Author Topic: 48v10Ah battery problems  (Read 8204 times)

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #15 on: May 26, 2019, 12:07:20 AM »
Thanks again Alan. I appreciate your wisdom and insights.
I did not test the pack on the battery because I would have to put new heat shrink on before assembly and I was not ready to trust it yet.
I went back over all the tests I logged and put them all in the same type of table that you used. I was able to see that the cell voltage differences have been getting smaller....slowly!
I spent today manually balancing the cells and this is what I ended up with:

5/25 Post Balance Measurement-------:            
Cell #   V      Cell #   V
1   3.63      7   3.61
2   3.62      2   3.62
3   3.65      4   3.62
4   3.62      5   3.62
5   3.62      6   3.62
6   3.62      10   3.62
7   3.61      11   3.62
8   3.64      12   3.62
9   3.64      13   3.62
10   3.62      14   3.62
11   3.62      1   3.63
12   3.62      8   3.64
13   3.62      9   3.64
14   3.62      16   3.64
15   3.64      15   3.64
16   3.64      3   3.65
            
Pack Tot   58.03      Pack Tot   58.03
            
Max   3.65      Max   3.65
Min   3.61      Min   3.61
Diff   0.04      Diff   0.04

This is as close as I could get because the cells would change as others were being charged or discharged so I quit chasing the rabbits.
I will let it on the charger overnight because I want to see if it gets better or worse before I put it back in the heat shrink and case for a road test.
Also, the total pack voltage is less than the 58.4v charger rating. Is that okay?
I hope this is progress.  :-\
Have a great Memorial Day weekend!
Kim



« Last Edit: May 26, 2019, 03:56:10 AM by KimABQ »

Offline Adamsavage79

  • Confirmed
  • Bachelor of Magic
  • ****
  • Posts: 172
Re: 48v10Ah battery problems
« Reply #16 on: May 26, 2019, 01:12:03 AM »
Thanks again Alan. I appreciate your wisdom and insights.
I did not test the pack on the battery because I would have to put new heat shrink on before assembly and I was not ready to trust it yet.
I went back over all the tests I logged and put them all in the same type of table that you used. I was able to see that the cell voltage differences have been getting smaller....slowly!
I spent today manually balancing the cells and this is what I ended up with:

http://5/25 Post Balance Measurement-------:   
         
Cell #   V      Cell #   V
1   3.63      7   3.61
2   3.62      2   3.62
3   3.65      4   3.62
4   3.62      5   3.62
5   3.62      6   3.62
6   3.62      10   3.62
7   3.61      11   3.62
8   3.64      12   3.62
9   3.64      13   3.62
10   3.62      14   3.62
11   3.62      1   3.63
12   3.62      8   3.64
13   3.62      9   3.64
14   3.62      16   3.64
15   3.64      15   3.64
16   3.64      3   3.65
            
Pack Tot   58.03      Pack Tot   58.03
            
Max   3.65      Max   3.65
Min   3.61      Min   3.61
Diff   0.04      Diff   0.04


This is as close as I could get because the cells would change as others were being charged or discharged so I quit chasing the rabbits.
I will let it on the charger overnight because I want to see if it gets better or worse before I put it back in the heat shrink and case for a road test.
Also, the total pack voltage is less than the 58.4v charger rating. Is that okay?
I hope this is progress.  :-\
Have a great Memorial Day weekend!
Kim

That looks much better. Less of a difference. You have a couple higher voltage cells, but overall the pack looks to be much more balanced. It's normal have a one cell @ 3.62 vs 3.61v. I believe as long as they are all with in .01-0.02 of each other, than that would be a fully balanced pack. Alan can correct me if I'm wrong.

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #17 on: May 26, 2019, 03:41:40 PM »
Good news and bad news.  ::)
Good news is that I left the pack on the charger overnight and it rose to 58.2v, the highest I have ever seen.
the bad news is in the cell voltage chart below:
5/26 Post Balance Measurement-------:            
Cell #   V      Cell #   V
1   3.59      16   3.54
2   3.66      13   3.56
3   3.71      1   3.59
4   3.60      11   3.59
5   3.75      4   3.60
6   3.62      10   3.61
7   3.68      6   3.62
8   3.72      9   3.62
9   3.62      15   3.62
10   3.61      12   3.64
11   3.59      2   3.66
12   3.64      7   3.68
13   3.56      3   3.71
14   3.72      8   3.72
15   3.62      14   3.72
16   3.54      5   3.75
            
Pack Tot   58.23      Pack Tot   58.23
            
Max   3.75      Max   3.75
Min   3.54      Min   3.54
Diff   0.21      Diff   0.21


It got out of balance again with a difference of 0.21v between high and low.
Any thoughts?
Kim


Offline Adamsavage79

  • Confirmed
  • Bachelor of Magic
  • ****
  • Posts: 172
Re: 48v10Ah battery problems
« Reply #18 on: May 26, 2019, 04:14:06 PM »
Is the charger itself throwing the pack out of balance ? The reason I'm suggesting this, is because you had it nearly balanced and then you charged it. Never heard of a charger doing that, but I've heard of stranger things...

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,495
Re: 48v10Ah battery problems
« Reply #19 on: May 27, 2019, 12:18:20 AM »
Hi Kim,

I think you are worrying unnecessarily, because I definitely don't see those results as "bad news".
If you ignore the values obtained while manually balancing the pack, compared to your original results (Pre manual balancing) the pack has definitely improved significantly:



The difference between the highest and lowest cell (was 280mV but now 210mV) has actually been reduced by 70mV, which is an overall improvement of 25%.

I suspect that another night on the charger (without manual balancing) will improve them even more (although it might only improve by a small amount) but I'm hopeful that the balance of the cells will continue to improve with regular charging.  ;)

The biggest problem with the manual balancing is the time it takes for each cell voltage to completely stabilise after you have manually topped it up, and the lower cells that have been topped up more, will typically drop more than the higher cells, even though you may have topped them all up to the exact same ~3.65V.

Even though the cell voltages will have dropped after they have fully stabilised, the resting cell voltages should still be nearer to the ~3.65V than they were before they were topped up.

Fortunately, the BMS is able to add capacity (using the same amount of current) to all of the low cells simultaneously at the end of the charge (but only when the charger cuts in again) therefore it should gradually increase the voltage of all the lower voltage cells if the battery is left charging for a long enough period of time.

The BMS does not have to "chase rabbits" one at a time, it simply slows all the fast rabbits down, which allows the slower ones to gradually catch up.  :)

Is the charger itself throwing the pack out of balance ?

It is definitely not the charger, as there is no way the charger could reduce the voltage of individual cells, only the BMS has the ability to do that.

As previously stated, the "post manual balancing" figures have definitely improved over the "pre manual balancing" figures, which is a positive and reassuring result.

Alan
 


Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #20 on: May 27, 2019, 02:54:45 AM »
Thanks Alan and Adamsavage799 for your willingness to continue helping. I really appreciate it.

Alan is correct that there has been improvement. My question was whether or not it was caused by the manual balancing effort. I am willing to give that another try if it speeds up the process.
But for now I will leave it on the charger for tonight. It has been there since yesterday so we'll see if there are any changes by tomorrow.

Should I try it on the trike or should I wait until it is in better shape? I don't want to make things worse.

And Alan, how do you post the spreadsheet table? I have not been able to figure that out even when I tried to post it as a graphic. So I end up with clunky, misaligned text instead of nice, color coded tables.

Thanks again,

Kim
« Last Edit: May 27, 2019, 02:59:35 AM by KimABQ »

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #21 on: May 27, 2019, 08:04:26 PM »
Here are today's measurements:
5/27 Post Balance Measurement-------:            
Cell #   V      Cell #   V
1   3.56      16   3.51
2   3.65      13   3.55
3   3.73      1   3.56
4   3.58      4   3.58
5   3.82      10   3.59
6   3.60      11   3.59
7   3.70      6   3.60
8   3.74      9   3.62
9   3.62      12   3.63
10   3.59      15   3.63
11   3.59      2   3.65
12   3.63      7   3.70
13   3.55      3   3.73
14   3.74      8   3.74
15   3.63      14   3.74
16   3.51      5   3.82
            
Pack Tot   58.24      Pack Tot   58.24
            
Max   3.82      Max   3.82
Min   3.51      Min   3.51
Diff   0.31      Diff   0.31

The pack stayed on the charger all night. It is a MEBO GY200-48AL rated at 58.4v and 3A. I have not seen it kick on to charge state since yesterday, but it may have when I was not watching.

The pack voltage rose by 0.01 but 10 cells fell in voltage, while only 5 cells rose. Number 5 is at 3.82v. That seems awfully high. The difference from high to low also rose to 0.31v.

I believe something is malfunctioning. Since some cells increased the power could have come from the charger or from other cells. Since cell #5 is at 3.82v that is well above the desired charge voltage for a balanced pack. And none of the cells changed in a way that bettered the pack balance, so I have to wonder if the BMS is the source of the problems. Is there a way to test the BMS?

The saga continues.
Kim


Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,495
Re: 48v10Ah battery problems
« Reply #22 on: May 27, 2019, 10:32:30 PM »
Hi Kim,

I have found some information for a typical 16s LiFePO4 BMS:



The typical overcharge protection voltage is 3.85V per cell. As your highest cell is at 3.82V, it would appear that the cell overvoltage protection function of the BMS is working correctly, but the balance function does not seem to be bleeding the excess voltage from the high cells as it should.



For the charger to cut back in again and begin topping up the lower cells, the highest cells would need to be below ~3.65V, but as you still have 5 cells that are 3.65v or over, this is unlikely to happen.

You currently have 7 cells below 3.61V (the typical balance voltage shown above) and 9 cells above it.
If the BMS was working correctly, I would expect the 9 high cells to gradually reduce to ~3.61V, but they obviously are not doing this.

As things have not improved as I had hoped after another overnight charge, it is beginning to look like the BMS will need to be replaced.  :(

I suggest that you contact Gary to see what needs to be done now, as the battery is presumably still under warranty.

Alan
 

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #23 on: May 28, 2019, 12:09:50 AM »
Alan,
I agree. I think the BMS is not bleeding the higher cells like it should so balancing cannot be achieved.
Thanks for the info - very useful. It might be worth creating a sticky thread explaining cell parameters, how they get out of balance, and how to diagnose and rebalance them This sure has been a learning opportunity for me. I hope a new BMS solves it!
I will contact Gary tomorrow.
Thanks again.
Kim

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #24 on: May 28, 2019, 06:30:43 PM »
On the charger all night. Here are today's measurements:
5/28 Post Balance Measurement-------:            
Cell #   V      Cell #   V
1   3.55      16   3.50
2   3.65      1   3.55
3   3.73      13   3.55
4   3.57      4   3.57
5   3.82      11   3.58
6   3.60      10   3.59
7   3.71      6   3.60
8   3.74      9   3.62
9   3.62      12   3.63
10   3.59      15   3.63
11   3.58      2   3.65
12   3.63      7   3.71
13   3.55      3   3.73
14   3.75      8   3.74
15   3.63      14   3.75
16   3.50      5   3.82
            
Pack Tot   58.22      Pack Tot   58.22
            
Max   3.82      Max   3.82
Min   3.50      Min   3.50
Diff   0.32      Diff   0.32

They are almost identical to yesterday. The BMS has not changed anything so I think it's safe to say that it has ben the problem.
Kim

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #25 on: June 01, 2019, 03:23:39 PM »
Hi all,

I contacted Gary on 5/28 and he said he would send a BMS to replace the one on my pack. So I hope to be back in operation again soon. The problem started the end of March and I have been down since then!  :'(

As for the BMS replacement, does anyone have any advice concerning does or don'ts when performing the removal of the old unit and installation of the new one? I want to be sure I do the repair properly since I will be working with enough current to cause damage to cells and/or BMS.

also, I am guessing it would be good to manually balance all cells before connecting the new BMS just to make its work easier to start with. Is this correct?

Thanks thanks again for your help, especially Alan. I would not have gotten this far without your feedback and insight.  ;)
Kim

Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #26 on: June 08, 2019, 07:00:45 PM »
I installed the new BMS and let it charge overnight. The pack came up to 58.2v this morning so that is a first.  ;D But the individual cells still varied by 0.26v so I manually balanced each cell and here are the current readings:

6/8 With New BMS -------:            
Cell #   V      Cell #   V
1   3.64      1   3.64
2   3.64      2   3.64
3   3.65      14   3.64
4   3.66      15   3.64
5   3.65      3   3.65
6   3.65      5   3.65
7   3.65      6   3.65
8   3.65      7   3.65
9   3.66      8   3.65
10   3.65      10   3.65
11   3.66      4   3.66
12   3.66      9   3.66
13   3.66      11   3.66
14   3.64      12   3.66
15   3.64      13   3.66
16   3.66      16   3.66
            
Pack Tot   58.42      Pack Tot   58.42
            
Max   3.66      Max   3.66
Min   3.64      Min   3.64
Diff   0.02      Diff   0.02

It looks pretty good so I will test it now on my bike with the wheel suspended. If it runs for an hour then I will heat shrink it and put it back in the aluminum case for a road test.

Thanks Gary for the replacement BMS and great customer service!

And thanks to Alan for his expert help on this mystery. In hindsight, I think the BMS was bad from the start but I never noticed it until the cells went way out of balance and began to shut the pack down. It has been a learning experience. I'll let you all know if it is finally fixed.

Kim


Offline KimABQ

  • Confirmed
  • New Member
  • *
  • Posts: 21
Re: 48v10Ah battery problems
« Reply #27 on: June 09, 2019, 02:22:55 AM »
Well, battery and BMS are working great. That's the good news.

Unfortunately the motor still only ran for about 10 minutes before stopping again. While nothing felt hot (other than the throttle which was much warmer than I remember), after a rest the motor would start again. I applied some freeze spray to different parts to see if I could find the thermal issue and as soon as I sprayed the controller the motor started again. I repeated this several times and the controller definitely did not like getting warm and started as soon as I sprayed it. The ambient temp was 90F, but I don't think that would keep the controller on the edge of shutdown.

I noticed that the front disk brake was slightly engaged and adding drag/ load to the motor because its cable was being pulled by part of the stand this time. I would think this would trip safety thermal or load sensors before any damage would occur, but now I believe it may have damaged the controller. Any thoughts on this?

Do I have to replace the controller or just some of the parts?

The saga continues!  :-\

Offline Adamsavage79

  • Confirmed
  • Bachelor of Magic
  • ****
  • Posts: 172
Re: 48v10Ah battery problems
« Reply #28 on: June 09, 2019, 02:57:11 AM »
Do you know how hot the controller was getting ?

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,495
Re: 48v10Ah battery problems
« Reply #29 on: June 09, 2019, 02:40:28 PM »
I noticed that the front disk brake was slightly engaged and adding drag/ load to the motor because its cable was being pulled by part of the stand this time. I would think this would trip safety thermal or load sensors before any damage would occur, but now I believe it may have damaged the controller.

A binding brake should not be able to cause permanent damage to the controller, as the thermal protection would prevent it.

A badly binding brake can place a great deal of load on the motor and this would cause the controller to get warm.  If the brake was binding enough to significantly raise the temperature of the controller, the controller's thermal protection would cut the power going to the motor to prevent further overheating.

From the motor's point of view, it would be similar to someone riding at full throttle up a 4 mile long hill at ~25 mph for 10 minutes  without easing off and with no pedal assistance from the rider (and also without the 25mph cooling breeze flowing over it), so it is hardly surprising that the inside of the controller eventually gets too hot, causing it to cut out. ::)

Although there it is a remote possibility that a faulty thermistor could be causing the thermal protection to cut in prematurely, I suspect that the binding brake is far more likely to be the culprit. ;)

How difficult is it to turn the front wheel by hand when the brake is being partially applied by the stand?

Do I have to replace the controller or just some of the parts?

If it is just the binding brake causing the cutting out, then it will not be necessary to replace the controller.  However, if the cutting out was due to a faulty temperature sensor (thermistor) then the entire controller would need to be replaced as it is not really practical to replace just the thermistor due to the construction of the controller.

I suggest that you ensure that the brakes are not binding and then see if the cutting out still occurs, if it runs without cutting out, then give it a trial run on the road and see how it behaves.

Alan