Guess the Battery !

[djococaud]

Hi boys and girls !

I have a solar garden light to repair (similar to those http://www.amazon.fr/d%C3%A9tecteur-mouvement-production-Projecteur-rechargeable/dp/B00MBDAKLI/ref=sr_1_7?ie=UTF8&qid=1435064391&sr=8-7&keywords=spot+solaire)
This LED projector was just great and worked fairly good...until now !

The culprit is the battery, which is completely dead ! (0V accross the cell, without the protection circuit)

I can't find anything on the battery markings, just It's a 3.7V 2000mAh...
I would bet It's Li-ion, but not sure (could olso be NIMH...). The charging circuit doesn't help as it's only the solar panel through a simple diode  .

Do you have an idea of the chemistry of this battery ?

Thanks guys !

[EEVblog]

Based on the size and shape that looks very much like a standard 18650 cell with a protection circuit.

[djococaud]

WOW, 2 min answer delay !!! awesome ! 
Thanks, i will check the batteries prices/availability in my electronics local store...

[djQUAN]

Yes, definitely 18650 Li-ion. But I guess the replacement batt will end up the same as it appears the charging "circuit" is relying on the over charge cutoff of the protection PCB (cuts off at 4.5V or higher) rather than terminate at 4.2V as proper chargers do.

I would add a shunt regulator IC / circuit to burn off the excess voltage. A TL431 and a transistor could work.

[amyk]

Batteries marked "ASO" brand is made by a Chinese company with no English name other than "Shenzhen Zhuoneng New Energy": http://www.szznp.com/

They seem quite honest with their capacity, according to this test I found

You can replace it with any other 18650 of same or greater capacity.

[djococaud]

I think I found the actual product I bought a few years ago...http://www.amazon.fr/Projecteur-solaire-led-lumens-panneau/dp/B00GNAUPW4/ref=sr_1_24?ie=UTF8&qid=1435066663&sr=8-24&keywords=projecteur+solaire

Surprisingly they wrote: 3.6V/2000mA NIMH, for 6 to 8 hours of light (I olso checked on solairepratique.com who sales this projector...)

Since the charging circuit is crap (I should say inexistant), I could test a cheaper NIMH and see if I get 6 hours of light... (no big deal if the price is low )

[djococaud]

Thanks Amyk,
So Li-ion It is...
Let's see if my local store has some...

[mikerj]

Thanks Amyk,
So Li-ion It is...
Let's see if my local store has some...

If you aren't going to fix the basic overcharging charging problem then it's very important vital that you use a protected cell or move the protection circuit onto a new cell.

[djococaud]

No worries, I used a (not cheap ! ) camelion ICR18650 Battery with an integrated discharge/overcharge protection.

Since I didn't find a battery with wires I just reused the wires of the old battery (with the connector) And I soldered them on a battery holder...

Just good as new ! (I must still wait a few charge/discharge cycles to be sure...) 





just for info : http://www.camelionbattery.com/attachments/article/133/ICR18650%20Announcement.indd.pdf

[amyk]

Amusing that the datasheet for that cell says in huge type "up to 500 cycles" and then in the small text says "more than 500"...

You may want to check the open-circuit voltage of the solar cell with bright light - if it's constantly hitting the protection cutoff the cell will be (slightly) overcharged every time and fail much earlier. Adding a shunt regulator would be a good idea.

For this sort of "accumulator"-type application, limiting charge voltage to a bit less than 4.2V will result in longer life.

[djococaud]

Indeed...well I will see...if it lasts 2 more years, then I will just replace the battery in two years and don't bother modify the circuit!   
And if it fails earlier...I will upgrade to a better system (with citizen LEDs and Buck driver...or anything more efficient than the current linear driver in this thing...)