Battery Pack 2x18650 3600mAh 3.7v 13.32Wh on solar led flood light not charging

Greetings everyone:

I have the following solar led flood light that seems to have stopped charging as I probably left it in the on position for months before putting it on the off position to charge instead of charging as soon as there is no energy left.

https://www.lowes.com/pd/All-Pro-18...ion-Activated-Flood-Light-with-Timer/50278197

When I opened it up, it is basically a battery pack that is labeled "GREAT POWER ICR18650 3600mAh 3.7V 13.32Wh", when I took off the wrapper, it is basically two 18650 batteries that has no writing on it that seems to be connected (+) to (+) and (-) to (-) between both batteries and then the (-) and (+) side goes to a circuit board that connects back to the mainboard and then the solar panel also connects to the mainboard. I have a BCT-3100 v2.2 charger. It appears the batteries are basically connected in parallel so they are really 2 x 18650 1800mAh batteries. In any case, it seems like the batteries are welded together so how does one separate it and can one just solder the tab back to the battery after recharging or is there another way to do it? As the solar panel just connects to a barrel connector so it seems I can use a DC Adapter to charge the batteries. Thanks!

They're probably generic ChiCom cells of dubious quality and since they're actually marked ICR (lithium-cobalt-round) I'll take them at their word and assume that they're now junkers, to be recycled.

If you have a voltmeter, you can measure the voltage and if below say 2.8v each, they're toast and don't bother mucking with them.

Now, you can buy tabbed 18650s from places like BatterySpace and go that route, but generally you need a tab/spot welder to get things securely in place. If you have a really good soldering iron, you might be able to rough up the ends and get the solder to stick, but you don't want stuff shorting out while it's in 'charge' mode.

When we get bored, we rip apart laptop packs and scavenge 18650s for those and they're tabbed. I just carefully pull the strings apart and use some small vice grips to pull the tabs off the ends, then hit them with a Dremel tool to smooth them back out.

I've never tried putting things back, as I'm better at destroying things.

If the batteries ran don't to 0.00v, don't bother charging them up as true ICR cells can vent with flame after sitting for a while at a highly discharged state.

Chris

Thanks ChrisGarrett, either they are junkers or they just got overdrained. Without dewelding it, I tried placing one of the cells into the BTC3100 and basically what happened is it said 4.2volts and then was trying to charge at 60mA but then within 10 seconds, it would say full, this is probably because there is still a second cell connected in parallel. Is there a difference between ICR 18650's vs the 18650's like the NiteCore rechargeables used in flashlights? I took apart batteries just like you did maybe a few years ago by using small needle pliers to pull off the tabs but then the devices were not kept so didn't really have to put the battery back together either. I always thought solar lights used battery packs that used rechargeable NiMH AA's so was surprised it was 18650's inside. Can I actually use 3xAA's wired in series instead which would seem to give a higher mAh but the voltage would be 3.6v instead of 3.7v meaning it's going to be 0.1v smaller but would that even make much of a difference but ofcourse the capacity will be at whatever the highest AA capacity is, for e Eneloop Pro, it'll probably be 2500mAh or so.

If I understand correctly that they're still welded together, then you actually inserted both into the charger, from an electrical perspective. One is inserted directly, the other is just as connected, but by way of the tab.

Theoretically you could charge two cells that way.

If the pack was fully drained, but showing 4.2V almost immediately after getting placed into the charger, then I'm worried that the capacity of the cells has dropped to nearly zero. Charger fills up nearly zero pretty fast, which is why you see 4.2V.

If that's the case, the cells are of no use, and possibly dangerous. Personally I wouldn't do more tests at this point, the cells can have gotten to the point that it would now be dangerous to add load to them by using them.

I'd say it's just not worth the risk. There are several companies out there though, that will let you pick the cells you want, and then they'll make a custom pack for you.

You need to measure individual cells and see what they show, voltage wise, to be sure. ICR/li-cobalt cells, if truly using cobalt don't like to sit in a discharged state for too long, since the copper separator can form copper dendrites and then when you go to charge them up, you get a short.

Not knowing all the details, I would just caution you to either get a new light, or buy tabbed 18650s. You can't use NiMH batteries, because the BMS (battery management/charging system) is pegged to li-ion cells.

It sounds like your Opus saw a 'fully charged' cell, whether both cells were at 2.1v while still tabbed together, or not. While they're mechanically linked together in parallel, we won't know for sure.

Anyhow, that's how I see it.

Here are the tabbed Panasonic 2900 PFs:

http://www.batteryspace.com/hi-powe...-44wh---ncr18650pf-0-87---un-38-3-passed.aspx

Chris

terjee, I figured the tab basically is seeing both cells as one because of the tab.

I still don't know if it is the pack that is fully drained or if it is a problem with the rest of the light as if the light was in the off position for some time and I turn it to on, the light would actually light up and immediately turn off which normally would mean the battery is completely drained but still has some current left.

ChirsGarrett, I am thinking about getting a new light under warranty but just wanted to see if there was a way I can charge the battery assuming it's good but the problem is how to get the battery connected back together after removing it off the tab.

The original is a ICR, what type of battery is the Panasonic you linked? I know this is what I'm using in my NiteCore flashlight:
https://www.amazon.com/dp/B01FBAAMDS/?tag=cpf0b6-20

Also, assuming I get the tabbed Panasonic 2900 PFs, do I still need to hook it up to the little circuit board that is the width of the solder and basically seems to just take the solder tabs and then black and red wires run to a connector with a plug that plugs back into the main board or does the Panasonic 2900 PF not even need that and just needs to either have the plug or I can just splice the red and black to plug it into the main board? Is that little board on the battery pack actually to provide protection while the cells themselves are unprotected?


free upload no registration

I tried to tear off the tab from one of the cells but basically disconnected the same cell by breaking the tab and lifting it so it no longer contacts the battery and with my Fluke 87v, battery 1 measures 4.229v while battery 2 measures 4.231v. I made sure when I measured that it is not bridged to the other battery so it seems the batteries are still actually full when it comes to capacity as the charger showed the same v4.2 when inserted and the tabs were disconnected.

Almighty1 said:

I tried to tear off the tab from one of the cells but basically disconnected the same cell by breaking the tab and lifting it so it no longer contacts the battery and with my Fluke 87v, battery 1 measures 4.229v while battery 2 measures 4.231v. I made sure when I measured that it is not bridged to the other battery so it seems the batteries are still actually full when it comes to capacity as the charger showed the same v4.2 when inserted and the tabs were disconnected.

Click to expand...

Now we've really screwed the pooch.

So basically you've broken the tab welds and have two separate 18650 cells that test at just over 4.20v, but are within spec (3.950v-4.250v).

They've been charged a bit high, but now your battery pack is broken?

What was the problem in the first place? What made you think that the solar panel wasn't charging them up? They seem to be full, so either the pack>light connection is broken, or your Fluke 87 is broken, or your Opus 3100 can charge up cells in the blink of an eye?

At this point, you might as well just try and get a refund. Your initial reaction to the Opus showing ~4.20v with cells welded in parallel was correct. The Opus was telling you that that cell was actually full, even though you/we couldn't be 100% sure because they were still joined: 4.2v/2=2.1v and that means a rejection.

I'm guessing that there's something wrong with the pack>light I/O switch on the PCB somewhere and the light's not coming on at dark, although the unit is charging the cells during the day via the solar panel and its I/O, but my degree is in English Lit. and Communications, so what do I know?

Chris

They're in parallel, not series, are they not? (Both pluses together, both minuses together).
That means a perfectly fine fully charged battery should show 4.2V. If they'd been in series (plus on one connected to minus on the other), it would make sense to expect the double.

What I can't really guesstimate here, is the risk of them being damaged. It seems entirely possible that they were depleted for a while, which could have rendered them dangerous. For that reason alone, I'd swap them.

There really only are two categories for ICR cells: those you trust, and those you should regard as potentially dangerous.

No formal electronics education, but I am into DIY electronics, designing circuit boards and the like, a licensed HAM operator, and FreeBSD user since around 2.1.1.

ChrisGarrett:

Battery pack is only semi-broken since as long as I can solder it back, then all is fine.

The problem was that the light is supposed to turn on and stay on for one minute but currently, it seems to only turn on and then go off less than a second later and it seems like there is a waiting period to get it to even turn on and then go off less than a second later, have not figured out how long the waiting period is. I doubt the Fluke 87v5 is broken and the Opus 3100 is not the issue either.

Can't get a refund as the light has been purchased for more than a year so can only be replaced under warranty which is what I'll probably do. Either the switch can be defective or it's that PCB itself. I thought it drained completely and just didn't have enough voltage to charge like how AA/AAA batteries are that requires connecting a fully charged battery in parallel just to get the voltage up so it will be recognize by the charger instead of null and begin charging. Guess that is not the case. I actually have another of the same light unopened in the box and can probably experiment by swapping the battery packs and see what happens with the new light. I just thought the problem was either the battery or the solar panel but who knows, the PCB could just be low quality since the two capacitors appears to be okay on the front of the PCB but who knows what can go wrong on the back side of the PCB.

terjee:

Yes, they are in parallel and not in series. I just thought that if one cell was bad in series, it can still be 4.2v.

You are right about those you trust and those you don't trust. Are the rechargeable cells used in flashlights like the 3100mAh-3500mAh actually ICR as well since it doesn't say on the battery itself other than it's high capacity.

I'm just a Astrophysicist with no formal electronics education and a FreeBSD user since 1.0-GAMMA, it was jkh (Jordan K. Hubbard) of the 386bsd before FreeBSD was available since I was interested in running 386bsd + patchkit back then.

Almighty1 said:

Yes, they are in parallel and not in series. I just thought that if one cell was bad in series, it can still be 4.2v.

Click to expand...

It certainly can, but it's not a likely and full explanation of the symptoms alone.

If one cell is failed open (as opposed to fail short for example), then the voltage could still be 4.2V. It's unlikely, but not at all impossible.

Also, what I'd expect then, is for the light to perform more or less as normal, just last bit under half as long before the battery was drained. Theoretically it's possible that one cell has failed open, the other isn't up to keeping the voltage as high as the light wants under load, and so you'd see a more rapid shutdown.

Almighty1 said:

I'm just a Astrophysicist with no formal electronics education and a FreeBSD user since 1.0-GAMMA, it was jkh (Jordan K. Hubbard) of the 386bsd before FreeBSD was available since I was interested in running 386bsd + patchkit back then.

Click to expand...

Yeah, you mentioned in another thread, but I forgot to mention 2.1.1 when replying there, sorry. It's not often I meet people that used FreeBSD that far back, so hat off to you etc.

I received a advanced replacement from Lowes via FedEx today and basically it seems like the replacement unit is made on December 2, 2014 instead of July 4, 2014 like the original. It had the same UPC, same model # as well as same product code. The replacement on the box seems to indicate it is 1000 lumens instead of 850 lumens for brightness and since it is 2 lamps, I would assume each one is 1/2 the total lumens. There are basically 10 leds on each lamp so it's 50 lumens each on the replacement instead of 42.5 lumens each on the original. The battery pack is still a ICR 18650 with the exact same 3600mAh capacity except this time it is Samsung and has a Q.C. passed label. It appears from testing, swapping the battery pack with the Samsung on the original lamp works so the original battery pack may test 4.2v but may actually not have enough current since it was similar to when my laptop power supply stopped working and while the voltage and even current tests correctly, there is a lack of load to test with just the digital multimeter. Ofcourse I wonder who actually designed these flood lights since it seems like everytime the battery needs to be recharged, you have to turn the light off before it will recharged as it will not recharge with the light in the on position and remember these are flood lights so usually for others, it is usually mounted in areas not easily reachable without using a ladder, I don't think other people really will be toggling the switch off everytime the battery is out of power and then wait four days atleast to fully recharge the battery before turning the light back on again, since the light is supposed to only turn on during night time anyways, makes you wonder why it couldn't just recharge with the switch in the on position during the daytime.

LOL, I remember someone commented about FreeBSD in another thread but didn't remember the name, sorry and then you mentioned FreeBSD here so thought I commented as well. I haven't been running FreeBSD or rather it is on pause since my Comcast/XFinity Internet connection since last March 2016 does not have a static IP and Im trying to run a mail server so need to probably do Static IP via VPN in order for the box to run again. Glad to meet another FreeBSD user here too so hats off to you too! =)