Beware all, the temp sensor on the Triton cannot save from all lipo disasters!

I've had a TP2100 3S1P pack that has puffed slightly on the last 5 or so flights (just one cell). I charged it ~halfway, then used it to do some 6S current measurements on my EDFs (previously was a 5S setup), noted 14.8A, 19V on one motor, and also noted that the abused cell was getting quite hot. Quickly outside with it to cool for an hour.

Later, I put it on the Triton in parallel with an identical pack, and started them off on a slow charge (0.2A per pack). The Triton's temp sensor was taped to the suspect pack. The cutoff was 100F. Results below. The Triton was just beeping away and displaying "Over Temp".

It seems damaged lipo cells can get in to thermal runaway that the temp sensor cannot detect soon enough.

Later, I put it on the Triton in parallel with an identical pack, and started them off on a slow charge (0.2A per pack). The Triton's temp sensor was taped to the suspect pack. The cutoff was 100F. Results below. The Triton was just beeping away and displaying "Over Temp".

How long were you charging?

Series-charging a pack with a known sick cell in it is playing with fire.

Paralleling the sick-cell pack with a sound pack does absolutely nothing to help the sick-cell pack.

So, fire ensued - no surprise - but two apparently healthy cells were burned up, too.

What's the point?

- RD

The point I got was "don't use cells/packs that have puffed". This question has been asked many times. My XXXX battery puffed up yesterday but it's normal size today. Is it safe to use it? There's your answer.

Well, next time anyone asks what to do with their puffed cell, someone can simply post this picture with the text: "dispose of it safely."

I was charging for about 2 hours, and was nowhere close to fully charged, at about 12.2 volts, C/10 rate.

RD, the whole pack was basically unhealthy, as it now appears 2 cells had actually puffed and burnt - the third was quite disformed after the fire and tossed safely.

I really just did not expect a fire with the temp sensor watching over things - that's my point. Live and learn.

The parallel pack can sometimes help to keep the voltage down when a single cell in one pack gets higher-impedance and has a tendancy to over-shoot earlier in the charge. I was hoping to balance/cycle the pack a bit and test it, then get some more flights from it but in a different model with lower current draw, or use it for bench testing power setups.

I will take apart the other 3S pack, and cycle each cell one by one to avoid the same disaster.

EXACTLY what I would have expected. The pack had puffed up 5 times already - why continue to use it? That's asking for a fire.

From what I gather - Once the thermal reaction starts you can't just cut off the charge and expect it to stop.

Sorry to hear about the fire. Feel lucky that it was only the pack and not your house.

Ted

update: I've taken apart the sister pack bought at the same time, which also showed slight swelling after flights.

I turns out it was a 2S pack, welded tabs, with a 3rd cell soldered on later. The 3rd cell is from a different lot with a slightly different part number! The oddball cell is slightly thinner than the others.

I'm now cycling each cell one by one to see if they are OK.

These cost a small fortune, and I never took them over 12C, mostly running them about 8C after a short 12C draw at launch. 10 flights per pack, maximum.

Does anyone know the correct part number for a TP2100 2nd Gen 10C cell?

These cells read:

CSL603496D 21ED (2 of these)
SL603496 13DK (1 of these)

The only cell to survive the fire was the SL603496 13DK. Both of the other cells exploded and flamed (apparently both the CSL603496D 21EDs).

The packs are made by tperc.com, bought at the LHS as 2nd gen 10C-12C Thunder Power packs. I'm not so sure they are even 2nd Gen packs.

Another update: I took each cell from the surviving sister pack, topped it off at c/10 rate (0.2A) individually and measured the capacity at 1C (2.1A) - all using my Triton.

data format: start voltage under 1C load, capacity at 1C discharge, top-off energy required (imbalance)

CSL603496D 21ED #1: 3.7V, 1008 mA-hour, 220 mA-hour
CSL603496D 21ED #2: 3.7V, 1112 mA-hour, 200 mA-hour
SL603496 13DK: 3.4V, 340 mA-hour, 35 mA-hour

The 2S packs from same supplier, purchased at the same time are still fine at ~1900mA-hour, always run in series with the 3S packs as a 5S setup.

What a mess.

I have a call in to tperc.com on possibly exchanging these, waiting for a call back from the owner.

Aside from verifying once again that puffed cells present a serious safety risk, the point about the Triton temperature probe is another valuable point to be remembered. We've debated in many threads whether or not temperature probes could catch Li-poly cells before they went into thermal runaway. Some people thought that the temperature probe would be a valuable safety device. I think the last thing I said on the subject was about my concerns that the temperature probe could lull some people into not paying as close attention to a charging pack as they would if there was not a temperature probe in use. This incident pretty much verifies my concerns about what could happen if too much faith were placed in a temperature probe to shut down the charging process in time to prevent a fire. I really appreciate you sharing this experience, gfcermak. This is another important lesson to be shared with all Li-poly users.

EDIT: My memory was working right today.

http://www.rcgroups.com/forums/showp...0&postcount=22

Dave, yes and thanks.

Lessions I learned:

1. never charge lipos overnight, even at a 0.2C rate
2. install a loud smoke detector in the garage, tonight before charging any more lipos
3. get a small ceramic bunker for charging to complement the storage bunker
4. unmatched series lipo packs can lead to problems in as few as 5 hard cycles

I got quite lucky in this mess. The batteries were on the aluminum plate on the edge of the bench, and when the bad one puffed and exploded, it rolled off the bench and dangled in free air, dripping flames and molten goo all over the concrete floor. I was at the other end of the house watching tv and did not find it until hours afterwards.

-Gerry

Without reading the entire thread, I had to post a comment..
You were charging and using a pack with a known bad cell?????
..a

AND you were charging unattended on the other side of the house!
This one needs to be stickied as a "what not to do!"
..a

Originally Posted by gfcermak

... the whole pack was basically unhealthy, as it now appears 2 cells had actually puffed and burnt - the third was quite disformed after the fire and tossed safely...The parallel pack can sometimes help to keep the voltage down when a single cell in one pack gets higher-impedance and has a tendancy to over-shoot earlier in the charge...

The deformed cell was likely the bad cell which was at a lower voltage than the two normal cells. The charger read the sum of the voltages on the three cells and the higher two were driven above limit voltage and burned.

All paralleling packs does is clamp both packs at the same voltage.

The charging profile of a sick cell within one of the series-charged packs cannot be modified in the manner suggested.

- RD

Gerry, I thought I recalled reading some posts from you in an earlier thread where you appeared to demonstrate pretty advanced technical knowledge of Li-polys. I just went back and read that thread, and it really seems odd that someone with so much technical knowledge would make several fundamental safety errors that nearly resulted in major collateral damage.

http://www.rcgroups.com/forums/showthread.php?t=253486