Best and safe battery maintenance/care/charging practices

[mrelwood]

The question is asked often, but the answers get drowned in partially if at all relevant topic titles. So here are the best known EUC battery practices:

- On all wheels except S22 and V13, take care of balancing the cells at least every 5th charge. I do this every charge with great results: Charge until the charger shows green, plus 1-2 hours. Then unplug and let the BMS balance the battery for several hours to a full day before going for a ride. Charging to full is required for the cell balancing to start, resting is required for the balancing to continue. Unbalanced cells kill the battery fast and are a fire risk.

- Don’t charge a hot/warm battery. Let it cool down 15-30 min after a ride, and let it sit after charging before a ride.

- Don’t attempt to charge below +5•C, or if it’s very hot. Charge only at low amps when it’s below +10•C.

- Don’t store the wheel in a hot car, or in a direct sunlight.

- Don’t charge faster than 0.5C speeds, preferably lower. 5000mAh x 0.5C x __p4s = 10A, for example for the Master/S22.

- Store the battery (> a week or a few) at around 40-60% charge. A day or two at 100% every now and then is not a problem.

- Be mindful that the lower the battery, the less balancing power and top speed you have.

- Check the voltage after charge every now and then. If it deviates, investigate. If it doesn’t charge to full, take a small ride, charge until green (+ 1-2 hrs if at 100%), leave it be for several hours, repeat four times.

- If the charging behavior changes in any way, investigate thoroughly immediately!

 

 There are a few great battery related topics, such as this:

 

Did I forget something from the list above?

[meepmeepmayer]

2 hours ago, mrelwood said:

Don’t charge a hot/warm battery. Let it cool down 15-30 min after a ride, and let it sit after charging before a ride.

Is that actually a thing? Not something I was ever concerned about. But I don't know how warm the batteries get either. Is it because chargers are faster nowadays?

[mrelwood]

That’s what Battery University (IIRC) recommends for Li-ion batteries.

[Critzlez]

Could you explain why letting it rest after unplugging is required for balancing? I haven't heard of this being recommended, only that you should let it charge a couple hours after green

Also, instead of charging for a few hours after green, couldn't we just stop charging after the reported battery voltage is at max?

Edited July 10, 2023 by Critzlez

[Chriull]

1 hour ago, Critzlez said:

Could you explain why letting it rest after unplugging is required for balancing?

It's one of the two stages of balancing. In phase one (during charging) cells with voltages above 4.2V get the bleeding resistor in parallel so the charging current is less for these cells.

In stage two (after charging) cells with voltages above these 4.2V get discharged by the bleeding resistors down to 4.2V again.

(These voltage limits of 4.2V are slighty different to reach some hysteresis, etc)

1 hour ago, Critzlez said:

I haven't heard of this being recommended, only that you should let it charge a couple hours after green

Led becoming green is a very relative sign. Li ion cells should be charged downto a charging current of approx. 50-60mA. Some wheel & charger combinations come near this value, some don't.

As (the first stage of) balancing happens during this constant voltage stage of charging by the adjustment of the green led (or extending the charge duration) this balancing stage can be extended or restricted (with the according minor pros and cons)

1 hour ago, Critzlez said:

Also, instead of charging for a few hours after green, couldn't we just stop charging after the reported battery voltage is at max?

This is also known as 80% charging (stopping after the constant current stage leads to some 80-90% charge) and is not recommendable as it prevents all balancing. Except one monitors the single cell voltages on a regular basis, of course.

[mrelwood]

8 hours ago, Critzlez said:

Also, instead of charging for a few hours after green, couldn't we just stop charging after the reported battery voltage is at max?

I think you were after that once the wheel reports that the battery is at 100%, you’d unplug. That’s only when the Constant Voltage stage of the charging process begins. You can look up CC/CV charging.

A battery doesn’t act like a gas tank, that has a certain measurable amount of petrol in it. The way it was explained somewhere is that a battery has a surface voltage and a core voltage. When charging, the surface voltage changes faster, but it takes a longer time for the core voltage to get up to the same voltage. If you unplug right away when you reach 100%, you wait 5 minutes and it has suddenly dropped down to 97%. You need to stay at 100% for longer for the core voltage to get up there as well.

 For the same reason if you check the voltage immediately after a ride and again after 10 minutes, it can recover maybe even 5% during that time.

[The Brahan Seer]

@Critzlez

https://batteryuniversity.com/article/bu-409-charging-lithium-ion

https://batteryuniversity.com/article/bu-803a-cell-matching-and-balancing

Quality Li-ion cells have uniform capacity and low self-discharge when new. Adding cell balancing is beneficial especially as the pack ages and the performance of each cell decreases at its own pace. A problem arises when a cell in a string loses capacity or develops elevated self-discharge. This can be attributed to high-temperature spots in a large battery. Low-quality cells may also be prone to unequal aging. Li-phosphate has higher self-discharge that other Li-ion, and this complicates cell balancing

Passive balancing bleeds high-voltage cells on a resistor during charge in the 70–80 percent SoC curve; active balancing shuttles the extra charge from higher-voltage cells during discharge to those with a lower voltage. Active balancing is the preferred method for EV batteries, but it requires DC-DC converters. The corrected currents are in the mA range only. Applying a heavy load during acceleration, followed by rapid-charging with regenerative braking requires well-tuned cells in a high-voltage battery to attain the anticipated life. For cost reasons, EV batteries use mainly passive balancing.

All Li-ion cells require a protection circuit that assures that serially connected cells do not exceed 4.25V/cell (most Li-ion) on charge and that disconnect when the weakest cell drops to 2.80V/cell or lower. The discharge disconnect prevents the stronger cells from pushing the depleted cell into reverse polarity. The protection circuit acts like a guardian angel that shields the weaker siblings from being bullied by the stronger peers. This may explain why Li-ion packs for power tools last longer than nickel-based batteries without a protection circuit. The protection circuit also safeguards the battery from excessive load current

[Funky]

11 hours ago, Critzlez said:

Also, instead of charging for a few hours after green, couldn't we just stop charging after the reported battery voltage is at max?

My charger turns green right when it hit's ~84volts.

But i have noticed it still pulling around 40-50Watts from wall. (I got the device that shows watts from socket.) Over the next 2 hours it slowly goes down from those 40-50Watts to around 2.3Watts. At that point my wheel beeps slowly 6 times. Letting me know it has fully done charging/balancing the batteries.

See wheel was already done charging, when the charger turned green - 84V. (Not really.) But it still was balancing the cells.. We all know balancing happens at the end of charge. That's why the "extra" 1-2 hours - for balancing.

It's putting the 40-50Watts somewhere, in those ~2hours..  

Edited July 10, 2023 by Funky

[alcatraz]

39 minutes ago, Funky said:

My charger turns green right when it hit's ~84volts.

But i have noticed it still pulling around 40-50Watts from wall. (I got the device that shows watts from socket.) Over the next 2 hours it slowly goes down from those 40-50Watts to around 2.3Watts. At that point my wheel beeps slowly 6 times. Letting me know it has fully done charging/balancing the batteries.

See wheel was already done charging, when the charger turned green - 84V. (Not really.) But it still was balancing the cells.. We all know balancing happens at the end of charge. That's why the "extra" 1-2 hours - for balancing.

It's putting the 40-50Watts somewhere, in those ~2hours..  

I think it has to do with a resistance in the cell, explained above as surface/core charge, or constant voltage charging. The surface is 4.2v (84v) but internally they're still below that voltage, so they still pull current. Also, as the wheel ages this times becomes longer.

Then later: The charger should have some sort of trickle charge protection and I'm pretty sure that's based off the power draw, not voltage (so it happens after the light turns green). When the pack pulls below a certain wattage, the charger then switches off the supply. This is probably the time when charging is complete. Pity the light doesn't indicate this moment.

Who knows, maybe non standard chargers switch to green at this point. I'm pretty sure my charger doesn't have trickle charge protection. (which essentially makes it an 84v power supply, not charger)

Here's a question for you. If a charger came with a 3600Wh commander, and it is used to charge a hypothetical tiny 500Wh 100v mten. Isn't it possible for the trickle charge protection to kick in before balancing has completed?

[Funky]

13 hours ago, alcatraz said:

I think it has to do with a resistance in the cell, explained above as surface/core charge, or constant voltage charging. The surface is 4.2v (84v) but internally they're still below that voltage, so they still pull current. Also, as the wheel ages this times becomes longer.

Then later: The charger should have some sort of trickle charge protection and I'm pretty sure that's based off the power draw, not voltage (so it happens after the light turns green). When the pack pulls below a certain wattage, the charger then switches off the supply. This is probably the time when charging is complete. Pity the light doesn't indicate this moment.

Who knows, maybe non standard chargers switch to green at this point. I'm pretty sure my charger doesn't have trickle charge protection. (which essentially makes it an 84v power supply, not charger)

Here's a question for you. If a charger came with a 3600Wh commander, and it is used to charge a hypothetical tiny 500Wh 100v mten. Isn't it possible for the trickle charge protection to kick in before balancing has completed?

Don't know.. All i know mine charger is showing that it's done charging. But wheel alarms me that it's done only ~2hrs latter. Also Watts don't drop down anymore from those 2.3W. Meaning it stops "charging" around 2.3W.

After the beeps i have notice it's starting to drop voltage.. Because bluetooth is still connected. But same time it isn't charging anymore, even with connected charger. (You need to re-plug it out/in.)

To me it's funny - because its shows green light, but still are pulling ~50Watts. And only after those 2 extra hours it has done charging. (I'm using the very slow charger.)

[mrelwood]

23 hours ago, The Brahan Seer said:

Passive balancing bleeds high-voltage cells on a resistor during charge in the 70–80 percent SoC curve;

That’s not how any EUC BMSs function though.

 

22 hours ago, Funky said:

Over the next 2 hours it slowly goes down from those 40-50Watts to around 2.3Watts. At that point my wheel beeps slowly 6 times. Letting me know it has fully done charging/balancing the batteries.

See wheel was already done charging, when the charger turned green - 84V. (Not really.) But it still was balancing the cells.. We all know balancing happens at the end of charge. That's why the "extra" 1-2 hours - for balancing.

It's putting the 40-50Watts somewhere, in those ~2hours..  

Small details, but the BMS doesn’t report to the mainboard when it’s “done” balancing. And just like the whole charging process, there isn’t a specific moment when it’s “done”. It only depends on what you refer to with being “done”. And the actual balancing doesn’t use any power from the charger. Only that you charge the packs full enough for the process to start. Then it bleeds the cell groups down with a very tiny current.

Well done checking the charger’s power draw! It shows you the CV stage of the charging process very well as the charging current goes down. Some chargers turn green when the charging current goes below 700mA, some only at 100mA. It’s not standardized. The 1-2h rule should account for all chargers though.

 The charging current keeps getting closer and closer to zero, never reaching it, since there are losses in the system. You can keep pushing a tiny current into the battery for a very long time if you so decide. Again it doesn’t get filled like a gas tank does that is factually full when no more petrol fits in the tank.

 Same goes for the balancing process. If there’s a need for balancing, it is only “done” when all cell groups have been bled down to below 4.2V after several hours after disconnecting the chargers, and possibly repeated the process a few times to make sure that no group stays below that. The tips in the opening post are there to ensure that the balancing is sufficiently done.

[Funky]

21 minutes ago, mrelwood said:

That’s not how any EUC BMSs function though.

 

Small details, but the BMS doesn’t report to the mainboard when it’s “done” balancing. And just like the whole charging process, there isn’t a specific moment when it’s “done”. It only depends on what you refer to with being “done”. And the actual balancing doesn’t use any power from the charger. Only that you charge the packs full enough for the process to start. Then it bleeds the cell groups down with a very tiny current.

Well done checking the charger’s power draw! It shows you the CV stage of the charging process very well as the charging current goes down. Some chargers turn green when the charging current goes below 700mA, some only at 100mA. It’s not standardized. The 1-2h rule should account for all chargers though.

 The charging current keeps getting closer and closer to zero, never reaching it, since there are losses in the system. You can keep pushing a tiny current into the battery for a very long time if you so decide. Again it doesn’t get filled like a gas tank does that is factually full when no more petrol fits in the tank.

 Same goes for the balancing process. If there’s a need for balancing, it is only “done” when all cell groups have been bled down to below 4.2V after several hours after disconnecting the chargers, and possibly repeated the process a few times to make sure that no group stays below that. The tips in the opening post are there to ensure that the balancing is sufficiently done.

But what about the beeps? They happen normally 1-3 hours after charger has turned green. (More ridden, over multiples days not charging. It takes more time to get the beeps to happen. VS when i have done 40km in on day and charge again.) My guess more unbalanced batteries - takes more time to get them balanced and the beeps.

There are something happening on mainboard.. Otherwise there would be no beeps. And they only happen when meter drops to 2.3Watts from wall. Maybe they simply happen when wheel doesn't get any charge incoming? Or drop to low.

 

Edited July 11, 2023 by Funky

[The Brahan Seer]

40 minutes ago, mrelwood said:

That’s not how any EUC BMSs function though.

That quote is copy and pasted straight from the link at Battery University. 

[The Brahan Seer]

@mrelwood  Different wheels use different systems. Some use proper BMS others passive balancing, many none at all*.

https://www.dnkpower.com/li-ion-battery-pack-balance/#:~:text=Parallel cells will naturally balance,be balanced between cell groups.&text=Battery balance is an important,pack has its own BMS.

Surely if you have good quality new batteries you shouldn't have any issues if you follow charging guidelines.

How many people have problems with failing cells and hence balancing under normal use?

 

 

 

Edited July 11, 2023 by The Brahan Seer
*hopefully this isn't right. Must be some form of BMS.

[Chriull]

13 minutes ago, The Brahan Seer said:

That quote is copy and pasted straight from the link at Battery University. 

The 70-80% should be around the change from cc to cv. That's about were passive balancing starts. This could be the meaning of the battery university citing.

[The Brahan Seer]

@Chriull At 70-80% SOC it starts balancing and at 100% reaches the balanced state?

[Chriull]

9 minutes ago, The Brahan Seer said:

@Chriull At 70-80% SOC it starts balancing and at 100% reaches the balanced state?

Maybe, but not necessarily. It continues balancing after charging, if necessary:

Or it could need some of such balancing cycles...

Edited July 11, 2023 by Chriull

[The Brahan Seer]

@Chriull I've just reread your discussion below:

[The Brahan Seer]

5 minutes ago, Chriull said:

It continues balancing after charging, if necessary

This is where I get confused. I don't understand why once its reached 100% SOC it needs to keep balancing. Surely it has to be correctly balanced to reach 100% anyway.

 If the EUC display is accurate then 100% should mean 100% balanced? No? 

If any cells are damaged or failing then they would never reach 100%SOC or 100% charged on the display.

[Chriull]

14 minutes ago, The Brahan Seer said:

This is where I get confused. I don't understand why once its reached 100% SOC it needs to keep balancing.

100% SOC can be reached by 19 cells a bit below 100% and the 20th a bit overcharged to compensate the difference...

If one cell is degraded and has lower capacity the whole battery can only reach the same SOC ("capacity") if the degraded cell is overcharged - so the battery would have to be out of balance ...

14 minutes ago, The Brahan Seer said:

Surely it has to be correctly balanced to reach 100% anyway.

Who told you that an EUC battery reached 100% SOC. Especially "which SOC". The number the apps and wheel reports are very rough approximations and not usable during charging or riding.

Ninebots had some joule counter and could give "real" numbers.

14 minutes ago, The Brahan Seer said:

 If the EUC display is accurate then 100% should mean 100% balanced? No? 

Most EUCs have a "huge 100%" range - comparable to fuel gauges. They show already 100% before they are really full.

14 minutes ago, The Brahan Seer said:

If any cells are damaged or failing then they would never reach 100%SOC or 100% charged on the display.

Balancing tries to keep cells without regard to their degradation state (capacity, etc) at the same final charge voltage.

During use, once every cell depletes by the same capacity the weakest cell reaches the lowest voltage. By charging all cells again with the same charge they should reach the same voltage again!

As the weak cell during burdening and charging degrades a little bit faster as the others it needs some small balancing again to stay at the same final top voltage.

Until the differences get too big for the balancing circuit.

[The Brahan Seer]

24 minutes ago, Chriull said:

100% SOC can be reached by 19 cells a bit below 100% and the 20th a bit overcharged to compensate the difference...

 

This makes sense. So with a smart BMS this wouldn't happen as It knows the charge of each cell. But that doesn't help non smart EUC's. 

 

26 minutes ago, Chriull said:

Who told you that an EUC battery reached 100% SOC

100% SOC is when the battery is 100% full.  So you are saying EUC's never utilise their whole capacity?  So  it stops it reaching this figure using bleed resistors? or something like that. But still balances the cells. Ok i can understand this.

35 minutes ago, Chriull said:

Most EUCs have a "huge 100%" range - comparable to fuel gauges. They show already 100% before they are really full.

I understand they can show 100% when not at exactly 84V (for S18) but does at 83.8-9 which is still very close.

 

39 minutes ago, Chriull said:

Balancing tries to keep cells without regard to their degradation state (capacity, etc) at the same final charge voltage.

During use, once every cell depletes by the same capacity the weakest cell reaches the lowest voltage. By charging all cells again with the same charge they should reach the same voltage again!

This would show up very quickly as not reaching the voltage required over-all.

So balancing extra is only required if you don't hit 83.8V or above by the time the red light goes green?

Any issue with the cells would show up as a deficit? in voltage figure.

In normal use, using good quality, initially balanced cells and charging within the guidelines that haven't been subject to major crashes/heat or water should have no reason not to work well for x number of cycles?

 

 

 

 

[Robse]

On 7/8/2023 at 4:48 PM, mrelwood said:

- Store the battery (> a week or a few) at around 40-60% charge. A day or two at 100% every now and then is not a problem.

 

Please specify: Is it unhealthy for the battery to be stored at 100% charged?  1 day, 2 days, 3 days, 4 days, 5 days?   If you are not "allowed" to charge the wheel to 100%,  so that it is ready for the next time you want to ride, even if it is only in 3 or 4 days, then you end up in a situation where you always have to know 4-5 hours in advance, that you want to go for a ride. No one can do that?  Shouldn't I also charge my razor before I need to use it - if I only trim my beard every 4 days?

[Chriull]

12 hours ago, The Brahan Seer said:

In normal use, using good quality, initially balanced cells and charging within the guidelines that haven't been subject to major crashes/heat or water should have no reason not to work well for x number of cycles?

There are rumours that the most important thing for a good battery are well sorted good quality cells.

As addition to the 100% SOC discussed above - this state is defined by the manufacturer for a single cell beeing charged at some low constant current then 4.2V constant voltage until the charging current drops to ~50 mA.

In a battery with cells put in series and in parallel most cell sees a bit different voltages and currents as each cell is a bit different. The bleeding resistors are used to reduce a bit the charging current for cells with already higher voltages. After charging all cells with higher voltages are discharged to 4.2V again.

This passive top level balancing is a very simple solid system to equalize small differences with every charge. It works solidly, but is very soon overburdened once the cells deviate too much.

9 hours ago, Robse said:

  1 day, 2 days, 3 days, 4 days, 5 days?

Its every day a very very small bit. After 3 month at 100% charge at 60°C cells degrade to ~60%. At 25°C they degrade to 80% after a year. (1)

So still in a region were one can decide to not care about it. But if one wants good batteries after years of usage this could help a little bit.

(1) for some unspecified li ion cell according to https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries

Edited July 12, 2023 by Chriull

[Critzlez]

On 7/11/2023 at 3:44 PM, The Brahan Seer said:

So balancing extra is only required if you don't hit 83.8V or above by the time the red light goes green?

Yeah from my understanding, it should be fine to not balance an extra few hours after green if it reaches the max voltage when it changes to green ,right?

[Funky]

On 7/12/2023 at 1:44 AM, The Brahan Seer said:

So balancing extra is only required if you don't hit 83.8V or above by the time the red light goes green?

6 hours ago, Critzlez said:

Yeah from my understanding, it should be fine to not balance an extra few hours after green if it reaches the max voltage when it changes to green ,right?

I don't know.. Yeah mine charger turns green right when i hit 84V... But my wheel signals me only after extra 2 hours has been kept on the charger. That it has done charging, as i mention it before..

My charger is still pulling ~50ish Watts from wall, when it turns green. And over the next 2 hours it slowly by 0.01 Watts goes down to 2.3Watts..

I figure it still is charging? No?