Here goes. I have three 36v 13Ah batteries for my ebike. One is 3 years old and the one that came with my bike. It is in excellent shape. I then bought two more from YosePower and they are about two years old. It’s these two I want to ask about. I’ll call them Y1 and Y2 to avoid confusion. They both charge up to 42v and show 41.9v when taken off the charger. Y1 then loses about .1v per day until it reaches 41.5v, then slowly to 41.3v after a week. I usually charge them up after a week or so. I haven’t messed with Y1. Y2 however, drops to 41.7v overnight and drops to 41.3v in two days. I thought it may have something to do with the Y2 BMS so I changed to a Daly BMS, but it made no difference. I then charged it up and disconnected the BMS and measured the cell groups daily. They all stayed at 4.0v and after a week all dropped to 3.9v. I realize that these may just be low grade batteries..BUT…since they are all connected, is it possible that some single cell in one group is weak and the rest of the groups are trying to charge it and slowly draining power? One other note. I tried buying one of those cheap active balancers but I have smoked two of them just plugging them in. Apparently this is a very big flaw with them and the manufacturer told me it can happen. Well gosh, I don’t want to keep taking that chance. So, if it is a weak cell, how to trace it? Any help always appreciated!
Battery troubleshooting-grab a coffee
- Thread starter Scott000
- Start date
Chalo
100 TW
My guess is that you have one bunk parallel cell group in both batteries, with a bad cell in the group dragging its mates down by self-discharge. If you open the packs after they've been fully charged and then sat around until they drop in voltage, you can probe each cell group with a voltmeter until you find the one that doesn't match. You can replace it (the whole parallel group) but probably not restore the cells you have now.
That is what I have been thinking. Unfortunately it involves disconnecting the cell groups and isolating the low group, then the cell or cells. Wish there was a more efficient way, but it is what it is.
Answer me this. If I bought a decent active balancer, and watched which group was always getting actively balanced, would that tell me in which cell group the weak cell(s) reside? Is that a valid test?
Chalo
100 TW
You don't have to separate cell groups, only measure between 1S series terminals. That won't tell you which specific cell is the culprit, but it will have beaten the poopoo out of its cellmates anyway and the entire parallel group should be replaced.
Ok, just so I understand.. I have the battery pack open and the BMS sense wire connector disconnected. I measure voltage of each group. But I get the same voltage in every group. 4.1v 6 places. A few days later I measure and all 6 places measure 3.9v. There is no group lower than the rest. That’s why I was wondering if one group had a weak single cell, if the cell wasn’t totally shot, could the other groups be gently making up for the voltage, and therefore drop the voltage of the entire battery slowly. Hope that makes sense. The battery is very usable, it just very slowly discharges when left off the bike and I check it.
Chalo
100 TW
Well, you should be measuring 10 groups, not 6. But if they're all dropping 0.2V per cell in unison, then that's either normal behavior for those cells with their age and condition (which seems doubtful), or the BMS is dragging them down pretty fast. If you leave them alone longer than a few days, do they continue to drop in voltage, even if it's at a slower rate?
Yes, they do drop in voltage and at the same rate after a few days, but let me go back and make sure I’m measuring where I’m supposed to be and document it. Be back in a couple of days!
lnanek
1 kW
- Joined
- Sep 26, 2022
- Messages
- 449
Maybe you could use an internal resistance measuring tool to measure each p-group and find which one has a bad cell. I'd also look for holes worn in the wrappers, leaks, corrosion, loose welds and shorts as well. Note that p-groups often zig zag through a battery, so you may have to alternate sides to test p-groups in isolation. That may be why you only see ~5 plates connecting the p-groups on each side of a 10s battery.
Re active balancers, I love how mine has a separate connector board. You wire everything up to the board, verify it's wired right with LED lights being even, then can slip the connector into the balancer diagonally so you can make sure the wires powering it from the lower p-groups connect first:
I use the bare board version, but there's fancier cases with LCDs and Bluetooth capability built around the same.
Re active balancers, I love how mine has a separate connector board. You wire everything up to the board, verify it's wired right with LED lights being even, then can slip the connector into the balancer diagonally so you can make sure the wires powering it from the lower p-groups connect first:
I use the bare board version, but there's fancier cases with LCDs and Bluetooth capability built around the same.
Yes! Could you please give me an example of an internal resistance measuring tool that you use? As you can tell, I’m not anything near a EE..I did not see anything corroded or leaking but that doesn’t mean there isn’t something there. Also, I need to explain how I ended up with 6 voltage measurements. The BMS has 11 sense wires, 10 and the negative. One of the groups is split between #1 and #11. So I was measuring 1-2, 3-4, 5-6, 7-8, 9-10, and 10-11. I should be measuring 1-2, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8-9, 9-10, and 10-11. I’m also on the lookout for a robust 10s active balancer. The cheap ones I bought were good for 1.2 amps and I think that’s what fried them. Some little burst of amperage just took out the board. Don’t think it was static, but anyway, I don’t trust them now!
P.S. and maybe I didn’t slip the balancer connector on correctly as stated….
I’m thinking, and when I think, I have to ask…..if I use a multimeter to measure the resistance in the separate groups, about what reading am I looking for, or, do I look for a reading that is off from the other groups, assuming there is only one group with one or more weak cells?
You can't directly measure cell or battery resistance with a multimeter. A multimeter measures it by placing a voltage across an unpowered resistance, and measuring the current, and doing some math to get the resistance.
Since your cells are voltage sources, the multimeter cannot do it's job, and may even be damaged or destroyed if the voltage is higher than that of the meter's output. (if it has a fused output, it may be protected by the fuse blowing, but no guarantees).
Since your cells are voltage sources, the multimeter cannot do it's job, and may even be damaged or destroyed if the voltage is higher than that of the meter's output. (if it has a fused output, it may be protected by the fuse blowing, but no guarantees).
To see what cells are problematic, then after your battery has lost voltage after being charged, then measure each of the 10 groups of parallel cells. You've only measured six groups, or rather, you've only measured six *places*, which may or may not even be different groups of cells (since we can't see exactly how you are measuring things, and you haven't described your complete testing process, step by step).
If you measure between each of the "balance wires" in the way you described above
"I should be measuring 1-2, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8-9, 9-10, and 10-11."
you should get 10 sets of voltages.
What are those, after you have charged the pack and let it sit until it's voltage has dropped as you usually see?
What are those, after you have ridden with the pack about half the distance you can usually get with it?
Any cell group that is different in voltage from the others is also different in characteristics. Which group or groups are problematic depends on what's being tested.
When you charge a pack, the worst cells get full *first*, and the better ones take longer to fill up, and during the actual charge process, when current is flowing into them, they rise in voltage faster.
When you discharge a pack, the worst cells get *empty* first, and during the actual discharge when current is flowing out of them, they drop in voltage faster.
If you find that all cells are dropping equally, then either the BMS is equalizing them trying to keep the dropping cells full, draining the good cells to refill the bad ones, or less likely all of the cells are equally bad, or very unlikely you could have one bad cell in each group that are all bad the same amount the same way.
If you measure between each of the "balance wires" in the way you described above
"I should be measuring 1-2, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8-9, 9-10, and 10-11."
you should get 10 sets of voltages.
What are those, after you have charged the pack and let it sit until it's voltage has dropped as you usually see?
What are those, after you have ridden with the pack about half the distance you can usually get with it?
Any cell group that is different in voltage from the others is also different in characteristics. Which group or groups are problematic depends on what's being tested.
When you charge a pack, the worst cells get full *first*, and the better ones take longer to fill up, and during the actual charge process, when current is flowing into them, they rise in voltage faster.
When you discharge a pack, the worst cells get *empty* first, and during the actual discharge when current is flowing out of them, they drop in voltage faster.
If you find that all cells are dropping equally, then either the BMS is equalizing them trying to keep the dropping cells full, draining the good cells to refill the bad ones, or less likely all of the cells are equally bad, or very unlikely you could have one bad cell in each group that are all bad the same amount the same way.
Similar threads
