mrfusi0n

We all know that when it comes to batteries, manufacturers stretch the truth when it comes to their cell's Ah rating. What I'm wondering is, who here has actually measured the output from their battery? How does that compare to the manufacturer's numbers?
Factors I can think of include:

• Manufacturer's simply stretching the truth
• Battery not fully drained (stopped either by user or BMS - protecting battery)
• Battery not charged to 100% (to prolong the battery's life)
• Using the ebike in a significantly non-typical way

So, you buy a battery that's listed as 10Ah or 15Ah or whatever it is. How much do YOU get out of the battery.


I'll start things off although my data might not be great:
My ebike's battery is sold as 52v 15Ah. The configuration is a little unusual in that it is 3s14p 2170 Tesla cells (I have not opened it up to verify). The math comes out to 780Wh claimed. I use an 850C controller with a TSDZ2 with the open source firmware which is fairly up-to-date but not the latest. Anyway, it tries to keep track of my watt-hours used for a given battery charge. After charging my battery up to 100% (which I will not normally be doing) and using the ebike until the battery voltage dropped to about 43v and the battery/motor were providing noticeably less assist, the 850C said I had used 507Wh.

If that number is accurate, I only get about 65% of the manufacturer's listed capacity! Now that I'm charging to only 80% or 90%, that number will be even lower. All of that being said, I'm actually pretty happy with the range this battery provides. I need to watch the numbers more closely but I think I got 40-some miles out of it.

CliffordK

I think the reality is that most of us don't have an accurate way to test "claims" of a gazillion lumens, or lux or amp hours, etc.

We can tell whether A is qualitatively different than B...

We can also tell if the battery "Good Enough".

One thing that I've encountered with tool batteries is that while they seem to work when they are fresh or new, they are unable to perform their duties as they age.

They might be OK for devices that have a trickle draw, but fail miserably with high draw items, and high voltage drop.

I just bought some new phone batteries. The dang thing really doesn't like the new batteries. Assuming the batteries are properly labelled, I think part of the issue is a different charging curve than the stock batteries.

MNebiker

The first issue is that of manufacturer credibility, which can only be addressed by buying a product from a company as reliable and reputable as possible and hope for the best! That's tough, as most of the packs come from China or contain cells from some unknown company. It's the old "Let the buyer beware!" thing. It would take a good lab to fully test the specs.

So hopefully we can assume that a new battery pack initially meets spec. After that it becomes a question of the design of the battery maintenance system, and cell balancing. See the post by Doc Wui in the "Back to the battery" thread. He explains it well.

In addition to the ebike, I also drive an electric car. In the car forums many complain of loss of range - like you, they see less power usage at cutoff. Some of this can be attributed to normal wear or decay of any rechargeable battery. After 125 - 150K miles it is reasonable to find some loss. But for the most part it seems to be a matter of cell imbalance, and in fact, the manufacturer issued a couple of recalls to reprogram the battery management modules to improve cell balance. This does not happen instantly - in fact it takes several full charge/depletion cycles to restore balance. That balancing is critcal for optimum power. (Fortunately, my car shows no degradation.)

Taking that comparison to the bike battery - the question is whether your BMS is balancing the cells adequately - or at all? If your BMS will try to balance the cells as Doc Wui indicated, try leaving the battery on charge for a period of time and then do your test again. (On my pack the charger runs and then shuts off - and then will randomly cycle a few times to balance the pack.) If this shows no gain, I would suggest that you run the battery through several full cycles and then run your power consumption test again. Li-ion batteries can be damaged by overcharging or deep discharge, but even a basic BMS should control the max SOC and cutoff SOC to protect the battery. (They are temperature sensitive also, so do not store them in hot or cold areas.)

Range is not a reliable benchmark for battery health as there are too many variables: Wind, terrain, speed, temperature all can affect the range. Windy days are a battery killer! And I seem to note a range difference between 45 & 85 degree days. The best you can do is compare similar rides, and as you say, be happy with the range.

dilkes

I believe the pack assemblers publish their specs just by taking the cell manufacturers specs and doing the math on it. These cell mfgr specs do not normally really reflect real life usage on ebikes. For example, the specs for the INR18650-25R cell say -
Nominal discharge capacity 2,500mAh
Charge: 1.25A, 4.20V,CCCV 100mA cut-off
Discharge: 0.2C, 2.5V discharge cut-off

In other words, they say the 2500mAh capacity is rated when discharging at 0.2C or (0.2 x 2500mAh) = 0.5amps and down to 2.5 Volts. If these were in a 6p config (15Ah), that 15Ah comes from a test at 3 Amps. 3 Amps doesn't get you much in the ebike world nor does a typical BMS let them discharge down to 2.5 Volts. At higher discharge rates, the capacity will be less. Many manufacturers provide a discharge curve showing expected/tested capacity at various discharge rates.

Doc_Wui

dlikes is exactly right here. The reputable ebike battery makers base their AH on the capacity of the cells in the packs. The disputable battery guys just make it up. I've seen packs advertised with so many AH that if you did the math, they need twice as many cells than are inside the battery.

As dlikes mentions, ebikes do run the internal cells under different conditions than what was used by the cell makers to test their cells. Ebikes do run them harder, but not as low in voltage, so I think the latter tradeoff compensates and total life remains the same. Still, I doubt you get 90% of the advertised AH out of a battery unless you ride really slow like me.

I have a way to check the AH from my battery packs. I measure with a wattmeter how much AH goes back in when charging.

2old

Real world data: 52V 10 ah Luna battery in its fourth (or fifth) year powering a BBS02 MTB as well as a 1000w, 48V DD rear hub errand bike. Last MTB ride was 17 miles, 2600" of ascending, about 2 hours and the end voltage was 52,5 suggesting that about half of the battery had been used. Took less than 2 hours for a 3 amp charger to restore to 100 %.The good news for me is the battery still has enough for my rides which usually don't exceed 25 miles.

billridesbikes

Can I suggest one more factor, this uses voltage vs. SOC relationship to determine the remaining charge in the battery? Depending on how this is done, load on vs. Load off, the internal impedance of the battery due to temperature, discharge during voltage measurement, and length of time the battery was disconnected from the load can all cause different voltage readings. Depending on the conditions the estimate of remaining charge can be off substantially (+/-25%) since it is a not like determining how much gas is left in a gas tank.

mrfusi0n

Thanks for all the great info, guys!

MNebiker, I'm really curious what BMS system I have in my battery pack. I was already planning on opening it up at some point to verify the exact cells in there but I'd really like to see exactly what BMS system is in there too.

Doc_Wui, measuring the watts used in charging the battery pack up is a cool idea. I'll have to look into that more. I suspect those numbers are a little inflated due to inefficiencies in charging but still a great measure. I was considering an inline one that would be on the bike but hadn't considered this...

2old, thanks so much for providing some real-world data! That sounds like quite a climb!

billridesbikes, that is an interesting point. We cannot really determine SoC given just voltage without knowing the conditions under which the voltage was measured. Interestingly, the firmware on my 850c display has a configuration setting for the battery pack's internal resistance. If I understand it correctly, this value is used to minimize the displayed voltage fluctuations under different load conditions.
For me, I simply determined my battery was "dead" when it was no longer able to provide the wattage I felt I needed for a "good" ride. In that regard, it is a somewhat subjective measure. That same level of charge might be considered 5% or 10% SoC for someone else (or different bike, motor, road conditions, etc.).

dilkes

Another technique I have used for measuring capacity is

• fully charge the pack (off the bike)
• discharge it with a load that causes about 10-15 amp draw (i.e. in same range as mu bikes draw)
• have my wattmeter inline
• measure watt-hr/amp-hr delivered when BMS shuts things down
• In my case, I have an inverter so create the load by using the inverter with a couple of appliances plugged in, (although it's not necessary to have the inverter, I already own one and, using it makes it easier to generate a larger current load from the pack)

I usually do this in the spring when my "riding season" starts and the batteries have been taken out of hibernation

mrfusi0n

So, you bring up a point I've heard a number of times before but haven't experienced myself: "BMS shuts things down". On my bike, when the batter gets low, I find it is only able to pull less and less current. I have a 52v battery and once it gets down under 44v, it starts getting more and more noticeable. Around 43v, I'm only able to pull about 4 or 5 amps (around 200 watts). This then continues to drop off but I haven't pushed it much more than this to avoid damaging the battery.

I was actually expecting the current to stay up and simply cutoff when the BMS determined the voltage was too low. When I look online, I see battery discharge curves for cells at fixed current levels but in my case, it is unable to maintain a fixed current. Is this normal?

2old

Today's numbers with BBS02/52V/10 ah: 12.6 miles; end V 52.6 suggesting 4 ah used, but use 5 ah in case that is an anomaly. I'll get a GPS plot from my friend, but seemed like equal miles of up and down & 2.5 miles of flat. Therefore used about 5 ah for 5 miles or 260 wh for 5 miles = 50 wh/mile and I was trying to pedal hard. 2100" of ascending so average ascent was 2100/25,000 (or so), 8% - 10%.

Frankenbike27

snow, wet or fluffy and cold also affects the power and range, a cold bike has to be pushed downhill to get it going, grease thickens
I like your comment "be happy with the range" just way too many variables

spartan_262

ive been building batteries my entire life, ever since they had Ni-cad battery, so i know almost everything there is to know about batteries.

i write down all the discharged/charged mah that each cell takes, calculate the IR, and match the cell banks before building a battery. BMS works best when cells are matched.



to answer the question of capacity, its important to note that "rated" capacity is at a specific current draw and at a specific voltage range.

if for example a 18650 has 3,000mah , thats at a 0.2c current draw and from 4.2v down to 2.5v lets say. if you took that same battery and over charged it to say 4.25v and discharged down to 2v at 0.1C then you would get something like 3,500mah.

vice/versa,

if you took the same battery and charged it to 4.15v and discharged to 2.9v at 0.2C you would only get something like 2,500mah out of the cell.

if you took the same battery and charged it to 4.15v and discharged to 2.9v say 3C you would only get something like 2,000mah out of the cell. 2/3 of the original "rated" capacity.



now its important to note that most BMS units only charge to 4.15v and discharge down to 2.9v. so right off the top u are losing 20% of your rated capacity.

on top of that most battery packs are 10-20ah in size and most ebike pull about 15A, (1.5C) .


so its not unreasonable for a battery pack you build to only have 60-70% of its rated capacity when riding like a moron. or if you are riding in ECO you might get as high as 80% of its rating.

spartan_262

its not all bad news tho, as charging to only 4.15v and down to only 2.9v will dramatically prolong the life of the battery,which is why most BMS units run 4.15v to 2.9v.

so end of day 3/4 of ur rated capacity, but like 3-4x longer battery life.


worth it IMO.

BattMan

Good explanation from spartan_262 on how rated capacity is true only in very specific conditions.

One advantage of more cells in parallel is lower current through each cell. My 14s2p pack is compact and lightweight, but I can only use 60% of its 300Wh rated capacity on my e-bike. By comparison, I also have a 14s7p pack. I can use 80% of its rated 1200Wh, because of the lower current draw from each cell. The large pack has 4x the nominal capacity of the small pack, but gives me about 5x the range on my e-bike.

spartan_262

yup. This is very true.

same goes for high current cells used in smaller battery packs. Cells that can handle high current draw but have lower rated capacity will actully give u longer run time simply due to the voltage drop under load.