Hating on BMSs

[Upnorth]

I’m about to put together a 16s 5p pack with LifePo4 3800mah cells.

i have a large suspicion that either the BMS or the cell level fuses killed my Luna Wolf.

sooooooo I want to not use them.

reasons are:
-I have my suspicions that the BMS just isn’t capable of balancing high output cells. I doubt that the resistors it would use would be able to balance the cells. At 100% it “should” just cut off charge to the high “parallel” groups but I’m betting the charger senses pack voltage before all cells are at voltage. Besides there is NO way a BMS can sense and correct a low voltage cell inside a parallel group if the others are over volting to compensate.
Detailed information on exactly how the BMS “balances” seems to be elusive.
-The BMS “should” prevent reverse polarity connection, with high current 60vlt power it is likely this would fail.
i plan to use XT90 connectors which are hard to plug in backwards.
-The BMS provides low voltage cut off, but when and what is it looking at?? I’ll watch the voltmeter on the controller and set its LVC high, and not power any device that doesn’t have a properly set LVC
-I’ll balance the cells at any point of charge with a balancer that has good resistance, removing the need to charge the cells to max peak volts
-the BMS has overvoltage control, again based on what voltage, I’ll use a Grin Saitiator set to the correct volts and amps....
-by wiring in two 8s jst connectors I can actually read the individual group voltages AND balance them at any point.
-following Lunas advice I only charged up to 100% every other week, which means I rarely balanced anyway, plugging in a checker to read the levels and balancing if required isn’t an ordeal at all. I’m betting that after a couple weeks of commuting without balancing the cells were so far off the charger would be cutting out before the BMS had time to balance the cells, which is why almost all vendors are telling “street level Ebikers” to “always charge to 100” it keeps them balanced and losing up to 50% of the cycle life is wayyyyyy better than killing a wolf in 175 cycles.......
- as for the fuses, I’ll wire in an 80amp replaceable fuse, if something shorts that should prevent a fire, in any case LifePo4 cells are suppose to be much safer.

https://www.amazon.com/Heltec-Capaci...ws_feature_div

[Upnorth]

I'm curious.
Would you share your experience with LUNA Wolf batteries?
Length of ownership, number of cycles, normal depth of discharge, type of charger used, storage state/length of storage.

Please just share your raw data.

I followed Luna's advice and my Wolf lasted 4500km, roughly 185 cycles, which adds up to $0.45cdn/km
It was damn fun and they have a great form factor and solid build but I just cant justify the high cost per km.

It is not just me read the following posts:

https://electricbike.com/forum/forum...ry-performance

https://electricbike.com/forum/forum...ng-pack-issues

https://electricbike.com/forum/forum...-dying-already

https://electricbike.com/forum/forum...rly-lvc-53vlts

https://electricbike.com/forum/forum...-with-problems

https://electricbike.com/forum/forum...-its-last-legs

https://electricbike.com/forum/forum...-the-right-one

https://electricbike.com/forum/forum...sag-under-load

It would be nice to know what kills them..... I listed my SUSPICIONS that is all, a non expert opinion, and I am hoping that future batteries last longer.

[Upnorth]

Thank you, this is good data. My pack at 185 cycle SHOULD have over 90% capacity. It has LESS than 20%.

The key to your success may well be charging to 100%. This likely helps the BMS keep the cells balanced. This is something that the community at large seems to think reduces capacity, common knowledge seems to be that depth of discharge impacts cell life. That is dependent on the BMS being able to keep the cells balanced. On mine it did not. LUNA told me I did not need to charge to 100% every use and I think the BMS was inadequate to keep the cells balanced.

I have followed LUNAs advice in the article you linked above, I use a stock controller and rarely use the throttle, NO overly high discharge rate. I used LUNA's "advanced" controller until this spring and a Grin Satiator after that.

More Data from other users would be interesting.
Had I know this I would have charged to 100%, storing at around 60-80% if I am not riding soon after the charge.

I have a Wolf V2 that I purchased last year. I will do that with it, and report back as time passes, or its behavior changes. IF after 700 cycles I have 83% capacity left I will be very happy.

This may be something that LUNA and others could chime in on.

Mike_V, you likely now have over 700 cycles, according to the graph your choice of LG over the 30q is now getting you at least 10% more capacity and likely that much more amperage. Again something to consider when purchasing a pack.

[AZguy]

Yeah, you really need to charge them 100% every few charges and frankly I frequently charge to 100% if I'm going to use it soon... I have a lot of battery's and so they all sit for days to weeks at a time and I would never do that at 100%, usually about 70%... from 70% I can charge it up to 90-100% in short order with the 5A charger if I'm in a hurry but I usually know to throw them on at 2A the night before I'm going to use them. I setup the "poor man's smart charger" like this too: https://electricbike.com/forum/forum...charger-reload

[K442]

Seems like I see a lot of suspicions and assumptions with regard to the BMS then the conclusion of "I'm better off to do it myself". And in reality, the BMS's in some of these packs seem to be pretty cheaply made. (I'm actually rebuilding from a BMS failure on my X1 pack) So I guess your proposal is one way to look at it, but are you really going to hook up 16 volt meters, then monitor them throughout your ride and get off and walk if the cells get too far out of balance or too discharged? Same on the charging side...sitting there watching the meters and plugging in / unplugging cell packs until all are balanced?

As for some feedback:

"i have a large suspicion that either the BMS or the cell level fuses killed my Luna Wolf"

Seems like this would be easy enough to check. Are the fuses blown? Is the BMS not charging discharging correctly? Are the cell packs out of balance?

"I have my suspicions that the BMS just isn’t capable of balancing high output cells. I doubt that the resistors it would use would be able to balance the cells. At 100% it “should” just cut off charge to the high “parallel” groups but I’m betting the charger senses pack voltage before all cells are at voltage. Besides there is NO way a BMS can sense and correct a low voltage cell inside a parallel group if the others are over volting to compensate."

What would lead to this suspicion? What data confirms it? BMS works regardless of cell output and doesn't use resistors. It's also not looking at individual cells within a parallel group. By default ('default' meaning the nickel strips welding the respective ends of the cells together) parallel cells have to be at the same voltage. The BMS is monitoring the groups of parallel cells in series and making sure each parallel group in the series has equal voltage.

One key here it that while charging typically takes place at multiple amps to 10's of amps of current, and typically completed in a few hours - cell balancing typically takes place at 20-30 milliamps of current (0.030 amps). This is usually not an issue as you would typically have only minor imbalances to level out, but if you have a large imbalance on large capacity cells, it can take a while. For instance, you mention 5p 3800mAh cells, so that is 5 x 3800 or 19,000mAh capacity. If one of those groups is out of balance by 25%, that is .25 x 19,000 = 4750 mAh. If your BMS is balancing at 30mA, then you need 4750mAh / 30mA = about 158 hours to balance the cell.

"Detailed information on exactly how the BMS “balances” seems to be elusive."

Web Search? https://duckduckgo.com/?t=ffab&q=how...attery+voltage

-I’ll balance the cells at any point of charge with a balancer that has good resistance, removing the need to charge the cells to max peak volts

Guess this is OK if you plan to dedicate a good portion of your day to watching 16 individual cell packs charge, then switching/swapping/measuring them to make sure each one gets charged equally. Plus you'll need to plan to have that sort of access to your battery. A lot of packs are covered in shrink wrap and additional layers to make a solid, electrically insulating, waterproof case. You'll need access to taps for all 16 cell groups and those wires will be live all the time.

-by wiring in two 8s jst connectors I can actually read the individual group voltages AND balance them at any point.

Yes, you can...assuming you have nothing better to do for a couple hours each time the battery needs to be charged. Me - I have lots more important things to do... watching paint dry....making sure the grass is growing properly....etc.

....etc

I'm not specifically trying to bash you - just trying to make a point that while all this manual work sounds good in practice, it rarely works out in day-to-day operation. If you are having trouble with the cheap OEM units, possibly a bit more robust BMS will work. But overall, a [good] BMS should be [is] watching your battery pack essentially continuously to spot minor issues and correct them, or also spot major issues and shut the battery down before they become extreme issues. ('extreme' meaning your bike or house burning down because a cell shorted due to extreme over or undercharging)

This reminds me of a friend I had who decided he didn't want to pay 20 bucks for an electronic thermostat to turn the cooling fans on in his car. "A manual switch is good enough for me! I'll do it myself!" he said. So after I rescued him twice because the car overheated when he forgot to turn the fan on, and once due to dead battery because he forgot to turn the fan off, I was about ready to fork over the $20 for HIS electronic switch!

Good luck with what ever method you choose.

[Upnorth]

I appreciate the posts, constructive criticism is why I put up a thread, so thank you to all who have replied

K442, the over plan WAS to use a quality cell checker and a quality cell balancer, not 16 multi meters, every other week, perhaps weekly, more if cell condition dictated, this would be a set it and leave it operation, similar to letting a BMS charge to100.
You stated that "BMS works regardless of cell output and doesn't use resistors" then linked to and article that says in the first sentence "The BMS will look for the lowest cell, then place a load on all cells which are more than the maximum difference in voltage above the lowest cell"
The problem is that if the WOLF BMS uses a load its buried in potting and would have to be a small load (there is no other load that would be used other than a resitor that would generate heat).
Secondly, no, none of this is easy to diagnose in the WOLF, it is fully potted. Had I waited until the very cool clear one was available I would be able to a least look for something obvious, but as it is the case is solid black and all of the goods are deep inside a solid block of potting. It is a great design, truly it is. Mine just died 185 cycles in, to me this isn't acceptable.

The case you make about the amount of time it takes to balance a battery it exactly why a BMS cannot be relied on to do it, unless it is done often enough to keep the imbalance very small. Most chargers have a system that watches the amperage drop off and after it goes below a certain level they either shut of right away or continue providing very low voltage for a set amount of time. As far as I know NO charger set for lithium ion batteries will "float" like lead acid chargers do. So if the cells are taken to far out of balance they never will get back by the BMS alone, the BMS will stop balancing when no current is being put into the battery. A dedicated balancer is far faster, I was looking at one with a 2 amp load, and does need there to be a charge current to run. AND yes they can work well un attended.

If the 80-20 rule so widely recommended, even by LUNA, (though I suspect by the number of users with early dead Wolves they have removed reference to it on their site) is valid and the BMS only needs to be balanced every other week then a dedicated balancer isn't a pain at all. Yes you are correct that a bank of two 9 pin JST connectors would need to be accessible. Given the fact that there is no way to diagnose or asses battery health without it I think its not a bad plan really.

HOWEVER. in the end your anecdote about the fan switch hits the nail on the head.
I don't want to be that guy when the novelty of balancing a battery wears off.


When I added up the cost of a cell checker and balancer I found that the same company that makes the balancer also makes a bluetooth BMS that allows a whole host of options and programming controlled by the user, and its cheaper than the other two......

Having the ability to look at series voltage quickly on my phone will give me an idea of the over all condition of the battery.

ALSO, I just got a case of 26650 LifePO4 rated at 3800mah, 12amp continuous load and the cycle life is in the thousands, not hundreds. 16s3p should happily give me the 1200watts I think my BBSHD would consume the way it is set now, though I think I rarely (if ever) load it up that high.

So yes I am changing my mind, I am readily willing to admit where I was wrong.

Now to come up with a case that I can get into should I run into a bad cell. Throwing away 55 good half life 30q cells because there is one bad one in the 14s4p Wolf is upsetting.

[Upnorth]

KK42, one more thing......
Ever taken apart a battery that has failed? I have and found that ONE cell read zero and the other two in its parallel group were OVER 4.2........ so while the theory is solid that all parallel cells can be considered one, with the sum of the capacity and amperage, theory goes out the window when things go wrong....

Also
”then monitor them throughout your ride and get off and walk if the cells get too far out of balance or too discharged? ”
Irony is that is exactly what happend when the Wolf BMS shut off with 53volts in the pack. Though I didn’t get off and walk I sucked it up and peddled the 12km home on my own power. I wasn’t given the choice to allow an imbalance/low voltage in a parallel group OR the information to make that choice.


With the design of the WOLF there is no way to know. Shame is that I haven't yet come across an "off the shelf" battery I would trust more than the WOLF. I just wish I hadn't drank the 80-20 cool aid....

[AZguy]

...Most chargers have a system that watches the amperage drop off and after it goes below a certain level they either shut of right away or continue providing very low voltage for a set amount of time. As far as I know NO charger set for lithium ion batteries will "float" like lead acid chargers do....

FWIW I have several cheap chargers and they all just simply do constant-current and then just simply switch over to constant-voltage and stay there even when the current drops off (I've watched with meters) and continue to apply the voltage when the charge is "complete" - i.e. the LED turns green... all the green light means on those is that it's delivering less than some "X" amount of current but it does keep delivering voltage and some small amount of current (enough to sustain the voltage) even when the LED turns green. But frankly none of this really matters.

Float on lead-acid is a different thing altogether since once lead-acid has completed charging you don't want to keep the charging voltage on the battery or you'll cook out the electrolyte (primarily by disassociating the H2O through electrolysis) so a lead-acid charger worth it's salt will reduce the voltage to a "float" voltage after charge is complete to maintain the battery at 100% state-of-charge and should only need to be supplying the battery leakage current at that point. Unlike Li-ion lead-acid is very happy to sit at 100% state-of-charge forever.

Some Li-ion chargers (I also have one of these) do turn off the output once the current gets low enough to trip the end-of-charge stage (i.e. the current falls below "X").

Here's why it doesn't matter. Balancing works by shunting current around cells that have reached the full charge voltage (generally 4.2V) so the battery (BMS) keeps pulling the current the lower cells require until the very last one reaches the full charge voltage and so they're still pulling well above "X" current and the charger still stays in the charging mode and won't trip the end-of-charge stage until all the cells have reached full-charge and are balanced at this point.

Some BMS's will begin shunting the current prior to the individual cells reaching full charge voltage which gets the balancing started earlier but again, it doesn't really make much difference


Having said all that there are more than a couple of ways users can destroy batteries

Likely the most common is letting battery discharge too low which will kill cells. Li-ion (lead acid too) just can't handle too low a voltage and it will kill them or at least significantly reduce their life and capacity (essentially kill them). NiMH are different and can be discharged to zero and don't seem to care. Most of the time this happens when someone puts the bike away with the battery low, leaves it for months, the battery self-discharges way too low and one or more groups of cells gets damaged.

Another is by leaving the battery on a cheap charger like the first ones I mentioned that just keep delivering the voltage and don't trip (turn-off) after reaching end-of-charge state for a long time. This generally happens like the prior case when someone stores a bike for months at a time but instead leaves it plugged in to one of the cheap chargers. Unlike lead-acid this doesn't cook electrolytes but instead promotes electrode migration (dendritic growth) to the point where the cells leakage resistance drops too much.

Not charging fully often enough so that a group of cells gets so far out of balance they discharge too low and become damaged is just another common way for a user to kill a battery

These are not the only ways to kill a battery but are likely the most common



Now I don't know why your battery died on you. No offense but just based on probability, odds are that it was one of the above or some other user error. Not saying that yours may not be an outlier and just plain died due to a manufacturing issue with the battery or one or more cells, but just that based on odds alone, that's not the likely cause...

[Upnorth]

From Luna directly, on this very forum:

“Generally for storage and usage you want to keep it between 90%-10%, and more ideally 80%-20%, “

https://electricbike.com/forum/forum/knowledge-base/batteries-aa/62526-luna-wolf-pack-battery-documentation

[Upnorth]

Mike V did you even go looking for it?

this is NOT the only place I have heard it but this is directly from Luna and in reference directly to this Wolf battery.

I’m betting there are many people following this advice, and they too will have a battery that does early due to cell imbalance.

why did you delete the graph you posted earlier, and your response with actual data, those posts were actually helpful.

[Mike_V]

Leave me out stupid

[Upnorth]

Have a look at the picture, that is Luna’s advice. NO mention at all about the BMS not being able to balance the cells if this advice is followed. For fun stay on that link, read the advice then scroll down a little bit, on the very first page there is a comment from another user with the same issues I am having....... irony no?

I saved the picture because dollars to donuts Luna will pull the section advising users to stay between 80-20%,
as the should!

[Mike_V]

You should be ashamed for neglecting to contact Luna for help instead of typing nonsense.

[paxtana]

What you are describing is a bms issue that affects a relatively small number of battery packs and is addressed in the new bms we are working on for next version of Wolf. Premature LVC is not caused by imbalance, not caused by keeping a battery between 80-20%, the bms we use does not even need to be fully charged to balance.

Keeping the battery within the most ideal band of voltage is good for capacity and prevents dendrite formation within the electrolyte, and puts less stress on the cells which can lead to imbalance, that is true of all lithium in existence, even batteries with no bms. As such following that guidance in the documentation does not somehow cause imbalance, but rather the opposite.

While unfortunate that sometimes a safety feature can cause an issue, we feel the solution is improving the feature to make it more robust rather than getting rid of it entirely, as the alternative would effectively make it less safe.

[Upnorth]

Thank you for responding.
i agree that is unfortunate, what advice do you have for the “relatively few”?

LUNA is top notch, through out all of this I have maintained that though this situation sucks I cannot find a retail battery that I would buy instead of the Wolf.

What do you recommend I do?

[paxtana]

If you contact support we can walk you through testing to get the issue on video for confirmation then propose a discounted replacement pending pics of it being responsibly recycled, the discount amount is relative to the age of the order.

[Upnorth]

Thank you very much.
LUNA proves again it is a stand up company!

This has been a frustrating situation, due to the fact it is far from a brand new battery and there are so many variables that are way out side the control of the manufacturer I didn’t consider contacting them.
I should have from the get go.

[Mike_V]

Because looked before I replied about what you likely neglected: Here are the instructions