The 80% charging "rule"

[null]

Not to nitpick here, conclusion remains the same, but:
According to e-wheels the 18XL uses Samsung 35E, who are apparently rated for 0.6C charge current.

Edited February 17, 2021 by null

[WI_Hedgehog]

Reminder: It's not just what the cells can take, it's also the heat the balancing circuitry can/can't dissipate.

[Planemo]

3 hours ago, null said:

According to e-wheels the 18XL uses Samsung 35E

Wow I am surprised at that. I thought just about all wheels out there used a higher output cell than the 35E. Indeed, I would think most folks would sacrifice a tiny bit of range for a stronger output cell.

The 35E is only 8A as opposed to the bread and butter Sanyo GA's (MSX for example) 10A. So in 6P you have at your disposal 48A with 35E's and 60A with the GA's. Thats not an insignificant difference IMO. I guess the lower speed of the 18XL lets them trade off a bit of power for a little more capacity.

[Hangman]

On 2/9/2021 at 12:34 PM, Rich Sam said:

Thanks @mrelwood, yeah that is how they sold me a fast charger! hehe.  But after some extended research into the topic I quickly learned about balancing and how it may not be the best choice to not charge to 100%.

Ultimately I'm not salty about the purchase of the charger as I still have a fast charger, and I can also take one to work and one at home.

I feel this info should be stickied somewhere in a maintenance section!

Perhaps I have missed something as I understood the main purpose of the fast charger was to take the battery to 80 or 90% quickly.  From other posts I've read, the standard charger slows down as it nears 100% anyways.?.  So I take the battery to 80 or 90% using the fast charger, then unplug.  Then a few hours before a long ride, I plug in the standard charger & slow boat it up to 100% & balance.  On my short stints, I don't bother with the 100% as the fast charge to 90% is sufficient.  I get somewhere between 3 & 5, sometimes 6 short stints before I get the chance for another gratifyingly long, smooth ride.

Would this then be considered bad practice?

[RagingGrandpa]

36 minutes ago, Hangman said:

On my short stints, I don't bother with the 100%

I get somewhere between 3 & 5, sometimes 6 short stints

Balancing happens after the EUC reports 100%.

Your habit is bad because you're only allowing it to happen once every 3 to 6 rides.

I think it would be better if you were to use a standard 'slow' charger in a simpler way, as described below:

On 5/6/2020 at 11:41 AM, RagingGrandpa said:

"Charge overnight before a ride"

Don't charge after a ride, unless you're going to ride again immediately. (Minimizes storage SOC.)

I truly think this is all that would be needed for most EUC owners to keep their packs happy. If they knew this and nothing else, I think we'd be in a better place than today.

[Mono]

12 hours ago, RagingGrandpa said:

Your habit is bad because you're only allowing it to happen once every 3 to 6 rides.

Why is that bad?

[mrelwood]

31 minutes ago, Mono said:

Why is that bad?

The balancing current on EUCs is very low, and if one unplugs soon after reaching 100%, which is already every sixth or so charge, there’s not much balancing happening. The packs may very well be out of balance already when new.

I religiously balanced my 16S every 10th charge, yet the first pack failed at 4000km (at roughly 80 full charge cycles), the other pack at 8000km. In both cases a cell group had already reached 0V.

The second failure caused a nasty crash when I loaned the wheel to my friend. We knew that the battery was badly balanced and that he mustn’t accelerate fast, but the 16S failed to keep up at a medium acceleration by a 70kg rider. He busted his face and required a few stitches. No full face at that specific time, since he was just on a very easy low speed cruise with his wife.

 The minuscule amount of harm done by charging every time to 100% + balancing, is nothing compared to the risk of not balancing enough.

[FinRider]

This battery topic has been discussed so many times and everyone seems to have a different opinion on whats "best" or "healthiest" for our beloved EUC's.

I think we can all agree that the worst that can happen is for a pack to be "destroyed", be it not taking a charge anymore or, god forbid, catching fire.

IMHO we should not make this into rocket science. We tend to upgrade our wheels way faster than the battery degrades anyways.

During my riding season (more than +5C) I probably ride every day, often on very short notice. So I let the wheel "cool off" for about 30 minutes after a ride and then I plug it in to the charger and leave it there. I will unplug the wheel when I see the light go green which sometimes is just when it has reached full and sometimes 8+ hours after it has turned green.

 

 

 

 

[Planemo]

2 hours ago, FinRider said:

This battery topic has been discussed so many times and everyone seems to have a different opinion on whats "best" or "healthiest" for our beloved EUC's.

Theres only one 'best' and thats balancing every time. Anything else will cause an imbalance. That's not opinion, it's fact.

Someone can feel free to challenge that if they want, but they will still be wrong. They may bring in the argument of 'small imbalances are OK' to which I would reply well yes they are but its not 'best' nor indeed 'healthiest'. The bigger the imbalance, the harder it is to get rid of it, to the point where in some cases it can't even be done anymore by the factory charger, no matter how long you leave it plugged in. If you are lucky, you may be able to save the strings by charging them individually outside of the wheel but it's a major faff at best.

2 hours ago, FinRider said:

IMHO we should not make this into rocket science. We tend to upgrade our wheels way faster than the battery degrades anyways.

Not rocket science, no, but @mrelwood's experiences show that not only can some expensive packs be ruined, but serious injuries can result. And contrary to your statement, both happened well before any wheels were due to be upgraded. His comments are very viable and an important aspect to re-iterate, no matter how many times battery talk gets brought up in conversation on the forum.

 

 

[null]

19 hours ago, Planemo said:

Theres only one 'best' and thats balancing every time. Anything else will cause an imbalance. That's not opinion, it's fact.

 if you’re bored you can debate that over at endless sphere, where I got attacked for wanting to charge with a charger and not a PSU. Apparently their battery packs are so great they only need balancing every 100-200 cycles. Good for them, ours aren't an ideal hobby battery in a man cave. (Note I never asked to debate about PSU charge or balancing.

Edited February 19, 2021 by null

[Hangman]

29 minutes ago, Planemo said:

Theres only one 'best' and thats balancing every time. Anything else will cause an imbalance. That's not opinion, it's fact.

The bigger the imbalance, the harder it is to get rid of it, to the point where in some cases it can't even be done anymore by the factory charger, no matter how long you leave it plugged in.

Thank you for the clarification, I really appreciate it.  I simply need to change my practices.

[Hangman]

As I was perusing the forum I realized something I'd been neglecting in this battery chat.  How much Regen occurs when casually riding?  If you charge to 100% all the time, then take the wheel out & get 120 yards down the street & need to slow down & idle along behind some stroller, how much regen are you pushing back to a 100% battery & isn't the regen injuring the battery at that point?

When I first purchased my V8 over a year ago it was described to me (I have to paraphrase as I don't remember the exact wording of the conversation) that unless you are going on a long ride, you really don't want to charge your battery to 100% because the regen will put a lot of pressure on the batteries.

I've read somewhere that the mainboard or BMS bleeds excess recharge off so you are losing energy but it can only bleed off so much within a certain amount of time. 

I can only assume the larger, more powerful wheels can regen better than the smaller wheels like a V8 but I honestly have no idea.  I've always like the concept of regenerating the battery but have not studied it indepth.

[Planemo]

27 minutes ago, null said:

 if you’re bored you can debate that over at endless sphere, where I got attacked for wanting to charge with a charger and not a PSU. Apparently their battery packs are so great they only need balancing every 100-200 cycles. Good for them. (Note I never asked to debate about PSUs)

I think I'll pass on that one my first thought though is that very few ebikes run the sheer amount of cells we do...nor the voltages either..

[WI_Hedgehog]

21 hours ago, Hangman said:

Perhaps I have missed something as I understood the main purpose of the fast charger was to take the battery to 80 or 90% quickly.  From other posts I've read, the standard charger slows down as it nears 100% anyways.?.  So I take the battery to 80 or 90% using the fast charger, then unplug.  Then a few hours before a long ride, I plug in the standard charger & slow boat it up to 100% & balance.  On my short stints, I don't bother with the 100% as the fast charge to 90% is sufficient.  I get somewhere between 3 & 5, sometimes 6 short stints before I get the chance for another gratifyingly long, smooth ride.

Would this then be considered bad practice?

Your reasoning is correct and cells will be fine with that. You're not over-heating the cells nor over-straining the balancing circuitry during fast charging, you're using the normal speed charger during the balancing phase, and you're not running the packs down enough to hit minimum voltage on any of the [newer] cells. As the cells age this might become an issue, but for now you should be fine.

Do note with fast-charging the first 80% (constant current phase) goes faster than normal, but the last 20% of the charge (constant voltage phase) and balancing goes slower. (Time is saved over-all.)

[RagingGrandpa]

Balancing is interesting. More discussion below.

https://forum.electricunicycle.org/topic/18374-how-bms-works-gotway-pack-teardown/?do=findComment&comment=357293

[Mono]

9 hours ago, mrelwood said:

I religiously balanced my 16S every 10th charge, yet the first pack failed at 4000km (at roughly 80 full charge cycles), the other pack at 8000km. In both cases a cell group had already reached 0V.

Thanks for sharing, I wasn't aware of these package failures. How did you determine that these failures would have been (likely) prevented by balancing the cells every charge? I assume we agree that packages can fail even when they are brought to 100% and balanced during each recharge and the underlying reason for the failure is not likely to be related to balancing?

Edited February 18, 2021 by Mono

[Chriull]

9 minutes ago, Mono said:

Thanks for sharing, I wasn't aware of these package failures. How did you determine that these failures would have been (likely) prevented by balancing the cells every charge? I assume we agree that packages can fail even when they are brought to 100% and balanced during each recharge and the underlying reason for the failure is not likely to be related to balancing?

Hi!

Harshly said one cannot conclude anything from such experiences (sorry @mrelwood). As they are just "anecdotal evidence/single cases", statisticly most likely not relevant, the starting state of the individual cells unkown, etc...

It is just to be hoped that the implemented balancing circuitry has some effect and if used could delay or even prevent such misbalancing.

So faults occuring after such not intented use (80% charge) lead to not recommending such usage anymore.

I have not seen till now any considerations/paper/tests in regard to effectivenes of such passive balancing. Especially in combination with cell matching. So until which amount of mismatch (relative capacity loss due to aging and or change of other characteristics) such passive balancing can work effectively or at least delay pack deterioration.

Just the more cells are in series the more "unstable" the system becomes.

There are regularly emotional discussions on electronic and related firums with quite hardened point of views...

Imho doing nice, regular full charges (==balancing) without trickle charging below the manufacturer specified charge  current threshold and discharging the batteries just to "medium" percentages should be the "nicest" treatment. The huge battery packs of modern EUCs should help with such proceeds.

But still effectiveness of such measures (if there are any at all) comparee to original cell matching and quality is not known (to me at least)

31 minutes ago, Mono said:

I assume we agree that packages can fail even when they are brought to 100% and balanced during each recharge and the underlying reason for the failure is not likely to be related to balancing?

There are many different reasons for battery pack failures. Our EUCs "allow' physical damaging, too high burdens, over discharge, etc - all irreversibly causing damages to individual cells over time.

Mismatched cells - due to aging or original bad matching - can and imho will worsen (faster) if balancing does not soften the situation.

The weakeat cell(s) will be the one(s) reaching the lowest values while discharging. Over time easily below the specified 2.5V. While charging they will reach as first the 4.2V again, activate the baancing resistor and get slowly "overcharged" until at 4.2xV the BMS cuts off or the other cells catch up.

So the already weakest cells get constantly treated the worst.

[WI_Hedgehog]

 

3 hours ago, Mono said:

How did you determine that these failures would have been (likely) prevented by balancing the cells every charge?

I agree with @mrelwood on why:

12 hours ago, mrelwood said:

I religiously balanced my 16S every 10th charge, yet the first pack failed at 4000km (at roughly 80 full charge cycles), the other pack at 8000km. In both cases a cell group had already reached 0V.

From experience, new cells no longer go through multiple charge/discharge cycles at the factory so they can be matched with similar cells, the practical reason being the cell assembly line is moving so fast (due to market competition) the paste thickness/uniformity isn't very uniform, so matched cells will go out of balance with use anyway. Paste thickness tolerance is incredibly tight, cell manufacturing machines wear quickly, and constantly upgrading bearings and the down-time associated with doing so is costly, so it's not done as long as the cells are "good enough" (chabuduo). With so many cells in an EUC, cell matching is impractical anyway.

The full discharge of unbalanced cells is the issue; any time a full discharge happens a balance charge should take place. An 80% discharge can only happen 2-4 times depending on the age of the cells before a balance charge is required--any more than that and there's a risk of killing off the weak cells (generally speaking for new cells).

Cells in parallel tend to balance each other out, packs in parallel do not, BUT, running one cell dead weakens the entire parallel cell group--not immediately, but pretty quickly if you run the packs to 0% because the remaining cells in the parallel cell group with the dead cell are overstressed.

EUC manufacturers that limit pack discharge to 20% instead of 0% manage to extend the life of weaker cells by creating "headroom." As long as @Hangmandoesn't use up that headroom he's fine.

Edited February 18, 2021 by WI_Hedgehog
Clarity

[Hangman]

So a question I posed earlier still has some merit;

Quote

[edited]
How does the regen affect a full battery if one constantly charges to 100% plus balancing?

I was told at one point unless I am going on a long ride, not to charge the battery to 100% because the regen will put a lot of pressure on the batteries.

I've also read somewhere that the mainboard or BMS bleeds excess recharge off so you are losing energy into heat?, but it can only bleed off so much within a certain amount of time. 

I can only assume the larger, more powerful wheels can regen better than the smaller wheels like a V8 but I honestly have no idea. 

I understand that we need not worry about the intricacies of these batteries & how they act & react per cell as others have said earlier, but at the same time I don't want to face plant the pavement because I was negligent in my charging practices.

I'm still new to these wheels as my profile points out I have one wheel, a V8.  That said, I have been following what I was instructed when I made the purchase & am now learning that my practices may be putting my face in jeopardy [dry humor, sorry ].

[Tawpie]

1 hour ago, Hangman said:

following what I was instructed when I made the purchase

hopefully you got good instruction (haha, it's a moving target)! If you balanced the pack every so often, I personally don't think you have anything to worry about but I would get in the habit of regular balancing. If you've never balanced the pack, do so before your next ride (maybe leave it on the charger overnight?) and then get in the habit of whenever you charge it, try to keep charging to full+balance. Just don't leave it charged and on the charger for days/weeks on end... try to ride it often. If you are going to store it, run the battery down to somewhere between 20% and 80% before putting it away and check on it every few months. Charge to full+balance before your next ride and have fun!

Keep an eye on the reported pack voltage after every full+balance cycle, as long as it's not dramatically below 84.4V (79V would be dramatic unless your wheel has many thousands of km on it) I think you've nothing to worry about. But if it's not reaching 84.4V, it's worthy of checking further.

CAVEAT: the above is a semi-educated, but thought through opinion. But it could well be fake news, and no content creators (aka: 'news' organizations) have fact checked it. Don't stake your life on it.

Edited February 19, 2021 by Tawpie

[mrelwood]

4 hours ago, Mono said:

Thanks for sharing, I wasn't aware of these package failures. How did you determine that these failures would have been (likely) prevented by balancing the cells every charge?

1) The max charge my 16S was able to receive dropped slowly from 100%, indicating that the BMS triggered a cell group overvoltage shutdown earlier and earlier as the pack got more and more unbalanced.

2) I have since learned by reading the forum that the charging current is indeed very low, at least on the examined KS 18XL.

3) There are numerous threads about EUCs (of all brands) “not charging past xx%”, many of them related to charging only or mainly to 80-90%. The wheels had been used for as little as 1000 miles / 4 months.

 So, clear signs of pack unbalance, evidence of insufficient balancing current, and several other incidents of even fewer balance charges killing the packs even sooner. To me this is all crystal clear by now.

3 hours ago, Chriull said:

Harshly said one cannot conclude anything from such experiences (sorry @mrelwood). As they are just "anecdotal evidence/single cases", statisticly most likely not relevant, the starting state of the individual cells unkown, etc...

You are absolutely right. I could even be lying. One wouldn’t imagine writing a science paper using these findings as proof, but one can and should reflect whether one’s charging behavior is following the same path, and whether that’s a risk one is willing to take. I started to preach for balancing over any other charging habit in order to give people that chance.

 

14 minutes ago, Hangman said:

So a question I posed earlier still has some merit;

It does. If starting at the top of a hill, one indeed can’t start at a full charge. The wheel would soon start beeping due to overvoltage, and if one disregards the beeping and continues, the wheel will turn off.

How people living on top of hills should take care of balancing is indeed a conundrum. Some only charge to 80% at home, and charge to 100% + balancing at work.

Regarding the burden the regen will cause to the cells, one should remember that there is no hard limit at 100%, over which is bad and under which is good. As presented earlier, even charging to 90% stresses the cells more than charging only to 80%. Charging to 100% even moreso. Having regen briefly take the pack above that is yet even more harmful, etc.

 Going briefly slightly past 4.2V is not a death sentence by any means (although the annihilation voltage is not very far off). It’s the amount of time the cells must endure the said voltage that does the harm. And 4.2V already does some.

 Imagine a graph with voltage at the X axis and amount of harm induced at the Y. The resulting graph is an upwards accelerating curve, like an exponential function. Nothing specific happening at the exact 4.2V.

[mrelwood]

Charging to 80% with a fast charger and continuing to 100% with a regular one doesn’t make much sense to me though. The charging current is limited in the CV phase anyway, so at the end of the charge the fast charger behaves exactly the same as the slow one does. Personally I wouldn’t change to a different charger, and just use the faster one all the way.

Edited February 19, 2021 by mrelwood

[WI_Hedgehog]

1 hour ago, mrelwood said:

Charging to 80% with a fast charger and continuing to 100% with a regular one doesn’t make much sense to me though. The charging current is limited in the CV phase anyway, so at the end of the charge the fast charger behaves exactly the same as the slow one does. Personally I wouldn’t change to a different charger, and just use the faster one all the way.

The V in fast chargers tends to be higher.

Safe maximum (and minimum) voltage on a lithium-based cell depends on that particular cell's chemistry, but typically it's 4.35V. The lifespan at 4.35V is only about 20 charges, sometimes 10, again depending on chemistry. The lifespan of the same cell at 4.20V tends to be 500 charges, and at 4.1V around 1,000 charges--but that's highly dependent on chemistry and manufacturing process. Most "decent" cells will last 1,000 charges at 4.0V maximum. 

That's resting voltage though. Cells have something called "surface charge" where a small percentage of the chemistry (at the surface) will accept and rest at the charge voltage, the remainder needs to be forced a bit. To get 4.20V into a cell it needs to be charged at 4.25V, and when the charge current slows to a pre-set value the charge is stopped. As the cell is allowed to rest the surface voltage dissipates into the cell and the cell "rests" at 4.2V.

If you want to charge the cell faster, charge it at 4.35V and force the electrons in there faster. Sure, you'll heat the cell up some toward the end of the charge and shorten its life slightly, but that's mainly a few molocules thickness of the slurry at the surface (each charge, so it's cumulative). Just be sure to cut the charge off at a higher terminating current so the resting voltage ends up at 4.20V.

The way to fast-charge but not wear out the cell is to charge the cell at constant current until 4.35V, then charge the cell at a constant voltage of 4.25V. Depending on cell capacity, the constant-current charge time is reduced by 50%, and the constant-voltage charge time increased by 50%. Since the constant-current time for a new cell is somewhere around 80% of the charge time, a net time savings of 25% is achieved. If you don't fully balance the cells you can save quite a bit of time...this is what @Hangmanis doing. 

But you're perhaps thinking, "Hedge, if you charge a cell to 4.35V, how can you then continue to charge it at a lower voltage when it's already at a higher voltage?" It's because the high 4.35V surface charge dissipates quickly, so as long as the cell has a little time to rest the voltage drops to 4.20V quickly and the cell can continue to be charged at 4.25V. This is how a really smart BMS works, pausing a bit every so often to evaluate battery state including surface charge; if the BMS sees the surface charge rapidly dropping it knows the cell isn't full and can accept more charge without damage. Some airplane batteries and high-end motorcycle batteries use this BMS technology, and some high-end R/C chargers do also. Since EUC chargers don't yet have this ability, two chargers can be used.

Edited February 19, 2021 by WI_Hedgehog

[mrelwood]

1 hour ago, WI_Hedgehog said:

The V in fast chargers tends to be higher.

That’s the first I heard of this. All the 4-5A chargers I’ve measured (Wate included) have had precisely the same non-burdened output voltage than the bricks provided with the EUCs.

Do the eWheels fast chargers for example have a higher charge voltage? If they really charge at 4.35V per cell even in the CV stage, they would probably cut the balancing process way short as the top cells would reach the BMS cutoff limit much sooner.

[Chriull]

2 hours ago, mrelwood said:

That’s the first I heard of this. All the 4-5A chargers I’ve measured (Wate included) have had precisely the same non-burdened output voltage than the bricks provided with the EUCs.

Mine, too.

2 hours ago, mrelwood said:

Do the eWheels fast chargers for example have a higher charge voltage? If they really charge at 4.35V per cell even in the CV stage, they would probably cut the balancing process way short as the top cells would reach the BMS cutoff limit much sooner.

No. eWheels do not sell any fast charger. For EUCs just slow, faster and normal chargers are available!

Charging with 4.35V would not be possible with the used BMS - it would cut off charging before the CC stage could be finished. 

1C charging was "traditionally" normal charging, everything below slow charging, everything above fast charging!

For our new cells ~0.5C is specified in the datasheets for normal charging, so the border shifted a bit, but fast charging is still (normally) not reached.

Just looked at https://batteryuniversity.com/learn/article/ultra_fast_chargers - naming "already" changed. 0.3-0.5C are now rapid chargers, 1C fast chargers and 1-10C ultra fast chargers...

For all the 21700 4p wheels normal 0.5C charging would be around 4*5/2=10A.

@WI_Hedgehog fast charging is highly interesting - but needs bms and charger designed and adapted for this and each other.

This "surface charge" effect can also be nicely seen while discharging (riding) - under burden (acceleration/incline) the battery voltage sags noticable more as predicted by the "normal internal resistance" model and recovers again once the burden decreases.