ruined my pack :(

[andrew.box]

Not sure how, but somehow I did something very dumb and left my pack connected to the controller and it was drained down to 2.5V (for the whole 18S pack). Normally I would always turn off the main DC out switch and often also disconnect the pack too, just to be safe. I must have left the DC switch on after messing with it and a slow drain from the controller pulled on it for ~2 weeks. Really not sure how this happened, I would normally NEVER leave the switch turned on when the bike isn't in use or charging so I must have gotten distracted and ran off with it in the ON position and ended up leaving it that way. So much for idiot-proof.

I tried the slow charge using NiMH mode which I've used to revive over-discharged packs in the past but no dice. They're not coming back. Just wanted to share this unfortunately experience with folks who will understand the stupid, frustrated, sick feeling this gives me. A $250 battery down the drain after fewer than 20 cycles. Well, it could be worse. No fire and at least they were cheap lipo and not high-dollar A123 cells or something.

I measured 30K DC resistance looking into the motor controller. That's less than 10 mA at 72V which should really not have drained them dry in less than several weeks (10AH pack that I left ~80% charged), but I don't see what else it could have been.

 

[Floyd1]

Can you split the pack, find the crook cells and salvage something? Good luck

 

[dnmun]

most controllers use about 55-60mA. your BMS shoulda cut it off when it reached the LVC of the cell that was lowest. maybe the output mosfets had shorted.

 

[andrew.box]

I don't have a BMS other than the cycle analyst LVC and some cell-log style monitors I use for checking cell balance. So there's nothing to cut it off except using my own good sense to turn the pack off when I'm done with it. I wish there was a better solution for a high power 16-20S lipo BMS, but such a thing is necessarily complicated (see expensive).

I might split the packs and test a few single cells, not holding out much hope though. They seem to be in pretty bad shape.

 

[amberwolf]

Well, it wouldn't be very complicated, but it isn't teh cheapest:
http://www.endless-sphere.com/forums/viewtopic.php?f=31&t=36414
Might not be worth it for a small pack, but a larger one might be.

It's not really a perfect solution because of it's LVC point, but it could stop drainign the cells to zero.

 

[andrew.box]

Ya, I've been eying the Methtek LVC solution for a while now. Might be good piece of mind to just go ahead and bite the bullet and do it once I get around to buying new cells to build a new battery. I've just proved that it's good insurance to have some kind of fail safe in place. If it saves a battery it will pay for itself in one use.

I can handle balancing and maintaining the pack myself, but having a reliable LVC to cut off the pack in the event of an unforseen drain condition would probably be wise.

 

[SamTexas]

A simple solution that might work for you: Install a bright LED that gets turned on when the controller is on. The LED should be positioned so that it can be seen easily when you are off the bike but out of sight when you're riding. Bright red tail light?

 

[Ykick]

Sorry to hear 'bout this. It just dawned on me but I've yet to use a switch, pre-charge, etc. 'simply disco the battery/controller Anderson's and directly connect to charger for charge. Leaving any type of active and particularly parasitic electronics connected to RC Lipo is risky IMO. Cruder the better if you wish to avoid the complications of proper BMS design and execution.

Personally, I'm just a lazy ass and go for the cheap and simple. Direct connecting can apparently offer some unforseen benefits. Hmm?

Sorry to hear this happened to you. You're not the 1st.

Does CellLog show any individual cell levels? Are they pretty even 2.5V/cell? You should qualify individual cells at this point to learn if you have any salvage opportunities.

A moment of silence please.......

 

[el_walto]

Ya, I've been eying the Methtek LVC solution for a while now. Might be good piece of mind to just go ahead and bite the bullet and do it once I get around to buying new cells to build a new battery. I've just proved that it's good insurance to have some kind of fail safe in place. If it saves a battery it will pay for itself in one use.

I can handle balancing and maintaining the pack myself, but having a reliable LVC to cut off the pack in the event of an unforseen drain condition would probably be wise.

I thought the methtek LVC solution would only cut off the throttle when a cell voltage got too low? If this is true, leaving on an auxiliary device, or the controller would still would destroy the battery pack.

 

[pkirkll]

Sad day for batteries,,,, i posted how i killed mine too....

condolences....

 

[liveforphysics]

I measured 30K DC resistance looking into the motor controller. That's less than 10 mA at 72V

At full voltage (rather than the ~3v your DMM is checking with) it could be 10x lower resistance than your reading. 50-100mA idle current is typical.

 

[Arlo1]

Methods wants me to test one of his HVC LVC boards under load and try to blow it on the dyno using it for LVC to cut the power to the controller under full power when one cell goes to low. I will try to find time to get this set up soon.

 

[amberwolf]

I thought the methtek LVC solution would only cut off the throttle when a cell voltage got too low? If this is true, leaving on an auxiliary device, or the controller would still would destroy the battery pack.

You can set it up so the HVC unit that is meant for cutting off the charger can also or instaed cut off the controller power at LVC.

 

[andrew.box]

Sorry to hear 'bout this. It just dawned on me but I've yet to use a switch, pre-charge, etc. 'simply disco the battery/controller Anderson's and directly connect to charger for charge. Leaving any type of active and particularly parasitic electronics connected to RC Lipo is risky IMO. Cruder the better if you wish to avoid the complications of proper BMS design and execution.

Personally, I'm just a lazy ass and go for the cheap and simple. Direct connecting can apparently offer some unforseen benefits. Hmm?

Sorry to hear this happened to you. You're not the 1st.

Does CellLog show any individual cell levels? Are they pretty even 2.5V/cell? You should qualify individual cells at this point to learn if you have any salvage opportunities.

A moment of silence please.......

I wish they were 2.5 per cell, it was 2.5 for the whole pack. I've decided I'll definitely be spending the money for the methods lvc boards and relay when I rebuild. Might need to wait until after the new year. Holiday expenses coming up and I just purchased a HF ham radio, wife might not be too keen if I drop another 5 bills on toys so soon. She was surprisingly empathetic about the busted battery, she knows how much I babied that thing - until I destroyed it.

 

[Ykick]

Um yeah, 2.5V for the whole pack is pretty much toast IMO. 'sorry if I didn't catch that before.

18S is a pretty good spark for crude connecting as I do with 12-15S. Although, same Andersons that have been in daily for several years and no signs of impending failure.

It's s tough deal to remember to just turn the shit off. Much easier to remember when it's part of the charging procedure.

Hell, even commercially available Lifepo4 packs can drain to near 0V from BMS if not charged once per month.

You're not the 1st and won't be the last. Gotta learn and move on. Easy for me to say right now but relatively cheap tuition in the grand scheme of things.

 

[dnmun]

that is not true. the battery will not drain to 0V if it has a functional BMS. the BMS will turn off the pack when it drops below the LVC on any cell. maybe kfong's LVC device woulda saved him on this one, but he did not have BMS to protect it from this situation which is actually very common.

 

[Ypedal]

I'd try and salvage it.. :lol: .. outside.. wearing some safety gear and hose at the ready.. with a sacrificial charger.. not much to loose other than some skin.. have video camera ready !

 

[neptronix]

Hell, even commercially available Lifepo4 packs can drain to near 0V from BMS if not charged once per month.

Yup.. seen it happen here many times. Or even better.. someone puts the commercially available pack away for the winter, and the BMS has completely drained the life out of the first 1-4 cells.

Another scenario is that the BMS actually does cut power to the controller at the LVC, but considering the LVC is usually set where the pack has 5%-10% capacity left, the BMS, even if it does drain all the cells evenly, can discharge the pack down below 0%.

BMSes just add to the list of stuff that can break.

 

[nicobie]

BMSes just add to the list of stuff that can break.

That's true. Having a BMS has ruined more stuff for me than not having one.

I agree with LFP and AmberW that using one of Method's excellent LVM/HVM relays is the very best way to go.