Please post all the ways a battery has been killed

[spinningmagnets]

Except for fires, there is a separate thread for that.

There was recently a posting from a friend who left his E-bike outside and there was an unexpected light shower, so the battery got wet even though it was inside a bag. There's several ways in which wetness could damage a battery, but needless to say wetness in any form (condensing humidity, etc) is very bad.

Someone posted recently that they left their E-bike system plugged in, and a "resistor" drained the voltage down past where it could be revived, so it sounded like...if I have an on/off switch for my system, that will prevent this kind of problem? (or was it a charger left on?) Of course if I want to put the bike away for a few months in winter, I would disconnect the battery and remove it from the bike (set voltage to halfway, and store in a cool dry place).

Thanks in advance...

 

[steelmesh]

It sounds like you have a good idea of these issues.

Corrosion like oxidation, electrochemical processes (H2O+minerals = electrolyte) and damaging your pack's structural integrity (ie. wet duct tape) would be more of the enemy when it comes to getting your pack wet. The packs should be well sealed from humidity and an accidental soaking, but probably not rated for sitting in pools of water or soggy environments.

Salt water test, electrolysis on packs coated with/without corrosionx:
http://www.youtube.com/watch?v=RVxDfQxgE1U

The most surefire way to prevent discharging your packs is to disconnect all battery connectors, including power and balance leads. You could have a large solenoid to cut off the power, but you have to make sure there isn't a parasitic draw (like from a DC-DC converter or BMS). Also consider mechanical damage and fatigue (like vibrations damaging solder joints). HVC, LVC, charge temperature, storage temperature, storage charge also have an impact on pack life, someone posted that video on premature lithium battery failures presentation. He recommended staying below 40C(104F) and not charging past 4.1 volts, 4.0 even better.

 

[dnmun]

wetness did not damage that pack. the BMS got wet and is leaking current onto the signal lines for the LVC so the BMS has shut off. if he had just taken the BMS out and dried it then it would work again.

not sure why someone said there is a resistor that drains the pack down. people have all kinds of opinions about why their battery dies but usually they always say the BMS is broken, without any knowledge or testing. just the stock answer.

 

[cal3thousand]

It should be more clearly stated that some BMSes (actually quite many) draw their power from only a --cell or two-- SUBSET of cells. That eliminates the need to have a separate DC-DC step-down circuit built in, making them cheaper to produce. But that has the nasty side effect of draining --those 1 or 2 cells-- that subset of cells. Left attached and powered on, many of these Battery Management Systems have earned the Battery Murder System moniker.


Edit: Fixed to be a more accurate statement

 

[spinningmagnets]

This post seems to suggest that if you charge up a battery, and then a timer turns off the house current, but the charger is still hooked up to the battery, then over time the LED on the charger can drain down the battery. Maybe not in a day, but...apparently over a few months can certainly kill a battery dead.

Dogman said he leaves his charger on all the time and hooked up to the E-bike battery unless he's riding the bike. 24/7. If the charger ever dies, it's MUCH cheaper to replace than the battery.

There may also be a "bleed down" resistor across the capacitors in the controller. They need to be charged up when plugging the battery into the controller, which causes a famous spark for 48V or more (requiring an anti-spark circuit). So...if the battery is hooked up to the controller and the charger is off, the controller can drain down the battery?

 

[dnmun]

5k ohms, 58V down to 40V, average 49V. 9.8mA from 20Ah pack is 2,041 hours or 85 days with the battery discharging through the drain down resistor on the input caps. but after that the BMS will turn off the output mosfets for LVC and the battery will remain there discharged but not ruined until charged up again.

there is no BMS that takes power of the first cell or the first two cells. none.

the Bestechpower BMSs i am using now take power off the top of #3 and the ping signalabs take power off of #4.

 

[cal3thousand]

5k ohms, 58V down to 40V, average 49V. 9.8mA from 20Ah pack is 2,041 hours or 85 days with the battery discharging through the drain down resistor on the input caps. but after that the BMS will turn off the output mosfets for LVC and the battery will remain there discharged but not ruined until charged up again.

there is no BMS that takes power of the first cell or the first two cells. none.

the Bestechpower BMSs i am using now take power off the top of #3 and the ping signalabs take power off of #4.

I never said 1st 2 cells ... EVER

I said 1 or 2 cells. Your indication of #3 or #4 cells is quite specific, but no more correct than my assertion.

Here's my post again in case you missed it:

It should be more clearly stated that some BMSes (actually quite many) draw their power from only a cell or two. That eliminates the need to have a separate DC-DC step-down circuit built in, making them cheaper to produce. But that has the nasty side effect of draining those 1 or 2 cells. Left attached and powered on, many of these Battery Management Systems have earned the Battery Murder System moniker.

 

[dnmun]

the circuit current is derived from a voltage between ground, which is B-, and the top of the cell it is taken from.

if it is taken from one cell or from 2 cells then it has to be taken from #1 or #2.

it cannot just be taken from one cell out in the middle of the pack.

there is no BMS that can take current from the top of #2 and use that current to drive any of the available mosfets used in the output of any BMS. they all require at least 5V to turn on.

 

[dogman dan]

So what made all those cheap 18650 cell packs that got shipped by slow boat for months show up with overdischarged 1 or 1 and 2 cells? Same thing seemed to happen a lot to packs stored over long periods, like while deployed to Iraq.

Way's I've killed batteries.

Lead, discharged too deep

NiCad, run at too high amps

Lico, dinged or punctured

Lico, overdischarged when a paralleled battery disconnected. Those went to 0v.

Pingbattery lifepo4 left controller turned on, with the battery 90% discharged. The entire battery discharged to 0v. BMS should not have allowed that, but it was very old by then. That was the day I came home with West Nile, and was not thinking with a brain virus. Normally I would have unplugged, and started charging.

Ways I've damaged or impaired a battery.

Ping, run at too high amps

Lico, run at too high amps and discharged deep. But kept above 2.7v.

I have also made the mistake of keeping lifepo4 in a hot garage. The pings still lasted 3 or more years, but could have lasted longer. The pingbatteries were kept fully charged because I used them 2 cycles per day.

 

[spinningmagnets]

Thanks everyone, and especially you dogman. Your list is exactly what I am looking for...as I find more when reading through the threads, I will post them here.

 

[liveforphysics]

Ground the pouches open from contact with tire from forgetting to bolt-down the battery mount in a race.

Crashes damaging cells from impact with the ground.

During normal use (which keep in mind most of you guy's aren't pulling >50C discharge and occasional 5C charging, so I expect my stuff to decay a bit quicker than most).

Nano-tech's infant mortality issues...

Nano-tech's puffing issues...

Nano-tech's self-discharge issues...

The sooner you can move to some high manufacturing quality and materials cell with excellent QC screening, the sooner you will get to only have the failure modes you induce yourself through poor management. It's pretty nice.

 

[dogman dan]

For some reason, my recent purchases of the cheapest possible 20c lico from Hobby King has been 100% good packs.

Years ago, a purchase of three packs of a particular cell size was 100% puffers. I didn't mention batteries that came killed in the other post, just ones I killed.

It seems that the hobby lico is somewhat a matter of luck, getting a good batch or not.

 

[drsolly]

I use Hobbyking Lipos, 4s 20c hardcase, because they're the best value for money that I could find. I put a pair in series to give me the volts for my 24v motor, and a pair is 5AH. When that pair is exhausted, I plug in another pair, and so on. I have seven such pairs, and I've never needed more that five ... so far. It's not that I travel long distances; it's that with geocaching you're starting and stopping every couple of hundred yards, and you're often going over soft squishy grass/mud, so I get terrible mileage numbers.

To tell me when a pair is exhausted, I plug in an alarm gadget into the balance ports. I have it set to beep at 3.2v. What could possibly go wrong with that?

Here's what. When an idiot plugs in a new battery, removes the alarms from the old battery, and then plugs them back into the old battery. which by now has recovered enough to be just above the 3.2v beeping level. So now I'm running on a battery, and I have nothing to warn me when it's depleted.

Until the bike motor stopped working. I very quickly determined the cause, swapped to the next battery, and carried on.

When I got home, I use Battery Medic to measure the cell voltages for each of the cells in the batteries. It wasn't as bad as I feared; the voltages were between 3 and 3.2 volts. Saved by the low voltage limit on the controller!

But when I tried to recharge those batteries, the charger reported that there were only three cells (in a 4s battery). It wasn't that a cell was dead, because Battery Medic would have told me. I couldn't work out what to do, so I powered everything off and went to bed.

Next day, I plugged in the charger set to "Charge" (I usually set it to "Balance Charge") and that worked. After several minutes, I did my usual balance charge on those batteries, and it looks like I've gotten away with my blunder.

Now I need to think of a way to idiot-proof my changing of the batteries.

 

[cal3thousand]

I use Hobbyking Lipos, 4s 20c hardcase, because they're the best value for money that I could find. I put a pair in series to give me the volts for my 24v motor, and a pair is 5AH. When that pair is exhausted, I plug in another pair, and so on. I have seven such pairs, and I've never needed more that five ... so far. It's not that I travel long distances; it's that with geocaching you're starting and stopping every couple of hundred yards, and you're often going over soft squishy grass/mud, so I get terrible mileage numbers.

To tell me when a pair is exhausted, I plug in an alarm gadget into the balance ports. I have it set to beep at 3.2v. What could possibly go wrong with that?

Here's what. When an idiot plugs in a new battery, removes the alarms from the old battery, and then plugs them back into the old battery. which by now has recovered enough to be just above the 3.2v beeping level. So now I'm running on a battery, and I have nothing to warn me when it's depleted.

Until the bike motor stopped working. I very quickly determined the cause, swapped to the next battery, and carried on.

When I got home, I use Battery Medic to measure the cell voltages for each of the cells in the batteries. It wasn't as bad as I feared; the voltages were between 3 and 3.2 volts. Saved by the low voltage limit on the controller!

But when I tried to recharge those batteries, the charger reported that there were only three cells (in a 4s battery). It wasn't that a cell was dead, because Battery Medic would have told me. I couldn't work out what to do, so I powered everything off and went to bed.

Next day, I plugged in the charger set to "Charge" (I usually set it to "Balance Charge") and that worked. After several minutes, I did my usual balance charge on those batteries, and it looks like I've gotten away with my blunder.

Now I need to think of a way to idiot-proof my changing of the batteries.

What you REALLY should be doing is hooking up all those 5Ah pairs in parallel to eachother. That way, instead of having 5 X 5Ah you will have one large 25Ah pack.

Why is this better?

1) You don't have the risks of any particular pair being drained down too far (over discharge is BAD for batteries)
2) Say you use about 20Ah out of the 25Ah in one day. That means that each pair will have 1Ah left and they will be happy there. Using your last method, some will have ~0Ah and 1 pair will be "full".
3) All that work of unplugging and replugging is gone and so are the risks (each time you do that, you open the door to human error)
4) Your batteries will stay closer in condition since you aren't overdischarging some packs and leaving other packs full.

 

[drsolly]

What you REALLY should be doing is hooking up all those 5Ah pairs in parallel to eachother. That way, instead of having 5 X 5Ah you will have one large 25Ah pack.

Why is this better?

1) You don't have the risks of any particular pair being drained down too far (over discharge is BAD for batteries)
2) Say you use about 20Ah out of the 25Ah in one day. That means that each pair will have 1Ah left and they will be happy there. Using your last method, some will have ~0Ah and 1 pair will be "full".
3) All that work of unplugging and replugging is gone and so are the risks (each time you do that, you open the door to human error)
4) Your batteries will stay closer in condition since you aren't overdischarging some packs and leaving other packs full.

Four good points. But I'd have to do it slightly differently. Because of what I'm using the bike for. I thnk most people here are commuting, or trail-riding, or shopping. I'm geocaching (and no-one else here has mentioned that they're doing that). Some days when I go out, I don't get back to the car until the end of the day, but on others, I do one loop, get back to the car, then do another loop. While I'm back at the car I can have lunch, drink coffee and change battery. That's one reason why I have seven separate batteries; so when I get back to the car, I can leave the ones I've used, and set out in the afternoon with all fresh batteries. Another reason is I can be flexible about how many I carry, based on how far I'll be going. So what I could do, is buy two more 4s, and make up two batteries of 8s4p. Or make up one 8s4p and one 8s3p.

I can't entirely predict how much battery I'll be using, because although I know the route I'll be taking (apart from when I get lost or go down a track that the map says is there but reality doesn't), I don't know the quality of the track I'll be on, and soft squishy grass/mud uses a lot more battery than tarmac. And gooey clingy mud (as you get when following a bridleway across a ploughed and harrowed field) can make progress more a matter of inching forward despite using a lot of power. Experience says, though, 8s3p is usually enough for a half day, and 8s5p is probably enough for a full day. And that's why I like to take 8s4p for a half day and 8s6p for a full day. Conveniently, 8s6p fits nicely in a large ammo can with room for a bit of foam padding round the sides, and 8s4p fits nicely in the other boxes I found on Ebay.

So why don't I just make up one big 8s6p, which would probably last me all day? Because it's big and heavy. I can just carry 8s6p in one pannier but it's very unbalanced to the side; I'd need to use two panniers, which means I can't use my kickstand (when you're geocaching, you're stopping every few hundred yards). Another advantage of using one battery-pair at a time, is I have a very accurate idea of how much juice I have left. But I think I don't need that, because I now use an amp-hour meter which gives me that info (before, I was relying on the number of blobs being displayed by what was actually a very crude and inaccurate voltmeter).

Another problem - I can't charge 8s, because A) the batteries are 4s and B) Hobbyking don't seem to carry balance leads or parallel balance leads above 6s. and C) I would have to solder up a balance lead that took two 4s as input and one 8s as output, and I'm not too confident about my soldering ability on that. So I'm currently stuck with needing to charge 4s, which means I still have to do the unplugging and replugging. But that's done at home, not out on the road, so I'm less likely to make mistakes.

Also, I've abandoned the 10-amp kettle-type leads, I'm pulling far too many amps for them. And I'm getting disenchanted with 4mm bullets; repeated plugging and unplugging is going to get progressively more difficult as the plastic boots start to wear. I'm now thinking XT60s or XT90s, and I've ordered some other types of connectors to try out. I tried Andersons, and found that I was unable to crimp them, maybe I need to get a special crimping tool. Plus, they're a bit too semi-permanent for me, with all the unpugging and replugging I'm doing.

When those additional connectors arrive, I'll decide which one I like best, and then I'll have to remake all my wiring harnesses to make use of them. And at that time, I'll rethink how I'm doing this, and maybe go for 8s4p.

So I could, for example, do an 8s4p and an 8s3p, and make suitable harnesses for each of those.

More thought is required!

 

[spinningmagnets]

I've read a couple posts in my searches where someone had single bullet connectors, instead of paired and polarised connectors (that cannot be plugged in backwards). One single moment of inattention, or being in a rush, or connecting red/black bullets in dim light...Plug it in wrong and there's a huge spark...and if the pack even works, it is never quite the same...

I almost didn't post this, because I take it as a given that I will always use polarized connectors. My favorite is the XT-90, easy to solder wires onto, and it cannot be accidentally plugged in backwards. There are other good connectors, Andersons that are properly orientated when they are paired, etc...

 

[biohazardman]

I have killed a few a123 26650s cells but it was difficult.
1- I ran the pack to LVC then parked the bike and let it sit for several days with the controller still on.
2- When I found the cells were at .6 to .8V on several occasions I charged my 20s 2p pack at ten amps without balancing.
3- I then immediately went on a hard high discharge ride often hitting 30+ MPH to all the way to near LVC to see dnmun the first time. I had popped a few cells by the time I got there, these cells make a very distinctive pop when they die. I quickly pulled my pack and cut out the bad cells which were getting very warm. I had killed my first four a123 cells after a couple years of use. I have a couple more years on the rest of the cells, in the same pack, now and have done the same sort of thing two more times and lost 3-5 cells each time. dnmun said I should have a BMS, he is probly right but cuz I only balance once or twice a year, but I have resisted the evil part as I seldom have problems and the loss of 10-12 cells in near four years is acceptable to me.

Cheap low-C cells are a different matter. I had a Foxpower 36V 10ah pack on my fist build with a 500w Golden motor and it werqed great. So, I bought a 48V pack and installed a new controller. The improvement in performance was great but the cells began to die quickly as they could not handle the 30A draw. So cheap cells die the death much easier and more quietly than the good stuff although they do puff up and get hot. Not really even any fun although a good way to get rid of some of that excess cash you might have burning holes in your pockets and ruining otherwise perfectly good pants.

 

[dnmun]

they were still hot when we cut them out. they had swelled up and popped the ends of the cylinders out too.

that was bad but i over discharged a ping pack because i left the BMS off 'briefly' while i did something else.

so then i was in the other room and i hear a noise, like something fell off the table so i was gonna go yell at the cat for knocking stuff over and i found that the ping pack swollen up to twice it's length and all the pouches were like balloons twisted in a spiral connected by the tabs on top, and draping over the edge of the table. pushed a big charger over the edge.

i measured a negative -16V on a 7S section of ping pouches. so that is why i always warn people to not run a discharge test without a BMS present.

all that work building that battery. all for naught.

 

[John in CR]

I've killed some of my Konions. They were used batteries to start with that had an expected life of 300-500 cycles in their original design use with power tools. I sorted and combined them into ebike battery packs in 2008 and 2009 and used some of them for 1500-2000 cycles (mostly less than 50% DOD and charged to about 95% SOC). Some of the groups in these packs went dead in the past year. I believe they were killed in the best manner possible, excessive low stress cycles.

At the other end of the spectrum, I've had only one RC Lipo battery death, and it died from being ignored. Upon receipt of 15 of the 4s hardpacks, I checked them all for balance and none were much out of whack. I balanced charged them to 4.1V/cell and set them aside to use later. About a month later one of the packs flamed out without have been touched much less ever used. Luckily I had the packs spaced far enough apart that the adjacent packs didn't go too. I was lucky in that it was the top cell that flamed out and the cell under it only puffed badly, so since the room was well ventilated there was little mess to clean up and no other damage.

 

[999zip999]

Lucky John ! Did you buy a lotto that day ?

 

[Drunkskunk]

Lead acid - discharging at 2C. That was an expensive lesson

Lipo - BMS failure: Mechanical damage. Basicly, it got shaken to death.

Lipo - BMS failure: The board burned.

Lipo - Age. 4 years and several hundred cycles.

Lipo - over discharge. I left it hooked to a controller that was off.

 

[Jonboy]

Lipo- run to low my fault miss calc on lvc killed 2 cells in in 2 6s packs
Lipo- pierced one! not on my bike I'd like to add! Powf .. up it went!

 

[docnjoj]

I used to run my restructured Fatpacks as 15S at 8 ah. Trike was way faster than I need. One day I did the standard "Stupid human trick" and left them on with the controller. Pack went way down to about 1 volt. Started recharge with and old NiCad charger and when the packs were up to about 3 volts the Bosch charger recognized them and charged them the rest of the way. I kept them off the trike for a week, and no cells fell very much. I recharged and went riding with Cellogs attached and they seem to be OK. They are now run in parallel with an old 10 amp Ping to power my wife's trike. Still going! This event took place around 2-3 years ago.
otherDoc