why would a LiFePO4 cell go bad?

13 posts / 0 new
Last post
queensland
queensland's picture
Offline
Last seen: 8 years 3 months ago
Joined: Thursday, May 29, 2008 - 16:33
Points: 55
why would a LiFePO4 cell go bad?

Hi all,

In more "general consumption" terms (because I am not an electric geek), could anyone explain why a specific cell in a LiFePO4 battery pack would go bad? I have two 1.5 year old packs made up of 8 cells each. They have probably been charged less than 30 times over the past year, properly charged after each use, proper temps in winter storage, etc. One pack seems fine but the other wasn't charging so we did the usual cell check. Each cell is supposed to be 3.2v. Three are basically dead. The other cells are OK. I know I need to replace the bad cells, but was curious why this might have happened.

Thanks for your help!

colin9876
Offline
Last seen: 11 years 6 months ago
Joined: Thursday, October 23, 2008 - 11:37
Points: 289
Re: why would a LiFePO4 cell go bad?

Get a single cell charger and Im sure those dead cells can be restored - if they are measuring over 1v you can definately get them back
My bet is on the BMS being the problem.

Single 3.7v LiFePO4 chargers are about £10 on ebay

queensland
queensland's picture
Offline
Last seen: 8 years 3 months ago
Joined: Thursday, May 29, 2008 - 16:33
Points: 55
Re: why would a LiFePO4 cell go bad?

The "bad" cells are measuring below 0.01v. The "good" cells are 3.2v.

"Tis better to remain silent and be thought a fool than to speak and remove all doubt." Abraham Lincoln

colin9876
Offline
Last seen: 11 years 6 months ago
Joined: Thursday, October 23, 2008 - 11:37
Points: 289
Re: why would a LiFePO4 cell go bad?

Theyve been shorted then if they r that low!
Does your battery have a BMS unit on there?
How many amp hours are the cells?

reikiman
reikiman's picture
Offline
Last seen: 11 months 3 weeks ago
Joined: Sunday, November 19, 2006 - 17:52
Points: 8447
Re: why would a LiFePO4 cell go bad?

To answer the original question - it's my understanding that LiFePO4 cells will, if they go above the high voltage mark, or below the low voltage mark, undergo a chemical change and be unable to do battery behavior.

Additional questions:

Whose pack? Did the pack come with the BMS? (if any)

queensland
queensland's picture
Offline
Last seen: 8 years 3 months ago
Joined: Thursday, May 29, 2008 - 16:33
Points: 55
Re: why would a LiFePO4 cell go bad?

I got the packs from FalconEV late in 2008. Used them in the Lepton all summer of 2009. Scooter ran perfectly. Removed them for winter, stored indoors at about 80% charge. I re-installed them myself in May 2010. Scooter ran correctly through 3 cycles, then stopped charging. As far as I could tell I didn't mess any of the connections up. We checked the packs at that point. One 8 cell pack shows normal 3.2v charges. The second 8 cell pack showed at least 2 reading below 0.01v. Two other cells read over 6v, but possibly we were detecting 2 cells together? No BMS, but using a "smart" LiFePO4 external charger from FalconEV.

A co-worker has been doing the diagnostic stuff as he works on my office's GEM cars. I am trying to learn the basics of batteries but have difficulty understanding it. I will certainly pass on any comments or suggestions to him. Maybe we do need to add a separate BMS in addition to using a smart charger? I am sorry to be so clueless, but am slowly getting it!

"Tis better to remain silent and be thought a fool than to speak and remove all doubt." Abraham Lincoln

Reid250
Offline
Last seen: 13 years 4 months ago
Joined: Wednesday, December 9, 2009 - 17:15
Points: 107
Re: why would a LiFePO4 cell go bad?

My experience now tells me that LiFePO4 cells must have a BMS if they are to survive. After the initial balance, my battery of 20 X 40Ah TS cells was fine for the first 10+ cycles and then, the charger sensed a problem and shut down. On checking the cells, I found 3 at 2V and the other 17 @3.45V I used a single cell charger and brought the 3 up to 3.75V. Two more cycles and the 3 are back at 2V. So now the expert, mountain chen, who sold this crap in the first place, advises that you need a BMS. His new scoots all have a BMS. The alternative is to purchase 3 new cells, but I believe we will always have a problem with LiFePO4 cells unless we have a BMS. This is the nature of these cells. The chemistry varies from cell to cell. Maybe we will get there one day with a "plug and play" EV but cheap Chinese Junk is not the way to go. Buying crap on eBay or from other hidden sources is just plain stupid. No support or warranty was dumb on my part.
I'm going back to the old fashioned model of doing business. Over the counter from a guy who has been there for a few years.

PJD
PJD's picture
Offline
Last seen: 1 day 17 hours ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: why would a LiFePO4 cell go bad?

Reid250,

You definitely have 3 bad cells, and based on their failure after only 10 cycles, they were probably bad from the factory, not due to lack of a BMS for charge balancing - but charge balancing IS important.

New cells and cell balancers can be had at the lowest price from elite power solutions. I've had good experiences so far doing business with them. The cell balancers will only provide charge balancing, not over-discharge protection, but, if the pack is reasonably healthy, and you ride with a pack voltmeter and keep the voltage above 2.5 volts per cell x #cells, you will probably be all right.

colin9876
Offline
Last seen: 11 years 6 months ago
Joined: Thursday, October 23, 2008 - 11:37
Points: 289
Re: why would a LiFePO4 cell go bad?

Reid,
Nooo that smart charger is only smart in the sense that it will go through different phases as it reaches the max 30v in charging.

Its not at all smart in sensing where the cells are at - as its only measuring the total
So its banging 6v into some of the cells to reach the total as some cells are holding none. That will end up damaging those too.

I really dont like BMS technology, its fiddly. But its the only way if ur doing series charging.

p.s. The real answer is parallel charging, all cells charge in parallel off a 3.7v source - perfect, they balance automatically. Only issue is u need a switch or harness change when they need to be used in series. Not many people have caught onto this yet, but it is the real answer

Reid250
Offline
Last seen: 13 years 4 months ago
Joined: Wednesday, December 9, 2009 - 17:15
Points: 107
Re: why would a LiFePO4 cell go bad?

I must disagree. When the cells are all balanced, the charger completes and shuts off and the battery measures 75V. If the charger stops and the red light is flashing, I will find there are the 3 cells that are low in the battery. The others 17 are at 3.85V. 17 X 3.85 = 65.45V 3 x 2.7V = 9.75 for a total of 75.2V. The difference may even be the ambient air temp. I understand that it is a dumb charger and only knows to shut off when it hits 75V, but please explain the flashing red light? It will also beep if I try to recycle it when the 3 cells are low. It does not beep when the cells are balanced. It shuts off and shows 2 green lights. I guess I will dig my Tektronix Dual Channel scope out and see what is what. Obviously, there is something causing the difference between a green and a flashing red indicator.
I have a bunch of 5V PS from cash registers so I'm looking for some 2amp 3.85 voltage regulators. I haven't found them yet. Maybe I will have to build an "H" pad on 4V regulators?

Reid250
Offline
Last seen: 13 years 4 months ago
Joined: Wednesday, December 9, 2009 - 17:15
Points: 107
Re: why would a LiFePO4 cell go bad?

Next trip to the BIG CITY I will buy 3 cells. When I can buy a good BMS from someone face to face, I will do that. No more "pig in a sack" for me.
Thanks for the reference to elite power. I know he builds good stuff. His original plan was to sell boards and allow assembly, but once all the collaboration was finished he pops up with a finished product. I wonder why they switched cells from Thunder Sky to GBS?

If/when mountain chen or one of the other orientals finally have a north american presence, the TS BMS will be $100 for 20 cells, as it is now in China. It is the $150 shipping plus you are screwed if you have to replace even one BMS bar. The cost of time and agony does not compute for me.

antiscab
Offline
Last seen: 9 months 3 weeks ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: why would a LiFePO4 cell go bad?

My experience now tells me that LiFePO4 cells must have a BMS if they are to survive. After the initial balance, my battery of 20 X 40Ah TS cells was fine for the first 10+ cycles and then, the charger sensed a problem and shut down. On checking the cells, I found 3 at 2V and the other 17 @3.45V I used a single cell charger and brought the 3 up to 3.75V. Two more cycles and the 3 are back at 2V.

Hi Reid250,

interesting, were you using a cell level voltage measuring device (like a paktrakr) at the time?
Or did you just go around with a multimeter after the fact and notice the 3 cells at low voltage?

I had an interesting experience back on my Emax when I installed a paktrakr.
it got the power to run itself off the first 3 cells and pulled them down relative to the rest of the pack.
my BMS compensated by running the shunts for longer, but i did find it surprising at the time.

the flashing red light on the charger could be because it is a little smarter than the average charger:
a lithium pack with all good cells will have a voltage that rises comparatively slowly once the total pack voltage nears 75v while still in current limit.
a lithium pack with a few cells not quite charged while the other are will have a pack voltage that rises quite quickly when the pack voltage nears 75v.

a charger that looks for this can see the pack is out of balance, and shut down, indicating the user needs to take action.
this is what I think has happened.

back to the original poster:
cells can go bad in pack when the BMS fails.
for instance, if the shunt in a BMS fails on, the cells could be drained to near 0.

if there is no BMS, cells can go dead when reversal happens (cell voltage goes below 0v).
cells will have reduced capacity when overcharged past 4.2v and possibly overdischarge (I haven't got around to measuring capacity loss due to overdischarge).

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

Reid250
Offline
Last seen: 13 years 4 months ago
Joined: Wednesday, December 9, 2009 - 17:15
Points: 107
Re: why would a LiFePO4 cell go bad?
My experience now tells me that LiFePO4 cells must have a BMS if they are to survive. After the initial balance, my battery of 20 X 40Ah TS cells was fine for the first 10+ cycles and then, the charger sensed a problem and shut down. On checking the cells, I found 3 at 2V and the other 17 @3.45V I used a single cell charger and brought the 3 up to 3.75V. Two more cycles and the 3 are back at 2V.

Hi Reid250,

interesting, were you using a cell level voltage measuring device (like a paktrakr) at the time?
Or did you just go around with a multimeter after the fact and notice the 3 cells at low voltage?

I had an interesting experience back on my Emax when I installed a paktrakr.
it got the power to run itself off the first 3 cells and pulled them down relative to the rest of the pack.
my BMS compensated by running the shunts for longer, but i did find it surprising at the time.

the flashing red light on the charger could be because it is a little smarter than the average charger:
a lithium pack with all good cells will have a voltage that rises comparatively slowly once the total pack voltage nears 75v while still in current limit.
a lithium pack with a few cells not quite charged while the other are will have a pack voltage that rises quite quickly when the pack voltage nears 75v.

a charger that looks for this can see the pack is out of balance, and shut down, indicating the user needs to take action.
this is what I think has happened.

back to the original poster:
cells can go bad in pack when the BMS fails.
for instance, if the shunt in a BMS fails on, the cells could be drained to near 0.

if there is no BMS, cells can go dead when reversal happens (cell voltage goes below 0v).
cells will have reduced capacity when overcharged past 4.2v and possibly overdischarge (I haven't got around to measuring capacity loss due to overdischarge).

Matt

I am using a multimeter for check cell voltage and some voltage selectable wallwarts for individual charging.

I think the answer to "why do individual cells go bad" is because of poor quality control on the mix of material in the cells. On checking the serial numbers of my cell, I see they are all labeled consecutively. This does not prove they were all from the same batch run, but maybe. In conversation with two different supplies of TS cells, they were both arguing that the cells I have, are not theirs because of the serial numbers. I was able to determine which of these two suppliers had in fact labeled these cells, so I think they are shipped from the manufacturer without a serial number on them. This then presents a further puzzle in determining the real date of manufacture of any TS cells. They could be sitting in a warehouse for years before they are labeled and sold to a scooter manufacturer, who then takes months to assemble and ship the scooter. My scooter was manufactured 15/11/2008 and not sold until @ 1/10/2009. Some or all of my cells may be from 2007, when we now know from experience, there were quality control problems.
It is our own fault for buying oriental junk, under their terms of no support. We are not prepared to pay for stuff made by our $50 per hour unionized industrial workers. Until EV goes mainstream, we are just tinkering with a very expensive, frustrating hobby. The ultimate in our stupidity is being fed by http://www.giobikes.com/Auction/?pcid=6 Buy a totally useless POS for $250, from a blind hole in cyberspace, with a huge "transaction fee". They do not allow pickup, so you are forced to pay 5X the real shipping cost (while they pocket 80% of it) if you live one block away. The subsequent "shipping cost" of any "free" replacement part is also several times the actual cost, so you have the pleasure of being screwed twice if something was broken, during assembly, shipping or use. Until we reject this "business model" we will continue to be victimized, by our own foolishness.
Cumoco is a tiny pretend step in the right direction, assembling the same oriental components of questionable quality (including Thunder Sky cells) in North America. I hope I'm wrong and it is not just another route to the public trough through venture capital. Off loading your risk, while being handsomely rewarded, is a great path in an uncertain market. Watch how soon the venture capitalists (if their any) take it public and bailout?

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • xovacharging
  • stuuno
  • marce002
  • Heiwarsot
  • headsupcorporation

Support V is for Voltage