Balancing a grossly imbalanced 14s14p 18650 battery

Lateksinen

1 mW
Joined
Jun 19, 2017
Messages
14
Hello. I have been reading various threads about balancing but have not found an answer yet so I decided to do a post.

I have a 14s4p 52v 14Ah battery with Sanyo 18650GA cells. My problem is that one of the cell groups (4 cells in parallel) is grossly imbalanced at 3.7v while others are fully charged at 4.2v. The reason for the gross imbalance is that the cable from this one cell group had detached and the cells were left at 3.7 volts. I then fully charged the battery after that so that all the other cell groups are at 4.2 volts. When I soldered the positive cable back to this one cell group, I didn't realize I would end up with a very imbalanced battery.

The BMS is either very slow at balancing the battery or it is not balancing. After an all night and day charge the one cell group still sits at 3.7 v. Any idea what the balancing current could be, approximately? What would be the best way to balance the battery faster than what the BMS is doing? Should I detach the wire again from the one cell group and drain the other cell groups to around 3.7v and then solder the wire back? Or somehow individually charge the lowest cell group to 4.2 v? I do not own a charger capable of that, however. Could it be possible to make one from a mobile phone charger? Is 5 volts too much for charging a single cell group?

Any other suggestions? Thank you in advance.
 
What would be the best way to balance the battery faster than what the BMS is doing?
Best way? Get 14 ISOLATED single cell AC to DC lithium ion chargers, make sure they are all running at exactly the same voltage, and connect them all. When they are all complete, pack is balanced and fully charged.

Other ways -

Get 1 AC to DC single cell li-ion charger and connect it to all 14 portions of the pack in order. Make sure to leave it on until charge termination (i.e. could be many hours.) Will take forever but be cheaper.

Do some fancy charge-transfer thing with a bunch of relays, a big capacitor and a controller.
 
billvon said:
What would be the best way to balance the battery faster than what the BMS is doing?
Best way? Get 14 ISOLATED single cell AC to DC lithium ion chargers, make sure they are all running at exactly the same voltage, and connect them all. When they are all complete, pack is balanced and fully charged.

Other ways -

Get 1 AC to DC single cell li-ion charger and connect it to all 14 portions of the pack in order. Make sure to leave it on until charge termination (i.e. could be many hours.) Will take forever but be cheaper.

Do some fancy charge-transfer thing with a bunch of relays, a big capacitor and a controller.

Thank you for your help. Could I also just connect a single cell charger to the lowest group and let it fully charge to 4.2v? That way it would be pretty close to the other fully charged groups and the BMS could probably take care of the fine tune balancing after that?
 
Lateksinen said:
billvon said:
What would be the best way to balance the battery faster than what the BMS is doing?
Best way? Get 14 ISOLATED single cell AC to DC lithium ion chargers, make sure they are all running at exactly the same voltage, and connect them all. When they are all complete, pack is balanced and fully charged.

Other ways -

Get 1 AC to DC single cell li-ion charger and connect it to all 14 portions of the pack in order. Make sure to leave it on until charge termination (i.e. could be many hours.) Will take forever but be cheaper.

Do some fancy charge-transfer thing with a bunch of relays, a big capacitor and a controller.

Thank you for your help. Could I also just connect a single cell charger to the lowest group and let it fully charge to 4.2v? That way it would be pretty close to the other fully charged groups and the BMS could probably take care of the fine tune balancing after that?
You could - but it's far better to bring them all to the same SOC.
 
I've had a similar situation before. I used a bench power supply as a single cell charger and just charged the low cells to make them match the rest.

I found some solid strand wire in my junk that fits perfectly into the holes on the BMS balance wire connector. I used a pair to charge the low cells. Moving the pair from one cell group to the next also made voltage measurements easy. It took a few days, but finally got all the cells within 10mV or so and I was good to go.

Whether you use a bench supply or a single cell charger, just make sure you charge to the same voltage and don't exceed 4.2v.
 
Yes to your question. You can charge a single cell with a 5v or 4.2v phone charger it will charge very slow. But when it hits 4.15v or so and will rack faster then you can get a beer from the cooler at your feet and hit 4.2v BUT won't stop and we'll race all the way to 5 volts because there's no place for the charge to go over 4.18v or so. And can go to thermal Runaway and might not stop Andrew Nall cells in that Parallel Group. So you man Enough To Try you have to sit and watch and monitor and sit and watch and monitor don't lose your span of attention . Or use a 4.2 single cell charger that stops at 4.2 made for the purpose.
Just because there is 5 V printed on the side of that phone charger it could be 7.2 volt or you might find a 4 volt charger that is 4.5 volt meaning cut the wires and check and next time you look in the trash can or the salvage store go pick up five or six of those for a buck or two or go dumpster diving. I just got 5 haven't checked them out but I couldn't let him go into the soil. Plus I guess there's a resistor or maybe a lightbulb that will pull it down ??? You play with battery you sit with battery. You have to take a piss unplug it.
 
Lateksinen said:
I have a 14s4p 52v 14Ah battery with Sanyo 18650GA cells. My problem is that one of the cell groups (4 cells in parallel) is grossly imbalanced at 3.7v while others are fully charged at 4.2v. The reason for the gross imbalance is that the cable from this one cell group had detached and the cells were left at 3.7 volts. I then fully charged the battery after that so that all the other cell groups are at 4.2 volts. When I soldered the positive cable back to this one cell group, I didn't realize I would end up with a very imbalanced battery

How exactly did you charge all the cells except that group?

If you hooked up a charger to the main + and - of the pack to do ti, then you also charged those cells with the disconnected balance wire, not just all the others.

It does not matter for that whether the balance wire to that group is connected or not, that group charges exactly like all the others in a series charge like that.

In a series "bulk charge" setup, using a BMS, the only thing that wire does is allow the BMS to monitor and *discharge* that cell group. (it doesn't charge thru that wire).

If that's the case, then something is wrong with that cell group; it probably has a defective cell (or cells) in it, that are internally leaking / dissipating charge, or they are so much lower in capacity than the others that they ran down really far during the ride and now can't fully charge up as a simple bulk charge, without potentially days/weeks of balancing (depending on how far down they are relative to the others, vs how much balance current the BMS can deal with per cell group for the other groups).


So you'll need to diagnose that cell group to find out which cell or cells are the problem, or just replace the whole group (easier) with a new one of the same kind of cells.





If you charged by connecting individual single cell chargers, or a multi-cell RC charger, to all the cells except that group, then you could indeed have charged just the other cells but not that one, if the charger was setup to ignore the balance wire voltages or it's also broken and not doing it's job.

If that's the case, you can single-cell charge the low group. The easiest way is just to reset the charger you already used for just one cell, and connect only the main + and - out of it to your imbalanced gropu.

I'd recommend using a charger made for that purpose, for safety.

Or if you have one, an adjustable voltage and current power supply "lab psu".


You can use a cellphone charger, but you must stay right there and monitor it with a voltmeter you are actually watching constantly, and disconnect it as soon as you see it rise to 4.2v (or less).

If you don't stay and watch it,, there's risk of serious overcharge, cell damage, potentially fire, etc. (and the fire risk is permanent after the damage happens, not just for right then).



Anyway, the likelihood is that there is something wrong with that group to be so low relative to the others, so you should fix that before doing anything else anyway.
 
amberwolf said:
Lateksinen said:
I have a 14s4p 52v 14Ah battery with Sanyo 18650GA cells. My problem is that one of the cell groups (4 cells in parallel) is grossly imbalanced at 3.7v while others are fully charged at 4.2v. The reason for the gross imbalance is that the cable from this one cell group had detached and the cells were left at 3.7 volts. I then fully charged the battery after that so that all the other cell groups are at 4.2 volts. When I soldered the positive cable back to this one cell group, I didn't realize I would end up with a very imbalanced battery

How exactly did you charge all the cells except that group?

If you hooked up a charger to the main + and - of the pack to do ti, then you also charged those cells with the disconnected balance wire, not just all the others.

It does not matter for that whether the balance wire to that group is connected or not, that group charges exactly like all the others in a series charge like that.

In a series "bulk charge" setup, using a BMS, the only thing that wire does is allow the BMS to monitor and *discharge* that cell group. (it doesn't charge thru that wire).

If that's the case, then something is wrong with that cell group; it probably has a defective cell (or cells) in it, that are internally leaking / dissipating charge, or they are so much lower in capacity than the others that they ran down really far during the ride and now can't fully charge up as a simple bulk charge, without potentially days/weeks of balancing (depending on how far down they are relative to the others, vs how much balance current the BMS can deal with per cell group for the other groups).


So you'll need to diagnose that cell group to find out which cell or cells are the problem, or just replace the whole group (easier) with a new one of the same kind of cells.





If you charged by connecting individual single cell chargers, or a multi-cell RC charger, to all the cells except that group, then you could indeed have charged just the other cells but not that one, if the charger was setup to ignore the balance wire voltages or it's also broken and not doing it's job.

If that's the case, you can single-cell charge the low group. The easiest way is just to reset the charger you already used for just one cell, and connect only the main + and - out of it to your imbalanced gropu.

I'd recommend using a charger made for that purpose, for safety.

Or if you have one, an adjustable voltage and current power supply "lab psu".


You can use a cellphone charger, but you must stay right there and monitor it with a voltmeter you are actually watching constantly, and disconnect it as soon as you see it rise to 4.2v (or less).

If you don't stay and watch it,, there's risk of serious overcharge, cell damage, potentially fire, etc. (and the fire risk is permanent after the damage happens, not just for right then).



Anyway, the likelihood is that there is something wrong with that group to be so low relative to the others, so you should fix that before doing anything else anyway.

I charged the battery through the charging connector in the case. It was not the balancing white that was detached from the first cell group, it was the main positive wire. How is it possible that the battery read 54 volts from the connector outside the case if the main positive was not connected? Also, I don't quite understand the wirings as I was also able to charge all the other cell groups even if that main positive cable was not attached.

I went to a local battery shop and they promised to fully charge the cell group at 3.7 volts with a special charger/tester and run some decharge-recharge cycles to test the capacity of the cells. Will see how that turns out in a couple of days. Hopefully I'll have a balanced battery.
 
Lateksinen said:
I charged the battery through the charging connector in the case. It was not the balancing white that was detached from the first cell group, it was the main positive wire. How is it possible that the battery read 54 volts from the connector outside the case if the main positive was not connected? Also, I don't quite understand the wirings as I was also able to charge all the other cell groups even if that main positive cable was not attached.

The part I've bolded/underlined above is crucial information that you did not provide in your original post. Knowing that it is the *first* cell group, and knowing that it is the main + wire (for the whole pack) changes the problem and the troubleshooting steps.


If you mean the main positive wire of the entire battery was not connected to the battery at all, then there isn't a way you could read any battery voltage on that wire. It would have to be connected to a voltage source to read a voltage. The only other voltage source would be the capacitors in teh controller, which will discharge quickly when the controller is on and the battery is not connected to it, so I doubt that is what your meter was reading.

So my guess, without pictures and diagrams of your pack and wiring, is that you are actually reading voltage thru the charge port, not the discharge port, and that there are two separate wires for the pack main positive--one for charge, and one for discharge, and only the discharge wire was disconnected.


I think you need to post pictures of the interior of the pack, pointing out the connections in question, so we can help you determine what is actually wrong with the pack and get it fixed correctly, so you don't have problems with it in the future.



Some more explanation:

If no main positive cable (series connection to rest of pack) is connected to a group of cells that is in series with other cells, then none of the cells in the entire series set can be charged *or discharged* via a series connection.

If it did actually charge via the main charge port on the battery, while that wire was disconnected, then there *must* be a second wire to that same point on the first set of cells, running from only the charge port to it, and not from the discharge port.


All that said, there is probably still something wrong with the group of cells that is low, that caused them to go low in the first place, relative to the other cells.

Without knowing more about your pack wiring, the only thing I can think of that could possibly have created this situation, without a problem in those cells is:
--during discharge, the main pack + wire came off the end of the pack
--bike stopped working
--main pack + wire moved and touched the negative end of the now-low cell group, essentially simply lowering the entire pack's voltage by one cell
--charging then charged up the cells that are still connected, and not the now-low group, so the now-low group didn't get charged because it wasnt' connected to the charger
--moving the wire back to the original position now leaves a pack that has one cell group still discharged just like it was when it came off, and the rest of the pack charged to full.

Assuming this is all the case, then moving the wire back to the accidental position and then discharging the battery by riding or running motor off-ground would let you bring the other cells down to where the low group is.

Then fixing the wire back to it's original position, and recharging the pack, and letting it sit for a while balancing, would fix the imbalance problem.
 
OP - check to make sure the balancing circuit isn't ON due to a defective switching transistor. This is often why cell banks go out of balance in the 1st place.

In other words, the bleed resistors are only supposed to on to drain charge from the highest voltage cells. If/when the switching transistor shorts, the resistor is simply ON all of the time constantly bleeding the cell group.

Simply locate and feel the bleed balance resistors and if the one connected to the problem bank seems warm then check for voltage drop across it. Compare readings to other resistor side by side on the PCB.
 
Something odd is going on here, and I'd advise you to not just get it fixed once, and forget about it.

Like Ykick suggested, it could be a faulty BMS.

There could be strain on your pack that will damage the positive cable again.

You could have bad cells.

Whatever the end result is, I would advise against just assuming it's permanently fixed.

I would check after every ride 3 times,

Then check every week 3 times

Then check every month 3 times, and if it's still good after that, accept that the problem has gone. Until then, I'd assume the pack could fail at any time, losing me an expensive battery if I don't watch it.
 
Alright. So I got the battery pack back from the repair shop. The first four cells had a lowered capacity (66%of 3500mAh) despite being almost brand new. They were replaced with new cells and fully charged to the same voltage as the other cells before spot welding in series. The battery worked well for one discharge, managed to ride 60 km on it before hitting the LVC. Then I fully charged the battery and rode a couple hundred meters. The display voltage meter started to show numbers between 48 and 58 volts, jumping randomly. Then the display shut off in a way it had hit the LVC. External volt meter read 57-58 volts and the display turned back on after removing the battery shortly. So now it seems to have this strange behaviour of showing jumping number of voltage under load (48-58v)and turning off suddenly. All the cell groups read same voltage, 4.1 volts, when I opened the battery case. So it doesn't seem to be unbalanced anymore. Could it be a faulty BMS? Any explanation for the voltage numbers jumping under load and sudden shutdowns?
 
Sounds like a loose connection somewhere. Could be the BMS or anywhere between the cells and the controller.
 
999zip999 said:
As the rule don't discharge your battery to lvc. It's better to leave something in the pack.

Well, the controller LVC was set to 45 volts and the pack actually ceased working at 47-48 volts (measured under no load). So I thought it was pretty safe. Wanted to see when it would cut the current. But yes, I'm planning on keeping the pack between 20-80% of charge to extend its life cycle.
 
Now this gets interesting. I got a second battery pack and the bike worked flawlessly with it for 50 km. Then the screen started to switch off suddenly. Detaching the battery or just turning the power on again helped for a short time (like 30s) and then shutoff again. Voltage was at about 50.I charged the battery fully again and was hoping for the symptoms to cure. But no. Even with a fully charged battery (both batteries) the screen (power) switches off when given some throttle. Then it won't switch on again until I detach the battery for just a couple of seconds. It does this over and over again.


So my question is: do I have two somehow faulty battery packs (unlikely I guess) or is it a broken controller? Or something else? Any way to differentiate between battery/BMS related cause from controller failure?
 
If there is something wrong with the controller such that it draws much more than the normal current, it might be tripping the BMS overcurrent protection.
Two bad BMSs seems unlikely, but not out of the question. This could happen if you had a shorted phase, either in the motor or the controller.

The best is if you have a way to measure the battery current when running the motor. You can also test the controller separately for shorts using a meter. This is done by disconnecting the motor phase wires and the battery, then use ohmmeter to measure each phase wire against the positive battery terminal on the controller, and each phase against the negative input to the controller. 6 total measurements. None of these should look like a short, but should look like a diode. Reversing the meter probes on each measurement you should see high resistance in one direction.
 
fechter said:
If there is something wrong with the controller such that it draws much more than the normal current, it might be tripping the BMS overcurrent protection.
Two bad BMSs seems unlikely, but not out of the question. This could happen if you had a shorted phase, either in the motor or the controller.

The best is if you have a way to measure the battery current when running the motor. You can also test the controller separately for shorts using a meter. This is done by disconnecting the motor phase wires and the battery, then use ohmmeter to measure each phase wire against the positive battery terminal on the controller, and each phase against the negative input to the controller. 6 total measurements. None of these should look like a short, but should look like a diode. Reversing the meter probes on each measurement you should see high resistance in one direction.

Thanks for help. I forgot to mention that I opened the controller and one of the phase wires (Green) had peeled insulation so that the wire was naked for about 1 cm. I fixed it but now that you mentioned about it, could that have shorted and caused bms failure?

Also, I wonder if it could be a faulty display? I found a thread with similar problem, which was fixed by replacing the display. https://endless-sphere.com/forums/viewtopic.php?f=28&t=75115

I guess I could see if detaching the display cable helps to get the power on again. I assume it would indicate a problem with display. I guess when I detach the battery pack it also resets the display.
 
Normally a controller fault shouldn't damage a BMS but anything is possible. You could try unplugging the display as a test. That might reset the display without resetting the BMS.

A way to measure the current is what you really need. Does the display show the current or motor power?

Does the motor seem to run normally before it cuts off? If you had a blown FET, it would not run normally.
 
fechter said:
Normally a controller fault shouldn't damage a BMS but anything is possible. You could try unplugging the display as a test. That might reset the display without resetting the BMS.

A way to measure the current is what you really need. Does the display show the current or motor power?

Does the motor seem to run normally before it cuts off? If you had a blown FET, it would not run normally.


Unplugging the Display after a shutoff didn't help. It still didn't power on until detaching the battery for just a couple of seconds. Then it runs okay for a few seconds up to one minute after cutting off again. I measured the voltage through the insulation of the wires coming from the battery after a cutoff. It read 7 volts. Then I detached the battery and voltage bumped to 57. So to the best of my understanding, it appears that the BMS is cutting off and then resetting after detaching the battery. Could it be faulty? Somehow hitting the LVC/some other cutoff under load despite all cell groups showin good voltages? And both batteries seem to have this same problem. Is there any way by measuring voltages from the BMS to see if it is faulty? I believe it shows normal voltage after I detach the battery.


The motor draws correct amounts of power under load, around 1500w with full throttle. Also works with no problems until it cuts off.
 
The BMS can trip due to a cell voltage being too low or due to current being too high. If the current was too high, I'd expect it to trip right away when you give full throttle. The pack voltage looks OK with no load.

If the BMS is accessible, you can measure the individual cell voltages either on the connector or on the row if resistors that do the balancing. If one cell is significantly lower or higher than the rest, it can cause the kind of behavior you are seeing.
 
Back
Top