Hah… so do you suggest that you found a brick charger which brings the current low enough for the bms to allow to balance the cells?
I assume this brings us to the thing that regular supplies some people use are not really intended to be used along with low-current bms’es and more suitable variant (like psu with low current capability) are desirable.
I was thinking about this… but it kind of is logical maybe, that these led light / regular psu’s were not meant to be powered in such low-power state… otherwise, maybe someone can come up with a solution how to fix this… and trick the charger to allow the current to flow…
A ‘‘parasite’’ load (resistor) maybe, which gets turned on only when voltage is high enough?
I didn’t. It just works. But based on what you’ve said I’m certain that that is your issue. You have a cc charger and not a cc/cv. I specifically looked for a cc/cv charger. That might help
There is nothing in the product description about cutoff current. In fact in the product description, it says
5amp constant current process and it does not mention a constant voltage process.
I looked at the charger I got from Battery Supports and there is nothing on the label concerning cutoff current
So, I got the board that I’m test charging down from 39v to 37.1 after a 3 1/2 mile ride.
I popped the lid and checked each cell for balance and this is what I found.
3.712
3.713
3.711
3.719
3.716
3.711
3.713
3.718
3.713
3.709
This is with discharging through the bms and includes some Regenitive braking.
I am recharging now with the Korad power supply
currently, my battery is at 38.6v and the power supply is in CC mode outputting 39.24v at 5a.
the voltage will increase as the battery’s voltage increases until it reaches 42v at which time it will automatically switch to CV mode holding at 42v with the current steadily decreasing until the battery is fully charged.
Charging with this power supply is like using a “Smart Charger”
My setup uses 12S lipos with a 60A BMS. I have never had a problem with them coming out of balance; throughout charging/discharging they are always within 0.01-0.02V of each other. I typically only charge to 48V (4.0V/cell) and discharge to around 45V (3.75V/cell), because this gets me around 12-13 miles. Note that when I charge to only 48V, the BMS balancing never actually occurs, but my cells always seem to end up balanced anyway
However today I wanted to test the lower voltage performance. I started at storage voltage of 3.85V/cell ± 0.1V and rode it to an (unexpected) cutoff. I thought I had plenty of life left, but suddenly everything cut out as I went to pass a cyclist uphill. Everything was dead, including the LEDs and voltage meters, which indicated it was BMS low voltage cutoff, not a VESC cutoff
I measured all the cells and found a huge voltage difference of 0.60V between the highest and lowest cell. Keep in mind they were within 0.01V at the start
3.60
3.45
3.20 !!
3.03 !!!
3.40
3.46
3.63
3.56
3.48
3.28 !!
3.34 !
3.45
So the BMS cut the power because cell #4 dropped below its 2.9V threshold during the high current load. My VESC cutoff was set to 37.2V (3.1V/cell) when the pack cut out at 41V
I charged at 45.8V with a 5A CC/CV charger. The voltage jumped up quickly from 41V to 44V and then slowly increased until the set voltage. Once it reached 45.8V, it continued charging for some time until cutting off.
But here’s the interesting thing… the pack rebalanced itself… seemingly without help from the BMS. I intentionally chose an end-of-charge voltage that was lower than the BMS balancing threshold. Yet the cells ended up nearly perfectly balanced, even though they were very uneven beforehand
3.82
3.81
3.80
3.80
3.80
3.80
3.80
3.81
3.81
3.80
3.80
3.80
I don’t know why the cells became so unbalanced after discharging past 3.7V/cell?
Nor do I know why they rebalanced themselves…?
Cool experiment,
Do you have means to measure the internal resistance of your cells?
It could be that the internal resistance is higher on some cells than others.
In this case that would be cell 4, that is why it charges and discharges faster.
This means that cell 4 has less capacity than your other cells. by keeping them in the 75-40% region you probably don’t see much of the effect, but once you go underneath this then it all goes crazy.
Take this with a grain of salt as this is mostly me making an educated guess
Check your nickel strips, one cell of pack 4 could have been detached.
That’s another example of bms not balancing at discharging, or maybe balancing only at low current, good for low discharge rate as in cruising.
@jmasta I’m extremely interested in runing a test that would change the idea of BMSs forever. If you won’t be able to do it I’ll do it anyways as soon as I receive my Korad power supply in two weeks.
The test would consist in charging a set of unbalanced packs with and without BMS intervention, wiring alteration would be necessary for the test, the used charger would be a cc/cv one.
If the pack is balanced wether you use BMS or not then, I’ll be irrelevant to use bms for charging.
Yeah @PXSS I think you’re right about the internal resistance. My 6S balance charger can measure the per-cell IR, but I would need to get to the original 3S balance leads to get an accurate measurement. They are pretty hidden away, so I’ll measure them the next time I take everything apart.
In the meantime I want to repeat this test, but pay more attention to the individual cell voltages throughout discharge. I need to find out when it starts getting crazy, and also to see if cell 4 will do this consistently
(BTW, that adjustable charger you found from China is awesome!)[quote=“Eboosted, post:110, topic:19302”]
@jmasta I’m extremely interested in runing a test that would change the idea of BMSs forever. If you won’t be able to do it I’ll do it anyways as soon as I receive my Korad power supply in two weeks.
The test would consist in charging a set of unbalanced packs with and without BMS intervention, wiring alteration would be necessary for the test, the used charger would be a cc/cv one.
If the pack is balanced wether you use BMS or not then, I’ll be irrelevant to use bms for charging.
[/quote]
I wouldn’t be able to do that test without rewiring my system. I intentionally wired it so the BMS must be installed for the DC charge port to be active. (The negative charge terminal is on the P- output from the BMS)
However if anything, I think this further validates the need for a BMS. Without one, I might have never known that one cell was dropping so fast. If I had ridden to my VESC programmed cutoff, that cell could have been damaged. Now imagine if you (unknowingly) did this every cycle
I think it’s largely due to the nature of LiPos. I’ve never actually taken these batteries below nominal voltage (44.4V or 3.7V/cell) before today, so I can’t say for sure without more testing
LiPos initially drop fast, then have very flat discharge line, and drop exponentially at the end (the dreaded lipo cliff). So small differences in resistance could have a large effect on the end voltage/capacity of each cell. Plus I am only running 12S1P, so each cell is on its own
Li-ion in general have a much more linear discharge curve, which I would suspect is more stable. In addition, you guys often have 3 or more cells in parallel. The parallel groups help balance out small differences among the cells. This is probably why li-ion packs can discharge to ~2.9V, whereas I need to cutoff my lipos at 3.4V
I hit the BMS voltage limit before the current throttling ever kicked in. That would have helped reduce voltage sag going up the hill, and I might have gotten more range before cutoff
Yeah for LiPos that makes total sense! I missed that part when I read your post last night. I thought you were using liions.
Lipos have about 10% remaining after 3.5V and yes, the internal resistance has a big effect on how fast they drop off after. Whenever we buy LiPos we request the manufacturer to send us matched cells so we never fly packs with unmatched cells. We buy Lipos from thunderpower though and they are a bit pricey.
Liions hit the cliff at ~3.1V but that by itself is more tame than the lipo cliff, also using quality cells from the same batch usually means a lower spread in internal resistance and lastly we use several cells in parallel which helps out average the internal resistance in case there is a weak cell.
Ok, so I charged my battery last night until the power supply output was down to around 50ma and the voltage meter on my board showed 42v, then immediately shut off and disconnected he bms to check cell voltages.
4.198
4.202
4.204
4.205
4.204
4.205
4.204
4.205
4.204
4.205
So then I plugged the bms back in and went to bed.
Today when I got home from work, I turned the bms on for a few minutes and the total pack voltage was at 41.9v
I turned the bms off and disconnected it to check cell voltages again and:
4.195
4.195
4.200
4.200
4.200
4.200
4.200
4.200
4.200
4.200
So it appears that the bms with its low balancing current needs ample time to balance the pack once its fully charged.
I would have liked to leave it on the charger longer to see if it would balance while charging the low cells but it got too late and I work the early shift.
It also appears that one of my packs may be a little slow though 5ma difference isn’t bad Imo.
I have it back on the charger now receiving 42v at currently 50ma to see if the low cells will balance up and wondering if the 4.2v cells will hold. I will update on that later.
Update:
So I left the pack on trickle charge (50ma) for about 2 hours and now the low cells (1and2) are at a full 4.2v
And now it’s time to go ride…
P.S. I think the extra 2 hours of trickle charge is totally unnecessary.
Simply charging down to 50ma the first time would have been plenty good enough.
I don’t understand what you mean by “then I plugged the BMS back and went to bed”
Were you charging with no BMS at the begining? Did you change the wiring and connected the BMS back in? I’m not following… Unplugging a BMS is not easy at all as you need to solder wiring
Did you measure the cell voltages before begining the charge cycle?
Check out my Lipo battery build, you will see that I have quick disconects on my balance wires for each 2s pack.
The bms was connected while charging then I disconnected the balance wires so I could easily check individual cell voltages.
Then I reconnected the balance wires and went to bed, it was late and I work the early shift.
The next day the cells where all balanced. I guess the bms was balancing them even though it was turned off.
I did measure the cell voltages before the charge cycle. those results where in an earlier post.
otherwise, maybe someone can come up with a solution how to fix this… and trick the charger to allow the current to flow…
