48v 15Ah LiFePO4 problems (insufficient charge)

[Godfreca]

Greetings Friends,
(First time poster, long time reader)

As the post states, I'm having a bit of problem with my 48v 15Ah LiFePO4 Battery. For 100+ cycles over about a year I've had no problems, easily riding the 12.5 miles to work on Wilderness Energy BL36. Recently, the battery was discharged (not by me) and was left for about a week. Ever since, I haven't been able to get more than 3 or so miles out of the pack, leaving me with a good deal of pedaling. When I plug back into my 2Amp charger, it starts balancing after about 1.5 hours (when it usually takes much longer).

I'm not very experienced, as you can likely tell, but I did test the individual cells on the BMS ribbon via the multimeter (they were all well over ~2v, and about even) as well as two different chargers. They were all fine.

Anyone have a suggestion about next steps? Would love some help troubleshooting!

 

[recumbent]

If you only get 3 miles from a 15amp/hr pack, it sounds like you have a serius problem, or just one bad cell tripping the LVC.

When you said that after charging "all cells were well over 2volts" this sounds disturbing, because they should all be 3.6 volts. 2 Volts per cell is critically low for lifepo batteries.

Maybe your charger is broken.

 

[Godfreca]

I have two seperate chargers, and have tested separately, so I'm pretty confident they are working.

With regard to the readings on the individual cells, I'm not sure my multimeter is correct - it displayed 2.5 for a AA battery! The readings were all over 5 on the LiFePO4, so I just estimated the number - perhaps it is just making a 1 digit error instead of a factor of 2 error? (like I said, it's amateur hour over here).

Any other suggestions?

 

[swbluto]

Either get a better multimeter or make a more secure connection (And/or, make sure you're operating it correctly.). Only trust the reading if the test unit for callibration (i.e. AA battery) seems "reasonable" AND steady.

 

[Russell]

I suggest you charge it normally, remove the charger, wait 15-30 minutes and then check and (accurately) record the voltages for each of the 16 cells. All of the cells should be right at the balancing point for your BMS, which for mine is 3.60V, and when all is right with them they will read 3.59V +/- 0.01V 30 minutes after the charge voltage is removed. If you see any cells significantly lower than the rest, and my guess is you will, then it would be best to charge that cell with a single cell charger to try to bring it up to the level of the rest of the cells. Here is an example of one you can use;

http://cgi.ebay.com/3-2V-2A-Smart-Charger-for-Single-Cell-LiFePO4-Battery_W0QQitemZ330320891189QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item4ce8a8cd35

Bulk charging a pack with a significantly low cell or cells in series can be futile as the higher voltage cells will cause the current to taper off long before the low cell(s) have a chance to charge.

If the difference between the cell voltages isn't too bad then plug in the charger again and let it run until it goes green, then once again remove the charger and let the BMS balance the pack. Repeat this sequence as many times as necessary to get all of the cells back to equilibrium. Doing it this way is actually faster than leaving the charger connected as each balancing will knock down the high cells and each charger application will raise up the low cells. Also if your charger isn't at least outputting 59V and it can be adjusted then do raise it to 59-60V which will help speed the process. If the low cells never come up then it's likely they are permanently damaged.

-R

 

[Godfreca]

Thanks for your help!

I followed Russell's suggestion, turns out my multimeter needed a new battery, and here's where I landed:

1 - 3.51
2 - 3.62
3 - 3.83
4 - 3.61
5 - 3.53
6 - 3.54
7 - 3.43
8 - 3.45
9 - 3.59
10 - 3.55
11- 3.69
12 - 3.55
13 - 3.56
14 - 3.55
15 - 3.56
16 - 3.29

They seem very erratic! Should I follow the procedure on charging then balancing or move straight to addressing the single cells?

Also, any guess on how this could have happened? Some of them (particularly cell #3) seem way off!

 

[Russell]

Thanks for your help!

I followed Russell's suggestion, turns out my multimeter needed a new battery, and here's where I landed:

1 - 3.51
2 - 3.62
3 - 3.83
4 - 3.61
5 - 3.53
6 - 3.54
7 - 3.43
8 - 3.45
9 - 3.59
10 - 3.55
11- 3.69
12 - 3.55
13 - 3.56
14 - 3.55
15 - 3.56
16 - 3.29

They seem very erratic! Should I follow the procedure on charging then balancing or move straight to addressing the single cells?

Also, any guess on how this could have happened? Some of them (particularly cell #3) seem way off!

Is that 30 minutes after you removed the charger?

The biggest problem cell is the last one at 3.29V. If you have a single cell charger or a variable power supply that can output 3.6V then by all means try to charge #16, otherwise follow the routine I outlined above and check the voltages after 2 or 3 times. Do make sure to remove the charger each time and wait at least 30 minutes before reapplying the charger or taking measurements. Also as I said above if your charger isn't ouputting at least 59V and it can be adjusted then take it up to 59-60V.

-R

 

[Godfreca]

Thank you Russell - I waited 20 minutes, but I'll wait longer and keep trying.

I don't have a single cell charger, I just ordered the one you suggested, so I'll keep on your suggested path until it arrives.

-christopher-

 

[Russell]

I asked because the BMS (you do have a BMS right?) should have knocked down the high cells in relatively short order so besides having the low cell (#16) problem the high cell (#3) is likely causing the BMS to cut-off the charge current when it reaches around 3.90V (the individual cell upper limit protect level for many BMS boards). If you have a 3V flashlight (edit) BULB you can discharge cell 3 (and 11) with it to bring it in-line with the others which will keep the charger active longer.

-R

After thinking about it I'm going to modify my statement and say the biggest problem right now are the HIGH cells. The reason as I eluded to above is #3 is going to hit the upper protect limit for an individual cell which will cause the BMS to open the charging circuit thus terminating the charge cycle early before the rest of the cells have a chance to get fully charged. Therefore you should first discharge cell 3 and probably cell 11 to the average voltage for the other cells. A 3V (2-cell) flashlight bulb is a great choice or pick up a 10 ohm 10W power resistor from RadioShack...but be very careful not to short leads!!!! Once you get those high cells down the charger will stay on longer and bring all of the cells up to a higher voltage...then do the multiple charge, remove charger, wait 30 minutes, and repeat cycles.

 

[swbluto]

Assuming the 3.29 volt cell has 3 Ah and the whole pack has 15ah, you're either going to be waiting a l0000000ng time with all that discharging, or you're going to discover a low capacity cell. Single cell is the way to be.

 

[Russell]

Assuming the 3.29 volt cell has 3 Ah and the whole pack has 15ah, you're either going to be waiting a l0000000ng time with all that discharging, or you're going to discover a low capacity cell. Single cell is the way to be.

I'm pretty sure that #16 has quite a bit more than 3Ah in it. In any event once he gets the HIGH cells down the charger will stay active longer and charge all the cells to a higher average level, #16 included. Since he has to wait for the single cell charger to arrive it won't hurt doing the multiple charging/balancing/charging cycles.

-R

 

[swbluto]

It might, but a 3 mile range as opposed to what's normal (20 miles?) seems to suggest that it would have a pretty low capacity assuming this charge cycle is comparable to past charge cycles. However, even if that's not the case, a floating voltage of 3.29 volts suggests a charge state of less than 70% and considering the other cells are at least 97%, you're going to have a lot of discharging with ALL of the other cells, and not just the top performers. This assumes the measurements were made after charging and letting the battery pack rest for a little while.

A lot of unnecessary work = not worth it, in my opinion.

 

[Russell]

It'll give him something to do...and I did say to check the voltages after a few cycles so if cell group 16 isn't budging then he can wait.

As it happens cell 16 is the group on my 48V/10Ah battery which died a couple of weeks ago. Mine appeared to be a tab failure on one of the pair of 5Ah cells comprising the last cell group. I ended up removing those two cells. One cell was a lost cause the other took a charge but could not hold more than 3.35V. I now have a healthy 15 cell LiFePO4 pack :|


-R

 

[Godfreca]

Hey guys,

I ran through several cycles as Russell suggested - connecting and disconnecting the charger, and couldn't seem to get any different values.
Should I discharge the battery again? Would that help?

Also, just a quick note on my battery, it is not 16 individual square cells, but instead three tiers of 25 tube-shaped cells. I'm not quite sure how this arrangement works and which ones i should run on the individual cell charger. I'm pretty scared to take this thing apart for fear of permanently damaging it/not being able to return it normal.

Any advice? I'm in Los Angeles, are there battery experts around that might be able to take a look?

 

[Russell]

Based on charging experiments I ran this afternoon I’m going to recant some of what I said earlier . In a nutshell it’s NOT faster to balance a LiFePO4 pack by the method I describe in earlier posts. The method I described is best used if a shallow discharge of the pack follows the balancing.

For everyday charging and balancing when the differences in cell voltages is not great (<0.20V) leaving a charger connected to the pack until the current flow reaches near zero (not simply when the charger turns green) will yield better results. This is however only true IF:

1)The charger’s output voltage is sufficiently high
2)The charger output remains active and high even when it indicates a green “charge completed” condition. A charger which completely turns off or switches to a lower level will not work as well.
3)The pack is not grossly out of balance.

If the pack is grossly out of balance then it should be manually balanced by bringing the cells up to the same point with a single-cell charger. This can be combined with discharge of high cells if necessary (a 10 ohm resistor rated 5W or better works well enough for smaller packs).


My recommendation then while you're waiting for your single-cell charger is to see if your BMS is working properly when charging and balancing. The best way to do this is to monitor voltage and current while charging (a wattmeter is an invaluable tool for this purpose as it does both simultaneously). Also record the individual cell voltages before charging, directly after charging and 30 minutes after the charger is removed.

Since you said you had two chargers use the one that is outputting the highest voltage (59V+ preferrable) AND be sure its output remains high even when the charge completed LED turns green. If you only have a DVM to work with then monitor the battery voltage connection on the BMS during a charge cycle. Be sure if it starts at say 54V that it increases over time to the output level of the charger. After the charger turns green make sure the battery voltage remains high (59V+). If during charging the battery voltage stalls and starts dropping sooner than expected, accompanied by the charger LED abruptly going green, then that's an indication the BMS is cutting off the charger (probably due to a cell going to 3.90V+). To verify this you can switch your meter to read AMPS (move leads to current position...and remember to move them back later!) and place it in-line with the charger and battery. Warning:If you don't know how to measure current then don't because you can blow things up if you get it wrong. If the battery voltage slowly approaches the open circuit voltage of the charger then the LED turns green and the voltage remains high then leave it connected for another hour or two. Again a wattmeter is great here because you can see the current slowly decrease towards zero as the voltage slowly rises. Once the charger current approaches zero then disconnect the charger and let the BMS do the final balancing but don't expect #16 to be up with the rest. If all is working properly the other 15 cell groups should balance however.

The alternative is to wait for the single-cell charger and use it to charge each cell to the same point so you know you're golden. Once that's done you'll want to watch cell #16 closely to be sure it can hold the charge. You'll also want to check the other cell voltages after a normal discharge/charge/balancing to be sure they are in fact staying close to one another.


-R

 

[dogman dan]

If you haven't allready, I'd be looking inside the battery to see if any bms wires are disconnected. Look for torn tabs if it's a pouch battery, or popped spot welds if it's a round cell one. Leaving it on the charger for a very long time could help some, and shouldn't harm it.

When my normal battery just gets a litte out of balance, a few short rides and long charges brings it back. Yours is much worse though, and some of the cells in a paralell string may have died unfortunately.

 

[dnmun]

russell was right, the high cell is shutting off the charge cycle. if you can drain some charge out of the that cell with a power resistor(i use a 1ohm 5W ceramic rctangular type) so the others can catch up, then it should go back into balance pretty fast.

 

[Godfreca]

Thanks again for all the help!

So here's where I am - I got the individual cell charger, but am quite afraid to disassemble my battery which is necessary to charge the cells. Their are 75 cells in a three tier stack of 25 each in a 5x5x3 set.

Still, I'm afraid to take it apart the battery for fear of permanently damaging it. Any advice? I've attached photographs of the battery.

Does anyone know where I could send it for some expert help?

 

[Godfreca]

Does anyone know of a place or person in Los Angeles that might be able to help me out?

 

[dnmun]

you don't have to take the pack apart to charge the individual cells. just put the alligator clips on each end while it is still in the pack.

we told you that you have one high cell that shuts off the charge, and how to discharge it so the others can catch up.

what is the big black burn mark on the heatsink in that picture?

you do not have 5x5x3. that is 16S pack so it will be multiple of 4 or 16 so i would say it has 3 cells in parallel for the 12Ah and 16 in series. so expect 48 cells total. or 64 if they are 3Ah wide.

 

[Godfreca]

I'm not sure where the black mark on the heatsink came from - it looks like it was burned in some way.

When you say: "Just put the alligator clips on each end", do you mean that i should put them on the end of each column of cells or on the individual cells? I assume you mean the latter. It seems that I would have to take pack apart to accomplish this - there is no other way to access them.

Thanks for checking in - I haven't gotten the ceramic resistor yet, so I haven't discharged that cell. Thus far I have been discharging the battery on the bike and plugging it back in as was suggested in a previous post. I'm about to sit down with my multimeter now.

Will send more information when it becomes available. Also, unless there is a filler in the middle of the second tier, there are 75 cells as I described previously.

 

[Godfreca]

Greetings followers!

Thank you again for all of the help.

Dogman was (as usual) correct - there was a frayed line within the pack. I resoldered the wiring (which was connected to the high cell) and since the cells have balanced (3.58-3.70) with the exception of the the one very low reading of 3.24v at the far end of the BMS.

I took the battery out for a quick ride and had the same issue - about 3 miles then cut-out.

I assume the next step is to try to individually charge the low cell to see if I can get it up. To do this I assume I should put the positive end of the cell charger to the line running into the BMS, but what about the discharge?

A little guidance would be helpful with regard to individual cell charging. Please help if you've done this before!

 

[dnmun]

you can't charge a cell through the sense wire. it will burn up.

 

[Godfreca]

I figured that all out. After taking the wrapping off my battery I am now charging directly on the cell.

Thanks!