My first day with Li-Ions, could have been worse

[vortex69]

Never thought my first post on cpf would be on this section...

So 2 or 3 weeks ago I ordered 2 cheap unprotected 18650 on eBay, today they finally arrived, first thing I did was check the voltages.
Both were around 3.5v, great, let's charge them, so I put them on the charger, and since they were unprotected, I was being careful and measured the voltage every 30-45 minutes and checked if either the batteries or the charger were getting hot, nope, all fine.

------ Here ends the "being very careful, safety first" part of this post -----

Meanwhile they finished charging and being an "inventive" guy, just for fun, I decide to convert a crappy 3xAAA flashlight to 1x18650 <--- First signs of a Bad Idea!
So I measure the "victim" (flashlight), inside diameter ~23mm, depth ~68mm, good, a 18650 will fit without problem, a bit of paper to fill the diameter and it's done.

So without much more planning or modifying I stick the battery in there and start screwing the tailcap, suddenly the light turns on - it scared me, but I clicked the switch and it turned off, good, thought it had short circuited, but no, at least not for now - so I finish screwing the tailcap, click the switch and a nice light starts shining (as nice as possible on a cheap, 5mm LED flashlight), nice, let's turn it off and put the 18650 in my battery storage box.

*click* "Hmmm, doesn't turn off" *click* "Oh, *censored*, short circuit."
I unscrew the tailcap as fast as I can, remove the battery and as I remove it I see a bit of smoke coming from inside the flashlight (all of this in ~5 seconds), I go outside as fast as I can and put the battery on the concrete floor, and stay a few meters away looking at it for a few seconds.

-------- Here ends the "Oh, *censored*" part of this post -----

Good, no damage.
So, what the heck went wrong?
I start looking at the inside of flashlight and notice the spring on the positive side, instead of being pointy side to the battery, flat side to the led circuit, it was flat side to the battery pointy side to the circuit.
I guess that, when I was screwing the tailcap I was putting too much pressure on the battery (didn't measure depth correctly), which was putting too much pressure on the spring, which caused the spring to spread or slide with enough strength to get under the protective plastic on the top of the battery and cause a short, another problem is that I didn't make sure the battery was tight enough on the sides, it could still move a bit.

Of course, all this text and no pics isn't acceptable (bad quality pics, but pics, be happy with that )

The spring that (probably) caused it all:

The poor battery:

So while this was quite a stupid error, I feel even more stupid because I was being as careful as possible until then.
No fire, no explosion, no damage, no nothing, only one (probably) dead battery and a bit of smoke, the battery didn't even get warm, impressive I would say, still, I better not keep pushing my luck and stop shoving 18650 into slightly modified 3xAAA flashlights.

In the end, my first day with Li-Ions, could have been worse.

*edit*
Some updated info in post #7

 

[SilverFox]

Hello Vortex69,

Welcome to CPF.

Actually replacing 3 NiMh cells with 1 Li-Ion cell is a good idea and a great modification.

However, as you have learned, you have to be careful to make sure everything lines up properly.

Tom

 

[LuxLuthor]

Vortex, actually a wonderful first post. You did show a reasonable amount of knowledge and care about what you were potentially dealing with, and give others experience of how quickly it began to develop problems. Best step since you had time was running outside and put it on concrete.

In some cases, there may not be enough time to even do that.

I'm also quite sure that even if you had a protected battery, the type of short you demonstrate between the (+) top and (-) upper shell would have resulted in a similar result. Once you tear the wrapper around the top of the metal can, allowing a dead short with the (+) nipple, you are bypassing the protection circuit.

Again, many thanks for sharing your experience.

:welcome:

 

[VidPro]

if at first you dont succeed,remember thomas edison caught a train on fire
http://www.nps.gov/edis/faqs.htmhttp://www.high-speed-internet-access-guide.com/articles/biography-of-thomas-edison.html

 

[mdocod]

Great first post vortex69!

Everything has pretty well covered the safety aspects and how well you handled things so I wanted to point out one other thing....

In my experience, flashlights that are built to run on 3xAAA cells in those little carriers rely on the high internal resistance of those very low capacity cells, and at the same time, depend on the high resistance in the carrier. Most of the shower-head style 5mm led flashlight that run on the 3xAAA pack are going to just run in direct drive or with a low value resistor in series with the circuit.

The 18650 size li-ion cell, while having a similar nominal voltage rating, will often have far lower effective resistance than the 3xAAA pack.

Using the DMM set in current measuring mode, you can complete the circuit from the body of the flashlight to the tail end of the battery pack and see how much juice is flowing. Most of the shower-head units are designed to start off around 20-40mA per LED in the head when running on alkaline AAAs. I've found that swapping to the 18650 can bring this up to 50-150mA or higher per LED in a hurry. When left on for more than a few seconds at a time, the LEDs will start to go into thermal runaway and draw more and more current until something gives.

The solution is often just swapping to a higher value resistor, or, picking out a buck regulator that would normally be found in a typical 1xHighFLux LED flashlight. (Picking something with an output current that is appropriate for the number of LEDs it's going to be divided up amongst.)

Eric

 

[Chauncey Gardner]

As said before, great first post & welcome to cpf:welcome:.

I have been following the liion battery threads pretty closely and can tell you I bought what appear to be the same batteries (from possibly the same seller) just to test them out.

So far no issues, and that includes falling asleep with one in a WF-139 charger overnight. Came off at 4.20v (with no issues as all of my 18650's do), protected or not. So far the batteries (the one in your pic) have worked fine and I keep an eye on the voltage (both in use and on the charger).

I also have begun to put the charger in a clay flowerpot (thanks livesforphysics for the tip) to err on the safe side.



What make & model of charger were you using on the unprotected 18650 if you don't mind.

Thanks & welcome. Sorry for your mishaps!


I know how closely some of us are following this topic so here is the battery shown above:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=130333407755&ssPageName=STRK:MEWNX:IT


To Eric, Silverfox or livesforphysics, I'd be happy to donate one of these & a couple other used liion's. pm if interested.

 

[vortex69]

Thanks for the welcome

mdocod: To be honest I never even thought about getting a driver to use in this flashlight, thanks for the idea, will get one if I ever try to modify a 5mm led flashlight again.

Chauncey Gardner: That's exactly the same battery and the same seller too , as for the charger I use a Trustfire TR-001, using the 12v input, too many reports of those chargers getting fried when connected to mains power, nice idea about the clay flower pot.

-------------------------------------------------

Also, since that first post was posted around 10 minutes after it happened, it lacked "further investigation".
The conclusions I ended with might be wrong, I don't know, if someone spots something not logical please correct.

Looks like one error ended up preventing bigger trouble.
For that modification I had to disassemble the tailcap, the switch has 2 metal contacts (one on each side), one for the - pole and the other to make contact with the body of the flashlight, one of them is covered by a bit of plastic (the battery one) so it doesn't override the switch.

The problem is that you can mount it the wrong way too, and that was what happened, when the light turned on the first time and I clicked the switch, I probably moved the tailcap (or the switch) enough for it to stop making contact, thus the light turned off (the tailcap was barely screwed, only maybe half of a turn).

So falsely confident about the switch I screwed the tailcap again, leading to the "Oh, *censored*" part of my first post.

In the end while it was quite a stupid error, mounting the switch the wrong way was good (at least I think so), since it made the flashlight stay always on, which made me freak out and open the flashlight as quickly as possible, completely unaware about the short on the + side of the battery.

Also it seems the battery survived the entire thing, the next morning, before throwing it away, I measured the voltage and it had around 4.0x volts (can't remember the exact number)

 

[Chauncey Gardner]

Thanks for the welcome

mdocod: To be honest I never even thought about getting a driver to use in this flashlight, thanks for the idea, will get one if I ever try to modify a 5mm led flashlight again.

Chauncey Gardner: That's exactly the same battery and the same seller too , as for the charger I use a Trustfire TR-001, using the 12v input, too many reports of those chargers getting fried when connected to mains power, nice idea about the clay flower pot.

-------------------------------------------------

Also, since that first post was posted around 10 minutes after it happened, it lacked "further investigation".
The conclusions I ended with might be wrong, I don't know, if someone spots something not logical please correct.

Looks like one error ended up preventing bigger trouble.
For that modification I had to disassemble the tailcap, the switch has 2 metal contacts (one on each side), one for the - pole and the other to make contact with the body of the flashlight, one of them is covered by a bit of plastic (the battery one) so it doesn't override the switch.

The problem is that you can mount it the wrong way too, and that was what happened, when the light turned on the first time and I clicked the switch, I probably moved the tailcap (or the switch) enough for it to stop making contact, thus the light turned off (the tailcap was barely screwed, only maybe half of a turn).

So falsely confident about the switch I screwed the tailcap again, leading to the "Oh, *censored*" part of my first post.

In the end while it was quite a stupid error, mounting the switch the wrong way was good (at least I think so), since it made the flashlight stay always on, which made me freak out and open the flashlight as quickly as possible, completely unaware about the short on the + side of the battery.

Also it seems the battery survived the entire thing, the next morning, before throwing it away, I measured the voltage and it had around 4.0x volts (can't remember the exact number)

Thanks for responding Vortex.

I have picked up 4 different brands to check out on the charger I have
to ensure the setup I'm using is safe and effective.

I'd happily spring for a more expensive charger, but figured I'd try it out with the cheapest stuff I could find to test it out (protected and unprotected).

My suggestion would be to ditch what you have. Leaving the unprotected liion's on the charger is not something I want to worry about.

Checking AW, Ultrafire, WOW & those cheapie blue unprotected batteries we both have.