new LiFePO4 battery has started dying at 50% capacity

[Mark Stosberg]

Hello,

I have a question about the battery performance I'm seeing with my electric Yuba Mundo. The summary is that the battery has started dying at 5 amp hours, when it is rated as a 10 amp hour battery, and originally lasted about 9 amp hours.

The battery is LiFePO4, 36 volts, from Cycle9.com. We have it attached a 35 amp controller by Infineon, a Cycle Analyst and a 500 Watt direct-drive Nine Continent motor in the front hub.

Originally we had a 25 amp controller and the battery lasted up to 9 amp hours (which we were happy with).

After just about 3 months, it seems that "suddenly" the max life of the battery is only 5 amp hours. There are a few factors which I think might be related:

• Our 25 amp controller, also from Infineon, was related with a 35 amp controller. (the button on the original quickly wore out)
• I accidentally fully reset the Cycle Analyst 2.2. Maybe it had some pre-set value in it that made a difference. However, I reviewed the setting and didn't see anything which seem like it would be at fault.
• When we charge the battery, we leave it plugged into the controller. (It's just more convenient that unplugging and re-plugging it) We have not otherwise found any adverse effects to doing this.

What happens how is that at about 5 amp hours, the system shows the same behavior as it did when it reached the 9 amp hour mark originally-- the assist slows down, and then eventually the electric system cuts out and won't power on until it's charged.

Under normal use, we only use 3 amp hours or less and then charge it again. We wouldn't otherwise run the battery all the way down, but we haven't been expecting it to quite after only 5 amp hours!

The LiFePo4 battery was relatively expensive, so I'd like to make sure it's performing optimally.

Thanks for any suggestions about what might be happening!

 

[Jeremy Harris]

My guess is that this battery pack has just been asked to do too much. Generally, most budget LiFePO4 cells don't like being discharged at more than about 1.5 to 2C; which equates to 15 to 20 amps for your battery pack. There are high current LiFePO4 cells around, but they are expensive and I am near-certain that your pack won't have them.

If you want to run at 35 amps from a small 10Ah pack, then you need cells that are rated for at least 3.5C discharge.

I'm afraid that you may have permanently damaged the battery pack by running it at too high a current, but there is a slim chance that all that has happened is that the battery pack has become unbalanced. To check, you'll need to pull the pack apart enough to be able to get a meter on to the individual cells. If you can do this and report back with the voltage across each cell in the pack, then we may be able to give some better focussed advice.

Jeremy

 

[dogman dan]

My guess is the same. 35 amp controllers and 10 ah packs of 2c discharge cells are going to die before they should. I don't even recomend 10 ah with any controller above 15 amps.

The pack is made of pairs of 5 ah cells most likely, and one of them has died in one or more of the pairs. So now you have a 5 ah battery, that is getting discharged at 5c now. So it's not going to last much longer even if it does have better 3c cells in it.

For that 35 amp controller, you need at least 20 ah of 2c cells, or better stuff, like headways, A123's, or the high c rate lipo.

Bummer.

 

[texaspyro]

If your pack does not have a BMS (or has a bad one) the cells could be out of balance. The higher voltage cells will prevent the lower ones from charging. Check the individual cell voltages. You may need to charge the cells individually with a single cell charger.

Also, bulk charging an out of balance pack can easily result in cell damage.

 

[magudaman]

+1 on possible mis-balanced pack need a good balancing.

 

[dnmun]

this guy disappeared too. maybe eventually they will start teaching this stuff in schools since the kids growing up now will be the ones who have to live with battery power in a few decades when the cheap oil runs out

 

[Kiwi]

Great photos.
I also would agree on pushing the batteries too hard. Looking at the loads you are pulling, chances are you are pulling constantly high amps often.

 

[Mark Stosberg]

Thanks to everyone for the replies! I sense there's a great, active community here.

I was in touch with my vendor and she agreed with the assessment here. A bit more detail: I think the battery pack is supposed to be one of the nicer quality ones, as advertised here:

http://www.cycle9.com/c9store/batteries-c-7/36v-10ah-lifepo4-batterycharger-p-129

The CycleAnalyst has a "max amp" setting that I was unaware of, which got set to "50 amps" when I accidentally did a full-reset of it. I'm unaware of what it was set to before.

Cycle 9 recommended changing this setting to 10 amps and then doing some "gentle cycles" with short trips and see if that helps with the battery life. If it does, she recommended I could slowly turn the max amps up to 25 amps. Otherwise, it seems the battery is considered broken, and I should keep the amp setting low permanently.

I don't see how I can test the individual cells without destroying the current plastic enclosure and creating my own make-shift one to replace it. (But I'm curious know how the individual cells are doing.)

I'm considering upgrading to a 36V/15AH LiFePo4 battery if I need to replace this. Cycle 9 says that's their current recommendation for batteries being paired with the Yuba Mundo cargo bike.

I will also try to take better note of the various stats that the Cycle Analyst is providing the next time there's a problem.

 

[jondoh]

Mark,

If you do set the current limit to 10 amps, be prepared for a huge change in performance. It will accelerate quite slowly. I haven't had experience with 10amps but I've ridden the magic pie motor which has been set by the factory to 15 amps and acceleration was fairly tepid. I'd think 10 amps of acceleration would be quite leisurely.

If the problem is a balance issue, your pack can be brought back. Pay attention to the charging of your pack. If your cells are way out of balance, and your bms doesn't shunt the excess current, your surviving cells may suffer uncomfortably high voltages. When LiFePO4 cells overcharge (past 4.2v or so) they heat up. If you have dead cells, you will want to pay attention to your pack and again watch for heating. My experience with LiFePO4 is that dead cells become short circuits. If charging a pack that's effectively 33v with a 36v charger, you run the risk of overcharging the surviving cells.

I think purchasing a 15ah (3c) pack is a good idea. Don't forget to set the CA to 25 amps and you should be fine.

 

[dnmun]

i am not sure what the assessment you are referring to is. i don't think you have measured the cell voltages yet so it is hard to even guess yet. but it does sound like you have a low cell with diminished capacity.

do you have pictures of the battery and the BMS?

if you measure the cells from the sense wire plug when the battery has discharged, the bad cell should be obvious.

 

[Mark Stosberg]

@jondoh: The first time this happened, I did notice that the battery was warm to the touch. I haven't really noticed it otherwise, but then, it's usually been functioning fine and we don't think to touch it. Thanks the feedback about 15AH option.

@dnmun: You are right that there was been no formal assessment yet. Everyone is just guessing. You can see a close-up photo of the battery here:

http://www.cycle9.com/c9store/images/attributes/Lifepo.jpg

If I don't return it, I will consider cutting open the plastic to access the individual cells. Right now I can't access them or the BMS.

Thanks again for the feedback. I'll weigh my options and report back when there's progress.

 

[dogman dan]

I'll assume that you blew any warantee when you put a 30 amp battery on a 35 amp controller. Bummer the CA setting didn't save you, that should have worked fine. So you most likely have little to lose by carefully unwraping the battery so you can start measuring the voltages, and try to bring it back using a phone charger or other 3-4v wall wart to try to bring the low cells back up.

 

[dnmun]

if you can get to the sense wire plug on the BMS then you can measure all the cell voltages there. but measure them when the pack is exhausted to the 50% level you mentioned originally. write them down. post up here.

take a picture of YOUR pack when you get the BMS exposed and post up here on your thread. since the battery worked when you got it, there is no way to send it back as defective.

 

[Anonymous]

Mark,
That's a very cool rig and great photos. Glad to see yet another i.c.e. sidelined for a ride that's lots more fun, healthier and earth-friendly.
I'll jump right in. Re:

I will also try to take better note of the various stats that the Cycle Analyst is providing the next time there's a problem.

Not.
The CA is there to keep you out of trouble.

Go to http://ebikes.ca/drainbrain.shtml section 7 and download your user's manual and read it top to bottom. Then print it out and highlight the important parts. The cycle-analyst is a remarkable tool and will keep you and your batteries happy. When you program in your max amps limit (just one amazing feature) do so with the specs. for that battery in mind: "It is designated for maximum discharge rate of 3C, which means 30A for a 10 amp hour battery like this. We've found that it can output more current than that, but we don't recommend it for long life... " - from http://www.cycle9.com/c9store/batteries-c-7/36v-10ah-lifepo4-batterycharger-p-129
I'd suggest keeping an eye on the Amp draw on the CA when riding.

And it's cool to track your electron usage with the Wh/mi at the end of each ride from which you can crunch your miles per gallon equivalent (mpge). 1 watt-hour = 0.000 027 322 404 372 gallon [U.S.] of automotive gasoline
source: http://www.onlineconversion.com/energy.htm

I wouldn't continue to use the pack as it is even at lower amps, since an imbalance can harm other good cells.

After the store tells you what's returnable and what's not you have little to loose opening your pack. Duct tape and ingenuity will create another battery box. So when the pack's tired open the pack (with care and with no rings or other conductive things in the area) number each cell and the put a digital multimeter to each in turn and record the voltage on each cell. That's first base. Then charge the pack and record voltages again. That's second base. Report back. The bright lights here will sort it out.

You'll probably see an imbalance and you might be able to fix it with a single cell charger (bringing up the low cell(s) to 3.65V. More V is harmful. With your multimeter you can monitor while charging. I just bought a purpose built single cell charger for $20. Others on this site say it's the best battery tool they have. You can make a single cell charger with what I call a "brick" and what others (I just learned today) call a "wall wart" (guess it looks like a wart on the wall when plugged in) from the bin at the thrift store - full of power supplies to portable phones and such. It should have an amp output close to that of your pack charger and a V output of 3.65 to 4.

Get all the cells as close to 3.65 as possible. Ride it. Repeat.

If the problem is damaged cell(s) (not balance-able after several balance attempts) perhaps they could be replaced.
Otherwise, the remaining good cell can be backup for your new 15Ah pack. Just keep them charged and away from your cute kid.

If your cells are revive-able or replaceable you could consider getting another 10ah pack and wiring them in parallel for 20(ish) Ah.

Good luck.

 

[Mark Stosberg]

Thanks to everyone for the incredible help and advice. I'm blown away by the community here!

To bring some closure to this: I did work out reasonable return terms with Cycle 9, so I won't be opening the battery to investigate this further. At the same time, I'll be upgrading to a 15ah battery that will be better paired with my 35 amp controller.

I have also ordered the ebikes.ca LED lights for the front and rear to try out.

That new stuff should arrive around Tuesday

( And for that matter, I upgraded the original tires on the Mundo to 2.35" Big Apples as well. My first impression of them is that they feel both fast and cushy, as I had hoped. )

I look forward to learning about watts, amps and volts on my e-bike adventure. I think owning a multi-meter may be in my future.

 

[Ypedal]

A multi meter is a must for any ebiker !!.. an ok one can be had for 10$, a few more dollars gets you more accuracy but at least get one for general diagnostics.. ( even if you have no clue how to use it.. hehe.. we can help ! )

The CA you have, does it plug into the controller with a 6 pin connector ? or does it plug in line with the 2 power wires ? Both models have all the settings in the menus but only the 6pin model can enforce these settings.. ( It has to bypass your throttle signal to do so.. and the 2 wire model canot do this )

Your controller is a 35 amp unit, even if the CA is set to 50 amps max the controller will not allow more than 35. You can Decrease the controller limit, but you canot increase it with the CA.

 

[Mark Stosberg]

I confirmed I have the 6 pin CA adapter.