bms Do Inmotion battery packs perform cell balancing? Yes!

[mrelwood]

4 hours ago, rcgldr said:

I used a small computer fan (AC Infinity AXIAL 8038, 7 watts) to blow air into the charger's fan opening.

I hope you checked whether the fan opening is the input or the output for the airflow. Don’t know about chargers, but some electronic devices only have an exit fan.

[rcgldr]

1 hour ago, mrelwood said:

I hope you checked whether the fan opening is the input or the output for the airflow. Don’t know about chargers, but some electronic devices only have an exit fan.

You're correct, it's an exhaust fan on one side, vent holes on the other side. I updated my prior post. The vent holes seemed like they might block flow from the external fan more than the single hole for the exhaust fan, so I tried the external fan on both sides, but I didn't see any significant difference. The charger maxes out at 84 volts, 5.00 amps, and at less than 0.30 amps, I assumed that any air flow would be better than none. The temperature got up to 37C, then decreased to 33C as the current went from 0.29 amps to 0.01 amps.  Normally I use that fan to cool an unvented trickle charger for my car.

https://rcgldr.net/misc/chrgfan.jpg

The eWheels description for their charger includes: "Reset button: once the selected charge % has been reached, the charger will automatically shut-off but remain in standby mode. When plugging the wheel back in, simply press the reset/power button, to restart the charge cycle again." However, the reset button just seems to reset accumulated amp-hours and watt hours in the display, so I power cycled the charger off|on to restart the charge cycle. The charger actually goes into standby when current drops below 0.30 amps. If the charger is set for 100% charge when it goes into standby, restarting it (power off|on) will just result in it immediately going back into standby mode. If the charger is set for 80% charge when it goes into standby, restarting it will continue a constant voltage charge, and current will slowly decrease down to a displayed 0.01 amps and then stay there. If the starting current is 0.30 amps or less, the internal fan does not turn on, and not knowing if this was intentional, I setup that small external fan so there would be some air flow.

Edited April 13, 2023 by rcgldr

[Chriull]

3 hours ago, rcgldr said:

The charger maxes out at 84 volts, 5.00 amps, and at less than 0.30 amps, I assumed that any air flow would be better than none. The temperature got up to 37C, then decreased to 33C as the current went from 0.29 amps to 0.01 amps.  Normally I use that fan to cool an unvented trickle charger for my car.

If the charger stops the fan below 0.3A its very plausible that no active airflow cooling is needed anymore. Especially if temperature stays below 37°C.

Switching mode power supplies are quite effective and do not need active airflow cooling at low amperage.

Electric circuitry works well at temperatures above "handwarm" without any problems.

Feasible designs allow up to 120°C for electric components like mosfets for elevated ambient temperatures without any problems assuming convection possible.

Edited April 13, 2023 by Chriull

[rcgldr]

4 minutes ago, Chriull said:

If the charger stops the fan below 0.3A its very plausible that no active airflow cooling is needed anymore.

The charger goes into standby when current transitions from above 0.30A to below 0.30A. Cycling power off and back on restarts the charger. If when it starts | restarts, the current is already below 0.30 amps, the internal exhaust fan doesn't turn on.

[rcgldr]

On 3/11/2023 at 8:27 PM, mrelwood said:

What baffles me in your situation is that the charging gets cut off at such high total voltage.

Have you ever used a rapid charger (like eWheels charger) instead of the stock charger on your V11? I'm wondering if the BMS on Inmotion wheels are relying on the characteristics of the stock charger to keep a pack a balanced, and that using a non-stock charger is leading to a pack imbalance the the Inmotion BMS can't recover from due to low discharge current (high resistance in bleeding resistors) and detecting overvoltage before there is time to balance the pack. 

[mrelwood]

1 hour ago, rcgldr said:

Have you ever used a rapid charger (like eWheels charger) instead of the stock charger on your V11?

No, I don’t see a point in exceeding 5A on the V11 or other similarly sized (or smaller) battery packs. 5A is the maximum manufacturer recommendation, and what the charging circuits, cables and connectors are spec’d for.

 Besides, the CV stage of the charge lowers the charge current just as it does with a low amp charger.

1 hour ago, rcgldr said:

I'm wondering if the BMS on Inmotion wheels are relying on the characteristics of the stock charger to keep a pack a balanced

What would these characteristics be? A charger provides a DC voltage, nothing else. Everything else happens in the BMS. And the balancing function in all EUCs except V13 and S22 is straightforward and simple: Cell group voltage over a threshold? Engage balancing resistor. There’s nothing the chargers can do to intercept that logic, except stop charging altogether.

Edited April 21, 2023 by mrelwood

[rcgldr]

11 hours ago, mrelwood said:

What would these characteristics be?

The maximum current and voltage. On my V8F, the stock charger is 1.5 amps, and with eWheels rapid charger I can use 1.0, 2.0, or 3.0 amps. Maximum voltage might be 84.2 volts on eWheels charger, and possibly 84.0 volts on stock charger. The eWheels charger will stop charging altogether once current drops below 0.3 amps. With the eWheels charger, after it stops charging and goes into standby, power cycling it off|on can continue an 80% (80 volt) constant voltage charge with current decreasing to an indicated 0.01 amps, but I haven't had this work on 100% charge yet. This could be due to V8F shutting off charging due to overvoltage

I tried switching to stock charger after a short ride, but in 5 to 10 minutes after stock charger shows green (less than 0.4 amps), pack voltage has reached a peak, reported voltage 84.0, and slowly declines at 0.3 to 0.4 volts per hour, but I've only left it connected for about 2 1/2 hours, so I don't know if the voltage decrease would stop once all cells above 4.15 volts were discharged down to 4.15 volts.

Edited April 22, 2023 by rcgldr

[rcgldr]

On 3/11/2023 at 8:27 PM, mrelwood said:

Another explanation could be that it’s the balancing resistors that are bleeding the voltage down to 83.0V (+/- measurement error). 83.0V / 20 = 4.15V, which does sound like a plausible lower balancing threshold voltage.

Have you tried to start charging again after the voltage drops down to 83.0V? 

I tested a long charge with my V8F. Stock charger LED transitions from red to green at 83.7 to 83.9 volts (400 mA), and about 8 minutes later, reaches a peak of 84.1 volts for about 20 seconds, where apparently overvoltage of some cell group is detected and charging stops, then voltage slowly declines until it drops below 82.7 volts, and it charges back up to 84.0 volts, a repeating cycle. I have two stock chargers, and the pattern is probably the same for both stock chargers, but I only tested for 3 hours with my other stock charger.

elapsed time - reported voltage (using EUC World):

00:00 - 84.0 - charging disabled, voltage declines

01:00 - 83.5

02:00 - 83.3

03:00 - 83.1

07:00 - 82.7 

08:10 - 83.3  - charging re-enabled, charger led changed to red

08:30 - 84.0 - charging disabled, voltage declines

09:30 - 83.5

12:30 - 83.1

15:00 - 82.8

17:00 - 82.6

17:10 - 83.3 - charging re-enabled, charger led changed to red

17:30 - 84.0 - charging disabled, voltage declines

So apparently once voltage goes below 82.6|82.7 volts it re-enables charging and charges back up to 84.0 volts, then voltage starts decreasing again. I don't know if there is any balancing going on.

In this test and prior tests, there is only 8 minutes between the time the stock charger LED changes from red to green, when current drops below 400 mA, and voltage only increases by 0.1|0.2 volts during that 8 minutes, so the current probably never gets much less than 300 mA.

Edited April 25, 2023 by rcgldr

[rcgldr]

On 3/10/2023 at 6:17 AM, RagingGrandpa said:

It's resistive top-balancing.

It's not resistive top-balancing in the normal sense, where once a group of cells reaches some voltage, a bleeding resistor discharges that group while other groups continue to charge. In order for this to work, discharge current has to be greater than the charging current going to the pack. Instead my V8F reacts in the same manner as described in Wrong Way's video. Some group(s) of cells reach some maximum voltage as pack reaches a short peak of 84.1 volts, charging stops, followed about 8 hours of a slow discharge which is probably due to the small amount of current used to drive blue tooth and other circuitry. Once the voltage drops below 82.7|82.6 volts, charging is enabled again, the stock charger LED changes to red indicating over 400 mA current, and about 12 minutes charger LED changes to green indicating less than 400 mA current, then about 8 minutes later, group(s) of cells reach some maximum voltage as pack reaches a short peak of 84.1 volts, charging stops, and the 8 1/2+ hour cycle repeats. 

The resistors on the V8F|V10F BMS are stated to be 1000 ohms, which translates in to a maximum of 4.2 mA discharge current, while the stock charger is probably around 300 mA when overvoltage detection stops charging. By comparison, a typical e-bike has a balancing discharge current around 200 mA (the charging current would have to be less than 200 mA for this to work).

[RagingGrandpa]

11 hours ago, rcgldr said:

In order for this to work, discharge current has to be greater than the charging current going to the pack.

Not actually.

Engaging a "bleed" resistor during recharging will reduce the charging current seen by that one cell group. (Because some of the charging current will bypass the cells by flowing through the resistor.) So the result is: the cells in parallel with the resistor will see a slightly lower recharging current than the rest of the cells in the pack, and this allows the other cells to gradually 'catch up' to the voltage state of the highest one. 

Doing this during the CC stage of recharging is usually not productive, because the variations in cell resistance cause temporary and artificial voltage mismatch between cell groups, as a function of the amount of current flowing. So, top-balancing systems wait until the end-voltage (CV stage) to engage resistors, where recharging current is falling naturally to low values and the resistive voltage drop across cells is negligible.

[rcgldr]

On 4/27/2023 at 7:29 PM, rcgldr said:

It's not resistive top-balancing in the normal sense, where once a group of cells reaches some voltage, a bleeding resistor discharges that group while other groups continue to charge. In order for this to work, discharge current has to be greater than the charging current going to the pack.

 

On 4/28/2023 at 7:02 AM, RagingGrandpa said:

Not actually.

"In order for this to work" - in order for a bleeding resistor to discharge a group while other groups continue to charge. There are other types of balancing. An EUC can also balance a pack during discharge (some smart BMS do this). 

On 4/28/2023 at 7:02 AM, RagingGrandpa said:

Engaging a "bleed" resistor during recharging will reduce the charging current seen by that one cell group. (Because some of the charging current will bypass the cells by flowing through the resistor.) So the result is: the cells in parallel with the resistor will see a slightly lower recharging current than the rest of the cells in the pack, and this allows the other cells to gradually 'catch up' to the voltage state of the highest one. 

From my 18+ hours of V8F charge test (starting at 83.9 volts), the charger is in constant voltage mode the entire time. There is only 20 minutes of charging about every 8 1/2 hours of which 12 minutes is over 400 mA (LED red), and 8 minutes from 400 mA (LED green) down to a bit below 300 mA. At 300 mA, with 1000 ohm resistors bleeding a maximum of 4.2 mA of current, at 300 mA, that's 300 mA without resistor, 295.8 mA with resistor, and not enough to prevent a cell group from reaching overvoltage which causes the V8F to shut off charging for 8 hours until pack voltage drops down to a bit less than 82.6 volts, before starting another 20 minute charge cycle. Reported voltage will jump from about 82.6 volts to about 83.5 volts at the start of a 20 minute charge cycle due to current from charger. If other cells are catching up, the current would have to be coming from within the battery, as the V8F cuts off current from the charger, other than just enough to detect there is voltage coming from the charger. I've confirmed this with an eWheels charger that displays 0.00 amps when the V8F is in this mode.

The V8F could be balancing via current from within the battery when charging is shut off and there is a small discharge current from the pack used to keep the V8F "awake" (blue tooth and other circuitry), but that is not resistive top balancing.

The eWheels charger will shut down charging once current drops below 300 mA, and the V8F will shut down soon after, which I assume would stop a discharge based balancing. The stock charger doesn't shut down, which keeps the V8F awake despite it shutting off charging. I can duplicate this behavior with the eWheels charger after it shuts down by switching to stock charger, waiting for V8F to stop charging, then quickly switching back to the eWheels charger where it will display 0.00 amps, and the V8F will remain awake.

Edited April 29, 2023 by rcgldr

[rcgldr]

On 4/28/2023 at 7:02 AM, RagingGrandpa said:

Engaging a "bleed" resistor during recharging

If the eWheels charger is set to charge to 80% or 90%, then after it goes into standby, I can briefly use the stock charger, then switch back to the eWheels charger, which returns into constant voltage mode but doesn't shut off due to current below 0.3 amps, and the displayed current slowly decreases to 0.01 amps, which I assume means 0.005 to 0.015 amps). So it only takes about 1 amp of current to keep my V8F awake and pack voltage constant. This would imply the discharge current after my V8F disables charging but remains awake is less than 0.015 amps, and the 0.04 amps discharge related to 1000 ohm resistors could be used to balance the pack during the slow discharge. I have no way to know if this occurs. I didn't see any increase in 100% charge pack voltage during the 18+ hour charge test, but my pack may already be balanced or close to balanced.

Edited April 29, 2023 by rcgldr

[RagingGrandpa]

Don't forget- after saturating at 4.20V, the NMC cells will naturally stabilize around 4.16V/cell when totally open-circuit and no bleed loads connected, within about an hour. 
That's 83.2V for your 20S pack, with 0mA parasitic load.

 

[v8nice]

On 3/10/2023 at 4:01 PM, lagger said:

I think that it should have BMS. I hate this controversy.
The community is actively discussing this for year while all it takes is someone to analyse the PCB of controlboard and especially google search the IC used around these resistor arrays. Yes the layout doesnt look so distinctive and the resistors used are quite small but it could be. All it needs is this and instead of doing it everyone is just keeping the fire of this confusion. I wont be openning my pack but I am certain there are people with packs available and it needs just like 10 minutes of research to get the IC name and search for it.
Anyway I looked on some photos of V5F V10F and all have the resistor arrays. What was Adam showing is just back side.

here we show you proof tested by a professional, there is no balancing of the cells, despite some implanted components. Inmotion said it himself, adams full of people who have been able to open the batteries of their wheel which no longer have any  .autonomy or full charge have seen the imbalance of the cells. Me even on several v8 v5f v8f batteries. There is no possible excuse, Inmotion makes very bad batteries, and almost systematically on all models the voltage calibration is completely  fake.

[lagger]

4 hours ago, v8nice said:

""

Thanks for thinking about me when you found that thread , finally some hard data. I cant believe Inmotion did this on one hand. On other hand I cant believe how well the pack works without any balancing.
Obviously this is not a good engineering design - pretty much a flaw.

I think this is something also for @mrelwood regarding the top charging to always balance the pack and videos regarding balancing of inmotion.

[mrelwood]

 

5 hours ago, lagger said:

On other hand I cant believe how well the pack works without any balancing.
...
I think this is something also for @mrelwood regarding the top charging to always balance the pack and videos regarding balancing of inmotion.

I can’t wrap my head around it either. How come all the V10Fs seem to be doing just as well battery wise than other wheels? My 16S battery died at 4k km despite balancing every 10th charge. Based on other wheels, non-balancing packs should be dying at 2k - 4k. Why aren’t they? And why do other wheels die so fast if the balancing is disabled by only charging to 80%? It doesn’t make any sense.

I hope someone knowledgeable enough gets to examine the V11 and V12 BMSs as well.

[v8nice]

On 3/11/2023 at 5:38 PM, mhpr262 said:

Did anyone truly think they didnt have a BMS/cell balancing  system? It is just unthinkable and impossible to build a big battery pack like in an EUC, with so many cells connected in series, without some form of battery balancing. It would be a recipe for disaster otherwise. Li-ion cells dont work like the old Ni-Cd or NiMH cells where you safely overcharge some cells a bit to let the others "catch up".

seeing this kind of message, I wanted to answer it. Inmotion did the unthinkable. Make large 84V batteries without including the components for balancing.  see on my wheels, that of my friends, problems with the battery which does not charge to the maximum (between 83.5 and 84v), batteries which charge less and less, or the autonomy which decreases, see the wheels which raise the pedals to  low speed with 30/40/50% battery left on the app.

[lagger]

6 hours ago, mrelwood said:

 

I can’t wrap my head around it either. How come all the V10Fs seem to be doing just as well battery wise than other wheels? My 16S battery died at 4k km despite balancing every 10th charge. Based on other wheels, non-balancing packs should be dying at 2k - 4k. Why aren’t they? And why do other wheels die so fast if the balancing is disabled by only charging to 80%? It doesn’t make any sense.

I hope someone knowledgeable enough gets to examine the V11 and V12 BMSs as well.

My question is if we really have enough data regarding balancing that the wheels are really dying solely because of balancing every 10th time. I am skeptical. I think a good factory balanced pack (capacity, internal resistance) could go well for some time before some cells start to drift out too much. However I am also very surprised that I made 8 k kilometers and it seems like the wheel still works fine. Thats more that I would ever expect.

[v8nice]

that's exactly it. When the cells incorporated are identical and come from the same batch, even without balancing, the battery will remain very balanced, even for quite a long time. On the other hand, it seems that not only do they not balance, but also not always batteries that come from the same batch... and the problem is obvious, the battery will lose capacity more or less quickly .

[mrelwood]

12 hours ago, lagger said:

My question is if we really have enough data regarding balancing that the wheels are really dying solely because of balancing every 10th time. I am skeptical.

And for a good reason. We don’t have that data. While my first pack died at 4k and the second at 8k, it’s still just one (or two) cases. But there have been several posts on this forum from riders that have only ever charged to 80%. And their packs died at 4k, 3k, even at 2k. Not an awful amount of data, but definitely a pattern.

12 hours ago, lagger said:

I think a good factory balanced pack (capacity, internal resistance) could go well for some time before some cells start to drift out too much.

Absolutely. But since the cells operate at different temperatures around the pack, some drifting is bound to happen.

12 hours ago, lagger said:

However I am also very surprised that I made 8 k kilometers and it seems like the wheel still works fine. Thats more that I would ever expect.

Li-ions are good for hundreds of full capacity cycles. V10 has a 650Wh battery, a rough estimate for the mileage might be 35km. Since EUCs don’t utilize the whole capacity, we’ll round it up to 40km. 8k is roughly only 200 charge cycles. IIRC, most batteries state 400-500 cycles before reaching 80% capacity. So you’re only halfway there, after which you could happily continue with the 80% (and declining) capacity.

 As long as the cells stay sufficiently balanced of course.

[rcgldr]

I'm wondering if the relative charge | discharge rates are similar. For example, a cell group that charges and discharges faster than the other parallel cell groups, so that the net result is cell groups take quite a while to drift apart, if the cells in a pack are reasonably close to each other.

Cells within a parallel group are always self-balancing. It's only the groups in series that need balancing. Based on that, there's probably less drift with a 6P 4P pack than there is with a 2P pack. 

Edited August 6, 2023 by rcgldr

[Chriull]

37 minutes ago, rcgldr said:

I'm wondering if the relative charge | discharge rates are similar. For example, a cell group that charges and discharges faster than the other parallel cell groups, so that the net result is cell groups take quite a while to drift apart, if the cells in a pack are reasonably close to each other.

Closeness of packs affects the cells mostly (just) by temperature influence - most processes of cells are heavily influenced dependent on temperature. Like for example degradation in a hotter environment.

37 minutes ago, rcgldr said:

Cells within a parallel group are always self-balancing.

Short circuited together should express their relationship better - their voltage is always exactly the same. Beside some negligible losses along the nickel strips.

37 minutes ago, rcgldr said:

It's only the groups in series that need balancing.

Yes, the more cells in series the more more easily they'll drift apart.

37 minutes ago, rcgldr said:

Based on that, there's probably less drift with a 6P pack than there is with a 2P pack.

The only problem with paralleled packs is that once one has to much self discharge/internal short circuits by plating the other paralleled cells "help" to self ignite this cell.

Both parallel and series circuit of cells have their very own advantages, disadvantages and dangers...

Edit: PS: in wheels with not too much cells in parallel single cells were more easily overburdened and by this faster degraded. By this higher p count can give batteries  a longer live by less burden.

Edited August 5, 2023 by Chriull

[mrelwood]

7 hours ago, Chriull said:

Edit: PS: in wheels with not too much cells in parallel single cells were more easily overburdened and by this faster degraded. By this higher p count can give batteries  a longer live by less burden.

In many modern wheels a 30s4p pack can have the battery split in four physical entities, making it four separate 30s1p packs. So the cells don’t enjoy the benefits of the parallels keeping each other in check.

[rcgldr]

10 hours ago, mrelwood said:

In many modern wheels a 30s4p pack can have the battery split in four physical entities, making it four separate 30s1p packs. So the cells don’t enjoy the benefits of the parallels keeping each other in check.

For example, both V10F and 18L have 80 cells, V10F is 1 x 20S4P, while 18L is 2 x 20S2P. 18XL is 120 cells, 2 x 20S3P.

Edited August 6, 2023 by rcgldr

[mrelwood]

4 hours ago, rcgldr said:

For example, both V10F and 18L have 80 cells, V10F is 1 x 20S4P, while 18L is 2 x 20S2P. 18XL is 120 cells, 2 x 20S3P.

Yes. Both very unmodern wheels though.