have to fit in there also. It seems this needs 12V input but what is the minimum current it can use? What panel did you use and what was its output?

Dave

Yes- it has an internal timer- I didn't look closely, but I suspect that one I linked does too. I'm sure the one I bought is obsolete now.

My solar panel was a bit larger - its a 15W panel- I think its 24V OC
My battery pack was 4S1P or about 14.8V and the cells are 3200mAH cells

I did this in 2016- and again, I have only replaced the actual LED floodlight since then.

I did add a waterproof enclosure and a small panel meter to actually see the voltage and current, each added to the cost

I spent about $150 on this.

Hope the info helps you.

Attached is a picture of my set up spread out before I "packaged" it.

After I "packaged" it:

 

[TT_Vert]

What voltage was the flood light you were powering? If you got that much runtime I have to assume LED?

 

[American Locomotive]

The only thing I will say is make sure your BMS and boost converter handle low voltage input gracefully. Make sure they don't behave erratically if the input voltage drops then comes back up (like if a cloud went by overhead)

 

[TT_Vert]

The only thing I will say is make sure your BMS and boost converter handle low voltage input gracefully. Make sure they don't behave erratically if the input voltage drops then comes back up (like if a cloud went by overhead)

It seems when the input current/voltage gets low the BMS continues to run off of the batteries until I think 9.x volts then it shuts off to make sure you check the batteries. You have to then restart it by applying 12V to the input/output terminals (P1/P2) I think they are called. What do you mean by erratic? I'd assume it'd just drop the current and then come back up. At most I'm seeing 500mA out of that panel.

Dave

 

[American Locomotive]

Some electronics do not behave well when the input voltage slowly drops or raises. They can get "stuck" in bad states and do unexpected things. So I'd slowly bring up the voltage using the bench power supply from 0v, then gradually drop it back down to 0, and make sure nothing crazy happens. Then I'd also bring it up to 5v, momentarilly drop the voltage down to ~1-2v, and then bring it back up to 5v and ensure it recovers and everything starts back up correctly. The last thing you want is a cloud to go come by and have the boost converter go nuts and start over-volting when the panel gets sun again.

 

[gpiggaz]

What voltage was the flood light you were powering? If you got that much runtime I have to assume LED?

Yes, LED and it's "12V" probably why it might have not lasted as long as it could have, The new one is over a year old and still going strong too though.