Bike works while plugged in

JaroMato

1 µW
Joined
Jan 13, 2022
Messages
4
So new here.

I have had an electric scooter for about 3-4 years now. Custom Made.
Here is what i know.
72V 30ah LIpo Newhop Battery Co.IMG_1342.jpg
the bike can go 40kmh with me on it.
bike was built on a 1981 honda trail 90 frame.
The BMS has been replaced once already.

Now i am having an issue where the bike will not work when connected directly to the battery. When the battery is connected and i plug the charger in than the wheel will spin.

When i turn the bike on when it is only on the battery power ...immediately the voltage drops to zero and than slowly climbs up. It does not work. The wheel does not spin, no lights. Battery also smells, and i noticed a brown spot on the BMS.

When i plug the charger in the bike wheel spins and the lights work.

I am a Battery newby. I will appreciate all help i can get.
I have a fluke volt meter.

Hopefully this is enough info to help.
 

Attachments

  • IMG_1343.jpg
    IMG_1343.jpg
    3.1 MB · Views: 451
  • IMG_1344.jpg
    IMG_1344.jpg
    2 MB · Views: 451
  • IMG_1345.JPG
    IMG_1345.JPG
    2.1 MB · Views: 451
  • IMG_1346.JPG
    IMG_1346.JPG
    2.5 MB · Views: 451
  • IMG_1347.JPG
    IMG_1347.JPG
    1.9 MB · Views: 451
What is the smell like?

Burned plastics?

Hot metal?

Acetone, or some similar chemical odor? Perhaps kind of sweet?

The last usually means leaking cells.

What was wrong with the previous BMS?

What was the specific failure and problem that led to that BMS being replaced?


If the BMS has separate charge and discharge ports (the BMS has a C- with the charge negative wire, a P- with the discahrge negative wire, and a B- to the cell main negative wire), then the specific problem you're having means that the BMS is "working", because it is allowing current flow all the way thru it if the charger is attached, but is disabling the output FETs when it's not (because of whatever problem it is detecting), so all the FETs work.

If it has a single charge/discharge port (C- and P- are the same port), then the BMS could be failed (discharge FETs failed) or it could simply be protecting the cells from overdischarge.

That could be the cells being badly out of balance, and one set is very low, or it could be a broken sense wire to a cell.

If cells are not balanced, especially if the smell is acetone-like, then one or more cells has probably leaked, and that could be from age, corrosion, physical damage, etc.


You can test for cell balance (and broken leads at the same time) with the Fluke. Set it to DCVolts. If it isn't autoranging, set it to 20VDC.

Black lead to B- of BMS. Red lead to B+ of battery; if it has this on the BMS use that, otherwise measure at the cell main thick positive (red) wire, or the main discharge port positive (red) wire. NOte this down as the main pack voltage.

Black lead to B- of BMS. Red lead to first balance wire closest to that. If the balance connector has a red wire and all the others are a single different same color, it's probably the wire at the opposite end from the red wire.

Note the voltage down as "cell 1".

Move the black wire to the next balance lead, and note down cell 2's voltage. It will read negative, don't worry about that.

Move the red wire to the next balance lead, and note down cell 3's voltage. It will read positive this time (just like the first cell).

Repeat this until you've measured all the balance leads.

List all the voltages you measured here.

The pack is probably 24s, meaning 24 groups of cells in series. If so, there will be either 24 or 25 balance wires. If there are 24, then you should read a voltage on every test done above. If there are 25, then the first test probably reads 0v (because the bottom balance lead and the B- are the same point in that case).

If the pack is fully charged (having been left on the charger long enough for the charger to indicate charging is finished, you should read about 3.6v on all the cells, if they are balanced. You should get around 86v for the main pack voltage.

If the pack doesn't appear to be charging either, then it's probably in the same state it was when it last worked (meaning if it was partly discharged, it's still like that).

If you get 0v on any balance leads, it probably means that lead is broken between the BMS and the cells. But it could mean the cell group is so damaged it has no charge at all.

If the pack is fairly empty, you'll probably read around 2.8v on each balance wire.

If there are differences between balance wire readings, especialy more than a tenth of a volt, the pack is probably pretty imbalanced. If there is more than a volt between readings, there are likely damaged groups.


Because the pack is made of 26650 cells, it has 24 sets of these that are probably connected together on positive and negative ends by nickel strips spotwelded to them, and then those sets connected in series to each other by more nickel strips.

For 30Ah, there's probably a minimum of 8 to 10 cells in each of those paralleled groups. If any spotwelds are broken or corroded, then those cels aren't connected to the pack any more, and that group is no longer 30Ah, and so the entire pack is now limited to the capacity of the smallest group. So it will not only run out of juice faster, it will also be harder for those groups to provide the current needed to run the scooter, and the scooter probably won't perform as well as it did before this type of problem happened.

The same thing happens if some cells are lower in capacity or otherwise damaged.
 
wow fantastic response!

I would call it a sweet smell.

I will have to digest the rest of your message before i respond to the rest of it.
 
JaroMato said:
I would call it a sweet smell.
Most likely a ruptured leaking cell. You'd have to open the pack itself up to check.

Can cells like 26650s don't typically leak easily; it usually takes significant mechanical stress to cause case failure enough to do that.

Corrosion could do it, which wouldn't be surprising given the rear wheel/shock condition, since it is tough to truly waterproof anything--usually when it's done insufficiently, it ends up doing more to seal the water that does get in in there so it can't get back out, than they do to seal the water out in the first place. :(

Overcharge or overdischarge (potentially sufficient overheating) could cause cell damage that eventually ruptures a cell.

Pouch cells rupture much more easily than can cells, especially since most "box" batteries dont' have any containment for them, or cell holders to protect them, etc.
 
so this is what i got.
voltage out of the main +/- wire 78.4V
voltage read from the red positive voltage wire.

3.267
6.535
9.80
13.07
16.34
19.61
22.88
26.14
29.41
32.68
35.95
39.21
42.48
45.75
49.02
52.29
55.56
58.83
62.09
65.36
68.6
71.9
75.2
78.4

I tried to get lower voltage numbers. They were all in this orientation. How can i determine what they are in the 3voltage range?

So when i add 3.267 + 3.267 = 6.534 and than continuing up adding 3.267 to each preceding number none were really far out of spec.

I could try plugging the battery in and charging it to full. However i suspect the bike will still not work. I really cannot remember why the last BMS was replaced.
 
JaroMato said:
so this is what i got.
voltage out of the main +/- wire 78.4V
voltage read from the red positive voltage wire.

3.267
6.535
9.80
13.07
16.34
19.61
22.88
26.14
29.41
32.68
35.95
39.21
42.48
45.75
49.02
52.29
55.56
58.83
62.09
65.36
68.6
71.9
75.2
78.4

I tried to get lower voltage numbers. They were all in this orientation. How can i determine what they are in the 3voltage range?
If you do the measurement as I described, moving the meter leads onto just each pair of balance connections, you will get the cell group voltage numbers directly. But since you have the numbers, you don't need to repeat the test.

So when i add 3.267 + 3.267 = 6.534 and than continuing up adding 3.267 to each preceding number none were really far out of spec.
If you want to find out what the values are for the numbers you actually have, you don't add, you subtract. Subtract the second-highest value from the highest, and that gives you the voltage of the cell group the highest number respresents. Continue doing this with each pair of numbers to get each cell group's actual voltage. The very lowest voltage will be the actual voltage of that cell group.

If you find that all of them are within a hundredth of a volt or so, then they may be well-balanced, however because 3.2v is the average voltage a LiFePO4 cell will read for most of it's discharge curve, it's not possible to say if they are balanced or not. They'd need to all be closer to full or empty to find that out. But it would mean that the BMS, if it is working, isn't shutting it's output off because of a low cell group.

If you find any that are significantly different from others, then that group is likely to be problematic for one reason or another. More investigation and pictures of things would be needed.

If you are measuring at the BMS PCB pads for the balance leads, and getting the voltages correctly there, then it does at least mean the wiring from cells to BMS is intact.

Note that cells could be leaking but still appear to operate normally, at least for a while, until electrolyte leakage drains enough from the cell to prevent it from doing so. If the problem leaves teh cell "open circuit", then other cells in parallel take up the load, leaving that group with less capacity and less current delivery ability, so they sag more in voltage than the rest under load. If they sag enough, the BMS detects LVC and shuts the pack off to prevent damage and fires.

Some BMSs renable the pack as soon as the laod goes away and the voltage goes back up, but some stay off until reset by connecting the charger.

If yours does reset by connecting the charger it means the BMS is likely operating correctly, but that there is likley a cell problem.

I could try plugging the battery in and charging it to full. However i suspect the bike will still not work. I really cannot remember why the last BMS was replaced.
If there is something wrong with the cells, recharging it could cause problems (rarely, dramatic ones, but it happens).

FWIW, the brown spot could be anything from coffee or soda/etc stain from anyone that previously handled it, to excess solder flux that never got cleaned off, to "goo" (electrolyte) from inside a cell or cells, or corrosion from an external or internal source, or even from an exploded FET.

If the cells are intact and working correctly, and the balance wires are as well, then the BMS could have blown the discharge FETs open, so that it can't output any current to the rest of the bike. This usually happens simply because of repeated overheating of the BMS inside the pack heatshrink and battery casing, or because more current was drawn from the battery for a longer time than it is able to handle (which also causes overheating), or because a voltage spike from regen braking damaged them.

If you can get good clear well-lit pics of the entire BMS on both sides (without the cardstock in the way of the bottom), with the aluminum cover on it unscrewed and removed (if it is removable and not epoxied in place), it may be helpful to diagnose potential problems it has. It may let us be able to describe tests you can perform on it to verify if it's failed or not.
 
so something interesting just happened. I was trying to turn the bike on and a spark happened on the blue wire. It was where the blue wire connected to the BMS. Than the characteristic smell...
it was where the left blue wire goes into the BMS. see photo
file.php


The heatsink on the BMS is epoxied on.

So i wasn;t able to get a photo of the BMS without the heatsink.

i was able to get the back side

the typing on the back says
bza25_65130_d16c16

maybe that helps?

thanks in advance!
 

Attachments

  • IMG_1372.JPG
    IMG_1372.JPG
    2.7 MB · Views: 295
Back
Top