48v10Ah battery problems

[KimABQ]

I have a new MP Edge with 48v10Ah LiFePO4 battery that I purchased from Gary in mid December last year. It has less than one hour total on the motor / battery due to installation taking a long time plus illness and weather. Since completing the install when I ride it only travels about 1/2 mile before it quits. It will start again randomly but only for a short distance and then it quits working again.
The throttle has all battery lights on and they stay on. I have disconnected all wiring except throttle and PAS and still the same problem. When system quits I measure battery voltage at output plug and it is typical ~53v, lowest was 52.2v. All wiring looks good.
As per Gary's instructions I have pulled the wheel and removed the controller to test it with power on. The LED gives one flash when power is turned on but there are no other flashes until power is removed. Then it gave two flashes consistently, but now only one flash when powering down.
When charging the battery it cuts off at 54.8vdc and has never been near 58.4v, which is supposed to be the output of the charger and I am guessing the proper charge voltage.
Gary had me disassemble my battery pack and check individul cells. Most cells were at 3.34+/-1. However, there was one cell at 3.66v so I put a small LED load across that cell and when it dropped to 3.62v the charger came back on and ran for a few seconds before shutting off again. During that few seconds three cells climbed above 3.66v and the charger quit when the first cell reached 3.85v. I then discharged all cells above 3.62v and when the last cell was drawn down to 3.62v the charger starts again and the cycle repeats - three cells above 3.62v, one of them reaching 3.85v and charger shutting down. The charger runs less than 5 seconds during each cycle.
The other cells gradually increase their voltage but after several hours of doing this cycling process the lowest cell was at 3.39v and the highest cells before discharging were at 3.76, 3.72, and 3.71 and two at 3.68v. The remaining cells ranged from 3.39 to 3.53v. Total battery pack voltage was 56.51v.
I am stumped at was is causing the problem. The BMS does not seem to be balancing the pack and when it is left on charge for hours without me discharging any cells. The charger light stays green and no low level cells seem to increase in voltage nor do any higher voltage cells drop. Everything seems to remain static. I do not want to disassemble the pack to charge discharge each of the cells with a walwort, but I do not know what else to do at this point. Should I hook it to the motor and see if I can get it to discharge and balance under that load?
this is a new battery and charger so I would hate to cause more damage.

Any help would be appreciated.

UPDATE: I reassembled the Edge on the trike and and put the front end on a stand so I could run the motor using the throttle. It started off at 55v, ran 12 minutes and quit. It was down to 53v but recovered to 53.2v. After 10 minutes it started up again and ran for 7 minutes. Votage fell to 53.0v and recovered to 53.1 after the motor quit. I let it rest for 20 minutes and it started again but only ran for 10 minutes and quit. It shut down reading 53.0 and recovered to 53.1v. Lights stay on for the throttle all the time. Nothing is getting warm; not controller, nor battery cells.
Three questions:
1) Should I keep doing this until iI drain the battery low enough for shutoff or will it cause damage?
2) Will draining the battery down help it to finally balance when I recharge it?
3) What is causing the shutdown? Controller? BMS?

[Bikemad]

Hi Kim andto the forum.

I suspect the high voltage cells (>3.65V) are causing the BMS to disconnect the charge current, and the lower cells are unable to charge up properly.

If a single cell is too low, it could trigger the Low Voltage cutoff function of the BMS under load to prevent it from discharging any further.

Your pack appears to be quite unbalanced, and I suspect it will require some manual balancing by either charging the lower cells, or discharging the higher cells until all the cells are at the same unloaded voltage.

I don't know how much current the BMS is able to bleed off from the higher cells while balancing, but if it is less than 50mA, it could take days or even weeks to properly balance with the charger:



A 12V 60/55W car headlamp bulb (or a 12V 50W halogen downlight bulb) can be used to discharge higher voltage cells, but you will have to keep an eye on the voltage to make sure it doesn't drop too low.

It should be possible to discharge individual cells via the balance lead plug which connects to the BMS.

The balance lead wires are usually rated for ~3 Amps, but the maximum current from a 12V 50W halogen bulb should be less than 2.5 Amps when connected across a single LiFePO₄ cell.

If the voltage of the high cells drops very quickly to the same voltage as the lowest cell with the lamp attached, it is probably a weak/faulty cell.

Unfortunately, until the cells are reasonably balanced, it is difficult to know whether the battery is going to work correctly.

UPDATE: I reassembled the Edge on the trike and and put the front end on a stand so I could run the motor using the throttle. It started off at 55v, ran 12 minutes and quit. It was down to 53v but recovered to 53.2v. After 10 minutes it started up again and ran for 7 minutes. Votage fell to 53.0v and recovered to 53.1 after the motor quit. I let it rest for 20 minutes and it started again but only ran for 10 minutes and quit. It shut down reading 53.0 and recovered to 53.1v. Lights stay on for the throttle all the time. Nothing is getting warm; not controller, nor battery cells.
Three questions:
1) Should I keep doing this until I drain the battery low enough for shutoff or will it cause damage?
2) Will draining the battery down help it to finally balance when I recharge it?
3) What is causing the shutdown? Controller? BMS?

1) It should not cause damage as the BMS will not allow any cell to be discharged too far, but I'm not sure if it will help either.

2) All cells are being discharged and charge at the same rate, the only balancing that can take place is when the highest cell reaches 3.65V. As I mentioned previously, if the bleed current is very low, it will take a very long time for the cells to balance correctly when they are their voltages are so different.

3) Unless you have the Low voltage setting in the controller set to 53V (I'm not even sure if it would even accept that value) then it is probably the lowest cell in the pack that is causing the BMS to cut the power. But I don't understand why the motor won't run if you still have 53V showing and all 3 lights lit on the throttle.

Alan
 

[KimABQ]

Alan,
Thank you for the thoughtful response. I think you are on to some important points. Let me offer a theory and let me know if it makes sense, or not! 
If the BMS is cutting off power when one cell hits the minimum voltage, that might explain why it quits for a while. I had thought the shutdown had to be thermally induced, but if the BMS can control output then a low cell cutoff of the power could explain the mystery shutdown while I still have 53v for the pack.
Then, would the low cells try to adjust themselves by siphoning power from the higher voltage cells? If they do, or the BMS does, that would explain why being shut down for a period of time allows the low cell to gain some voltage and it's good to go until it gets pulled down to minimum again. This is intriguing.
As I mentioned in my first post, I have been draining the high voltage cells with a small LED circuit which pulls down from ~ 3.75 to 3.62 in 10 minutes or so. The more high voltage cells there are, the longer it takes to get the pack ready to receive another charge. I do this simply measuring across the terminals in the fully assembled pack, then clip the LED load to those terminals until the cell drops. I do this to avoid taking the pack apart (see attached picture) but maybe this is not the best approach? If nothing else it does take a long time, especially as more cells get above 3.62 and also have to be discharged.
I currently do not have any ability to reset the controller. I was considering the Android / Bluetooth option but I don't know if it offers those capabilities. It sounds like it would not be able to diagnose this problem or even capture error codes from the controller.
I will get a headlamp bulb to start draining cells with. as long as my wife is not watching her car too closely!  Can I continue to use my method for draining individual cells with the bulb approach?
Thank again!
Kim

[Bikemad]

Would the low cells try to adjust themselves by siphoning power from the higher voltage cells?

When a cell is placed under load, its voltage will sag in relation to the current being drawn, but if the load is removed, the cell's voltage will usually rise back up again.

The following graph shows just how much a weak cell can drop under heavy load:



The above graph plots the voltages of 2 individual cells in a 5 cell 5Ah 20C LiPo pack being used on my electric strimmer.
You can clearly see that the cell voltages gradually increase at 41m 40s when strimming is paused, and again when the strimmer finally stops completely at 49m 10s.


When your battery is put under load, as soon as the lowest cell drops below the BMS minimum cell voltage setting, the BMS will automatically cut the power going to the controller and the load on the cell is then removed.

The voltage of the low cell will then gradually rise above the BMS minimum cell voltage setting, allowing the BMS to reset the power output.

The MP Edge can be programmed via the Bluetooth dongle and suitable Android phone, but it would not be much use for diagnostics as the Bluetooth connection would also be lost as soon as the BMS cuts the power to the controller. 

Can I continue to use my method for draining individual cells with the bulb approach?

Yes, but make sure that you monitor the cell voltage while the bulb is clipped on, and also check how much the voltage rises again when the bulb is removed. 

You will probably need to discharge them further until the resting voltage is about the same as the lowest cell.

I was also surprised at the layout of the cells in your battery, I thought that they would all be in a single row like the ones in the 48V 10Ah LiFePO₄ battery pack that Gary dismantled.

Alan
 

[KimABQ]

Thanks again Alan. Very useful info for troubleshooting this problem.
I have been continuing my experiment with discharging the battery by running the motor while the wheel is suspended. Basically I just clamp the throttle wide open and let it run until it quits. I monitor runtime, start voltage, shutdown voltage, and what voltage it returns to within 30 seconds. I wait for one hour between runs (except overnight) and it starts every time. I have more than 15 runs so far that started with 9-11 minutes before shutdown and start voltages were around 53.2. Now my runs are lasting 20-30 minutes with start voltages now around 52.6. with these accumulated runs I now have a total of almost 4 hours of run time and it is gaining run time with each cycle. It is like the cells are getting closer to being balanced.
However, I may be violating a basic rule because I have not charged the pack at all between runs because that seemed to reset everything when I did it before. Should I recharge or can I continue this way for a while longer?
I too was surprised by the cell arrangement. Maybe this assembly eliminates case shorts? It does make it awkward to test individual cells.
Good point about the BT / Android not functioning with BMS shutdown. Is there any display that does track these issues?
I found that 220mm heat shrink is a great size for this pack, but 250mm works well too.
Thanks again!
Kim

[Bikemad]

Hi Kim,

I would be tempted to charge the pack until the green light comes on and then measure/record the cell voltages. Then I would leave it connected to the charger overnight and measure/record the cell voltages again to see if there is any difference.

It might be worth leaving the battery connected to the controller and charger while both are turned on. The residual current draw from the controller and Throttle LEDs etc. should cause the high cell voltage to drop slightly and allow the charger to cut in again. Each time it cuts in, it should begin to very slowly balance the cells until the charger cuts off again.



I am not aware of any display that could be used to diagnose this particular problem, but it will be interesting to see if fully balancing the cells completely cures the fault.

Alan
 

[Adamsavage79]

If your pack is unbalanced, then I would suggest putting it on the charger and leaving it on after the green light comes on. It would be good to do this overnight. The more you discharge the pack, the longer it will take to balance the pack.

[KimABQ]

While I do not understand why discharging keeps the pack unbalanced, I did as you suggested and let the pack stay on the charger overnight without any load. It did not seem to change the pack since it can still only get up to about 55.v. 

[Bikemad]

I would be tempted to charge the pack until the green light comes on and then measure/record the cell voltages. Then I would leave it connected to the charger overnight and measure/record the cell voltages again to see if there is any difference.

Did you also measure the cell voltages before and after the overnight charge? If so, was there any noticeable difference between the highest and lowest cells on both checks?

The more you discharge the pack, the longer it will take to balance the pack.

I do not understand why discharging keeps the pack unbalanced

Passive balancing will only commence when the highest voltage cell in the pack reaches its maximum charge voltage (typically ~3.65V), therefore it will obviously take longer to reach this balancing point if the battery is more deeply discharged prior to charging.

For a reasonably well balanced battery, an overnight charge might be sufficient to balance the cells, but as your cells are so far out of balance, it would only make a very slight difference to the balancing of the cells.

Also, if you have the "UL" version of the charger, leaving it on overnight is not likely to be beneficial as the charger would not cut in again to continue charging:

The main thing is the charger Alan is showing is the CE version of the chargers. I did have those in the past but switched to the UL versions as that is what is required in North America. The UL version does not have a switch for 240. It also does not cycle back on as the voltage drops when the cells are balancing. So with the CE version it will keep coming back on topping up the cells and the UL version it does not.

Manual balancing using a light bulb to discharge the high cells is probably the easiest (and quickest) way to balance your battery.
The professional way (but a lot more involved) would entail dismantling the pack and connecting all of the cells in parallel for a few hours until their voltages were all stabilised at exactly the same level.

Running the battery down with no load is basically draining some of the available capacity from all of the cells (both high and Low) by a similar amount, and this is unlikely to have any significant effect upon the balancing of the cells.

At this stage, we are still basically guessing as to what is causing the cutting out, if you were able to confirm that one (or more) of the cells was unusually low prior to the motor cutting out, then it might confirm a weak or imbalanced cell.

If it still cuts out with no obvious sign of a very low voltage cell, it might even be a controller problem. 

If the BMS is cutting the power, I would expect the LEDs on the throttle to go off at the same time, like they do on Nathan's video demonstrating his BMS problem.

Alan
 

[KimABQ]

Here is a synopsis of the progress so far:

• Battery began cutting out after a few minutes of riding. It finally got to a very predictable distance (~1km) before it quit. After a period of time it would run again for a few minutes. Nothing was hot, battery measured ~53v, all throttle lights were on.
  Conducted standard troubleshooting of wiring, etc. No problems found.
  Removed battery case and shrink wrap to measure cells and found significant imbalance between cells. Tried to pull down higher cells with LED load and charger would kick on for short periods, but lower cells would not gain enough charge.
  Connected battery to bike motor and ran until it quit. Let it recover for at least one hour then run again until it quit. Run times started at 9 minutes but increased each time until 18 minutes of run time was achieved on the seventh run.
  Battery set overnight without load or charge. First run the next day was 37 minutes. Completyed five runs that day with all but one above 20 minutes.
  One run the next day of 23 minutes.
  One run the following day of 61 minutes. Have not charged the battery yet.
  Due to advice from forum I began to charge the pack between runs and got two runs of 28 minutes and 22 minutes. Not gaining much ground.
  Spoke with Gary about doing a complete tear down of the pack to analyze and charge each cell. He recommended charging cells in the pack without tear down and mentioned a single cell charger. Light bulb went off for me after the conversation because I remembered I have a Nitecore D4 Digicharger and it has a LiFePo mode on it. So I used it to bring the low cells up to ~3.6v and they all settled back to ~3.5v. I also discharged any cells above 3.6v.
  after everything was completed I put the pack back on the charger and it charged for a short while but the pack voltage came up to 58v! As of this moment the pack is at 58.2v. And the charger does kick on occasionally - I do not see a UL rating on it, only CE so I may have gotten lucky!
  

I will let the charger work on the pack overnight and recheck cell voltages in the morning. I may have found a solution but I will have ot run it on the trike a few times to be sure.
Definitely looks like progress! 

[Adamsavage79]

It sounds like you found a way to balance the pack. Work in progress. Keep up updated!

[GM Canada]

I have been watching this thread from the sidelines as I like hearing other opinions on topics without injecting my own sometimes.

You likely have a CE charger. We only carried the UL ones for a short period of time. They are more expencive, lower quality and don't work well at cell balanging.

If the pack is up to 58.2 and the chargers does keep cycling it sounds promising the pack is balancing. Once you are on the road again with this pack charge it fully after each ride and leave it on the charger to continue to cell balance. I generally leave my packs on the charger unless its been a week since my last ride. Then I charge them once a month minimum and top them off before a ride if the pack was off the charger.

Gary

[Adamsavage79]

I don't use my bike to go to work or  anything, so I always leave it on the charger overnight. So about 8-9 hours and then I let the pack sit for several hours before using it. I'm determined to get the max life out of this pack, this time. Anyway, hopefully you are on the road to getting this fixed.

[KimABQ]

UPDATE: I left it on the charger last night and this morning the pack measured 57.2 for a loss of one volt from last night.


I finally got time to measure the cells and the pack voltage had dropped ome more to 56.8. Here are the cell voltages:

Cell #   V      Cell #   V
1   3.47      9   3.51
2   3.57      10   3.52
3   3.63      11   3.49
4   3.50      12   3.55
5   3.72      13   3.44
6   3.52      14   3.68
7   3.61      15   3.52
8   3.64      16   3.45
   
TOTAL 28.66      28.16
            
      Pack Tot   56.82

Considering the wide variations I do not think the pack is balancing. 

Thoughts?

[Bikemad]

Hi Kim,

Now that all the cell voltages are at a more even level, have you tested the battery on the trike to see if the motor still cuts out?

As the cells are still out of balance, it might be worth leaving it on the charger overnight again and remeasuring the cells to see exactly what changes.

I have entered your last results into a spreadsheet to see how they compare and here is the outcome:



Most of the cells are below the typical 3.65V ±0.025V maximum cell voltage, but there is no obvious sign of a rogue cell that stands out from all the others. Although, placing a load on the pack and measuring the cell voltages while under load might produce a different result.

If the BMS is not balancing the cells, it is either faulty, or it simply needs a lot longer to reduce the 0.28V difference between the highest and the lowest cell voltages.

Another overnight charge is not going to fully balance the cells, but if the BMS is working correctly, it should at least show a slight improvement in the current 0.28V difference between the highest and lowest cells.

Balancing cells does take time, My LiPo balance charger can drain up to 300mA per cell while passive balancing, and I have it set to balance throughout the charge process rather than just at the end, but it still takes a long time to balance a fully charged 4 cell pack with a 0.15V difference:



If your BMS is only able to drain ~30mA, it will obviously take a lot longer, especially when you consider that the low cells can only increase in voltage while the charger is actually charging, in other words, not while the green light is on.

Alan