Thursday, 20 June 2019

Electric Vehicles Battery Facts (helping make the most out of yours)


Having had a few discharge cycles on my Mercane scooter I've been getting some experiential evidence which (strangely ;-) supports the facts about Li-Ion batterys that I have known for some time at a theoretical level. None of this will be any surprise to a seasoned electric bicycle, skateboard or scooter rider but none the less may provide some value to help you get the most out of that battery pack before it becomes landfill.

TLDR;

Modern folks are impatient, so:

  1. batteries are not cells, by grouping them we have greater voltage and therefore greater power, so therefore what is done to charge is different (more complex)
  2. let your battery fully charge (at least every third charge) to allow cell balance to occur (I fully charge >90% of the time
  3. discharge more shallow if you can: meaning don't let it get as flat as a lizard drinking water all the time, so charge it early rather than late.


In this post I'm going to reference two really good articles:

I get that much of that is way more technical than most riders want to get into, but if you want to get the most out of the money you spend on your gear then hopefully understanding these points in my blog post will help.

So lets start with

power and power loss

Electric motors are not engines, they only convert electrical power into rotation. This power is stored in the battery, the less electrical power you have the less "power" your bike or scooter or skateboard has to get up hills (or to accelerate). In electricity power is the simple multiplication of Volts and Amps, so as your battery drops in Volts the power available at your back (or front) wheel drops too. This is the typical power curve of a Lithium battery (note: a single cell):


So you can see that this drops from 4V down to 3.5V in a fairly gradual and linear manner, but falls off a cliff as you go below that. This is important because of two reasons:

  1. you will degrade the amount of life your battery has if you let it go below 3V (and some battery controllers will let it go to 2.8) - so don't deeply discharge as a habit.
  2. power available takes a big dive fast after that 3.5V level which is when you only have 20% of the charge left - meaning you may end up walking suddenly in that zone.

Batteries VS Cells

A "battery" is a bunch of cells together (here), and so to understand your battery you need to understand cells. Below is a "discharge" diagram of two cells, showing how quickly it falls from a fully charged voltage to a fully discharged level.


Note that they both really stop at between 3 and 2.8V This is because the test is truncated there, going lower than that will damage the cell. Also note that ffrom about 3V things get a bit hairy and they fall really fast.

Most eBikes and eScooters have electronics in them to prevent damage to the battery from trying to drain too far or charge too high; this is a Battery Management System (BMS).

Juggling requirements

Cells both weigh and cost; so all eBike and eScooter and especially eSkateboard makers will want to balance weight VS cost VS performance very tightly. They may also opt for clamping you in a tighter range of the battery using electronics that s only down the the 40% state of charge. This is what a "good" battery manager will do, and this will mean you get years out of your battery instead of months. The best of my knowledge this is only done in higher end EV's like cars.

Why: well this is because depth of discharge has a great impact on cell life.

Now also that above graph does not give voltage above 4V, which seems a bit surprising when a fully charged 18650 cell will show 4.2V on the volt meter when its fresh off the charger. There are some possible reasons for that but lets just say that (as you can see) its unimportant because the voltage drops pretty fast off the charger.

Below is the curve you'd expect from actual battery (a bunch of cells) tests (see this site for its excellent discharge database):



To get larger voltages you put a string of cells together in a Series, and to increase how many Ah you get you put them in a Parallel arrangement (all the +ve connected and all the -ve connected) Now  I have a 13S4P in my scooter; this means it has 4 cells in parallel and 13 of those bundles arranged in a Series.

My Mercane Widewheel 500W model is 13S4P, so that means that to get 1A per cell (in the above diagram) you've got to be sucking 4A from the pack and to get 3A per cell 12A from the pack.

The voltages are what you can read when its under load, not sitting around, also is why some battery packs are reported as 48V and others 52V. Because depending on what you want to call standard battery voltage the lithium cell in your battery pack may be called 3.6, 3.7 or as much as 4.2. My Mercane shows 52.4V* when its fresh off the charger which is close to 4V per cell (*note: I've come to suspect this is some sort of fault ... either of the BMS or what I don't know, because its not common NOR does my dual motor version do that I have since replaced the BMS and repaired an issue in the pack).

But (if you have a voltage gauge like the Mercane does) we all experience that pretty quickly the Voltage drops to the high 40's (or in that 4 to 3.5 range of linear decline).

So when my Mercane is fresh off the charger (or my mates skateboard) it will climb hills nicely, but when its about "half way" discharged (based on how far I normally travel before recharging) its already noticeably sluggish going up that hill. This comes back to that power equation of Volts x Amps that I mentioned earlier.

  • At 52V (and lets assume 10 amps) I have 520Watts available on my rear wheel (and the Mercane has a 500W motor so that's unsurprising),
  • But as my voltage falls to 46V I have only 460W available. (and this is where I choose to recharge my battery, because like; why run it down if I don't have to).
  • 44V = 440W ...
you get the picture. So those of us who like the fresh power hit will recharge about then, which is usually less distance than the maker says the bike will do. So as I mentioned, I typically put it on the charger at 46V and this turns out to be a good thing for the battery and for my wallet (recharging early).

Looking at the second article we find some interesting points, firstly this chart showing how the batterys stood up with full discharge cycles:


From that article they write about that:
The 1,500mAh pouch cells for mobile phones were first charged at a current of 1,500mA (1C) to 4.20V/cell and then allowed to saturate to 0.05C (75mA) as part of the full charge saturation. The batteries were then discharged at 1,500mA to 3.0V/cell, and the cycle was repeated. ... All packs started at a capacity of 88–94% and decreased to 73–84% after 250 full discharge cycles. The 1500mAh pouch packs are used in mobile phones.
So they relentlessly hammered them down and then recharged and then repeated for 250 times.

Just take a moment to think about that, to equate this to real world riding you'd have to fully saturation charge (the so called 100%) and up the steepest slopes it can take (because at higher speeds you don't draw as many amps) until it was flat, then charge and continue going up hill again without any break except to fully charge.  Ask yourself "is that likely"? (I'm guessing no)

Further they were charged at C which is Capacity, which for a 8Ah battery pack would be being charged at 8 amps ... can you even do that? I'm willing to bet NO because a standard charger is 2A

So basically this piece of evidence used by people (who don't get what they're reading) to justify don't fully charge is irrelevant to real world eScooter and eBike riding.

Even then they found that by going the extra distance and taking the battery to "Battery Management System" shut down (meaning your walking) that after 200 cycles some of those batteries had dropped 20% of their capacity from new. Whoopdie do. At the risk of stating the obvious there's 52 weeks in a year, so I'll leave it to you to work out if that's years or a years worth of use.

Next they found (presented below) that only discharging as deeply as 60% Depth of Discharge that the battery could be cycled 1,500 times before falling into that failure above compared to a less than half the charges by taking it to the limit each time


... which is about where I take my battery to with what I wrote above of taking it to 46V (or about 3.5V per cell).

The solution it seems simple; charger earlier and only use that full discharge as an unexpected need

This leads me to my last point, which I mentioned earlier and that is the 4V level (per cell) in the first graph. Given that LiION batterys can be charged to 4.2, why not charge them to that? Why 4?

Well again its longevity but to be honest its not as simple as the raw data would suggest. It seems that by charging the cell to 4V we get a return in a increase in the number of cycles possible before loss.


While this seems attractive it ignores an important point ... we are not charging just one battery.

In the real scooter world we use battery packs (not just a single cell) which are comprised of (around and for example 4 bundles of cells in parallel and 13 of those bundles in series (daisy chain)) to make the humble 18650 cell (which is 4.2V fully charged delivering typically 2.6Ah) into a beast that can deliver 52V and 13Ah ... such a pack is called 13S4P.

The parallel bundles keep themselves balanced (by the physics of being in parallel) the string of series bundles will tend to get themselves into trouble over discharge / recharge cycles (because only in a computer simulation is everything ever perfectly identical). Thus we have BMS systems which actually take care of this, the rub is however the more simple ones (meaning costing less) we usually see in our battery packs do their balance at the end of the charge cycle (moving voltage from the full ones beside the ones slower to charge). This is why its significant to recharge full.

But, don't wait till its super low to recharge as that's just pushing the battery cycles deeper and that will also contribute to lower cycle life ... This is very positive (yep, you picked it, that's a pun).

Whats in my pack?

This is where its important to know how many cells in your pack not just its voltage, because knowing both tells you about how long your likely to get (assuming the cells themselves aren't crap). Why? As mentioned above,  battery packs are combinations of Series and Parallel, this of course dictates the voltages found.

For instance a 12 cell pack charged to 4.2V per cell would be showing 50.4V and so when coupled to a 48V motor would seem right, right? But equally a 13 cell pack charged to 4V will be 52V and as we've just seen will mean that pack will last more cycles (or years instead of months) and not to mention not die as fast. (*note: in practice this isn't what's done in the vast majority of systems, but I'll get to that).

Die as fast in both ways:

1) run out of grunt (because when the cells drop off to nominal not full it will be 3.7V per cell or 44.4V and your motor won't have the 500W anymore it'll only have access to 444W)
2) life cycle till the battery won't hold much charge (due to excessive electrical wear on the battery).

But this isn't the whole picture

Batteries as mentioned above are really a string of cells (which just like a bunch of chickens) is called a battery.



As it happens most eBikes and scooters have something over 10 and maybe as many as 16 cells in a string to make that battery. Like this:
So if each cell is as mentioned above nominally 3.6V or fully charged at 4.2V then

  • 4 cells would be 16.8V
  • 13 cells would be 54.6V
  • and the 16 cells here would be 67V 

As you can imagine, it would only be in ones dreams that each of those 16 cells (or even 13 of them) would accept charge identically and discharge identically, which will lead to what's called "imbalance" and the pack would soon become unusable.

Enter the Battery Management System (BMS), whos job it is to keep that pack from over charging some cells and leaving others under charged.

Discussing the BMS is actually a complex subject but for our purposes here lets simply it by saying that the BMS starts forcing the "stragglers" to take some of the power that the full batteries have and thus keeping the pack level when its finished.

But (and here's the important bit), it only does this balancing (well lets exclude the high end ones not found in eBikes or Scooters) right at the end. This is important if you're inclined to not fully charge the pack before removing it from the charger because it can lead to an unbalanced pack, which will definitely result in your experience of less charge.

So by not understanding the details and by listening to people without a clue you're probably more likely to be damaging your battery than protecting it.

So if you are inclined to give your battery a little top up now and then (rather than fully charging it) that's not a bad thing, but you really do need to let it fully charge most of the time to ensure your batteries are brought back up to "being full" most of the time.


Lastly while I don't normally feature other peoples work I recommend you watch this youtube presentation on the subject (which I found recently in answering some twit on Reddit)


Its really worth sticking it to the end of that because (like my article above) he starts with the theory first (the ideal situation) and then moves into the practical and then lastly "what he does".

Outcome

So hopefully you'll be able to make what you have last more now and be better equipped with knowledge on the next time you are researching eBikes, scooters or skateboards. Be careful when reading specification (because most of us can't go to a shop and see these things nowdays) and ask some important questions of the maker (like how many cells, what voltage does it charge to ...).

Best Wishes

15 comments:

matt ward said...

Thanks for all these really helpful articles. I just took delivery of my 2019 twin motor widewheel. When fully charged the voltage meter reads 54.7. Based on my understanding from this article I shouldn't be charging it to full. So what is the voltage that I should be charging it to? i.e. what should the voltage on the key ignition be reading? Thanks again!

Anonymous said...

I have dual motor WW. I’ve been riding it for 3 weeks now. At first, the battery would charge to 54.5 or so, now it climbs to 52.7 max.

I’m worried that it might be battery degradation, but that would be so soon, don’t you think?

I charge it everyday overnight for about 6 or 7 hours. My daily commute is 20 km. By the time I reach home, it would have lost one dot, and the voltmeter reads around 46 (standing). One day I forgot to charge it at night and it couldn’t go halfway to my work (about 5km it gave in) and it had dropped only two dots, but the voltmeter read 39.

How’s it going with yours? Are you experiencing battery wear too?

obakesan said...

Hi there

its hard to be sure as it could be a few things. Your experience of pushing it hard and the BMS kicking in and denying power from the battery is something I've have once on my single motor. However your description of your usage is that you're doing a heavy discharge cycle on the battery daily (driving it down to low levels like literally just before cut off). This is how you minimize your battery life cycle count.

I think that you've made a decision to buy something which is right on the edge of its design parameters (and that's fine) and so you have to juggle (which you may have) the cost benefit of that as well as the cost to step up to something more expensive (like a dualtron{and still have other types of failures}).

As an aside, I noted the other day that Aliexpress had batteries for US$270 for the mercane, which if you're not the electrnically competent type means you don't have to rebuild it yourself.

My first advice is buy another charger and charge it at work if you can, they cost about $25 on eBay and would mean that you can charge at work which will 1) reduce the discharge depth 2) give you an opportunity to remove one critical failure point (for right now if you forget to charge its critical right?).

The plug type is XT60 and so you can just buy any charger fore 54V lithium (and there are plenty, they even look like ours) and just solder on a new plug. Just make dam fuckin sure you put the +ve and -ve on the right way round ... they should be marked. Unplug the charger and let it stand for a few minutes before you cut off the end. I see XT60 plugs on ebay for low prices too, so it shouldn't take more than $30 for a complete "other charger"

I'd suggest that based on what you're saying (such high usage) you would anticipate about 2 months before you started to observe battery loss because you ARE cycling to about full discharge every day). Getting another charger will be the cheapest way to ensure that you get over 500 cycles (so more than a year) on that battery before losing significant capacity.

As a benefit you'll also be able to suffer that degredation because you will only be doing 10Km per charge.

I also STRONGLY recommend that (given your miles) you check things like all bolts and screws on the scooter for being firm. Loctite where you can. I'll do a post on this soon but I discovered a small amount of play in my rear end (because the brake made a sound in one lean direction but not the other) and took my covers off to check. Turned out to be one swing arm bolt needed a quarter turn to snug it up and the problem went away.

A stitch in time ... right?

Best Wishes

obakesan said...

oh, I meant to write "periodically check" ... its not a one off thing. Same is true for a bicycle ...

Anonymous said...

Hello,

I'm so sorry to asking this and i'm sure it must be somewhere in your text but 1 English is not my native langauage and 2 this is all very technical. Is this what i try to conclude right?

- always charge to 95%
- never drive battery empty
- once a week charge to 100%

Or is there the need to once a week fully discharge also?
Thank you in advance for you answer!

kind regards,
Phil

obakesan said...

Hi Phil

1) no I very seldom charge to 95% I mostly charge to 100%
2) correct
3) personally that's not my style, but as long as you charge to 100% at least 1 charge in 3 I am not sure if it will be a problem. However I'm unwilling to risk battery pack imbalance (having had to pull a pack apart to balance charge) I'd rather just fully charge it and never have that problem.

My approach is to wait till its at 40% before recharging, if that's a few small trips then great, if thats only one trip then great.

Anonymous said...

Many thanks for your help!

Ali Dalkus said...

Hi, thanks for the very informative article you prepared. I am having my Widewheel Pro since 3 weeks and till now I was discharging it till almost zero when the speed was falling less than 10km/h( by especially doing extra miles to finish the battery).after that I was doing full cycle charge. Now I see that was very wrong. Now I will follow your advise. Thanks. Ali

obakesan said...

Hi Ali, thanks for your kind comment.

My view is to take it down to no less than 44V while under power (full cruise control speed with Power mode not ludicrous mode).

Not recharge until its about that (often 3 or so trips, but maybe one longer trip) and make sure its fully charged. I avoid recharging till the night before I know I'll be using it the next day (so that it doesn't sit for days at full charge, minimising any potential for Nickle plating out).

Happy Scooting

Ali Dalkus said...

I just noticed(after doing 300 km in my first month, when I charge I see 54.4v. I never seen 54.6v( first 2 weeks I didn't check or don't remember) but since last 2 weeks as starting to have more obsession to battery (As you mentioned in one of your articles 😀 ) now I am checking what volt I have on the LCD screen. The performance seem to be ok as I can get ranges 38-41 km in average.

obakesan said...

Ali, before you get concerned at splitting hairs over insignificant amounts that you can't be certain of, just check the >actual voltage< that the charger itself puts out and remember that i) its not a hand calibrated "NASA standards" device ii) the battery voltage will be different because you are seldom anywhere reading off the battery because the BMS is always part of the circuit.

Next, read this post: https://cjeastwd.blogspot.com/2020/02/tweaking-up-your-charger.html

Best Wishes

Ali Dalkus said...

Thanks for the advice. I am afraid without a voltmeter I won't be able to tweak it healthy

obakesan said...

agreed, but the reality is that its probably quite sufficient as it is. I expect that such a minor V difference (0.2v / 13 = 0.015V) is an insignificant issue. What will be more interesting to watch for is that you do not find that (when unplugging the charger with the scooter turned on so that you can see the Volts) that Volts do not instantly plummet 1 or 2 V. This indicates a problem in the cell pack (probably an inbalance). This is a video I did to send to the seller (https://youtu.be/n8N2xySOA6Y)

Ali Dalkus said...

I remember the problem you had. With mine there is no change in volts with charger connected or not. I still see 54.4v. maybe it's just a charger tweak issue. Will monitor if anything changes in near future. Thanks for the advice

obakesan said...

I was simply trying to illustrate what a problem will look like. Until then I see no significant issue and as you have said, just know what it is and then look for changes