16S Battery Level Variance

[Zanoni]

I've had the KS16s for about a month now.  The thing that I noticed when I first took it out of the box was that the battery % readout in the APP was jumping all over the place (varying up to 5% while just rolling it around by hand).  Another thing that I've noticed in this month of riding, is that the readout just before I turn it off in the morning can be 10% lower than what it says when I turn it back on to do my return commute.  This is my second EUC, following the V5F, and I'm wondering if this is a normal King Song thing (bad battery level algorithm), or is it just something to expect with devices that have larger battery packs?  Or...nobody else is experiencing this, and my device has a bad battery pack?

[Chriull]

@Zanoni i just did a quick test with my KS16S - pushing it by hand forward and back i got the voltage varying between 58,8 and 59,8. So +/-0,5V of ~59V makes a bit less than +/-1% variation or ~2% total.

The main reason for this voltage jumps should be the internal resistance of the batteries - by changing load (current) the output voltage of the batteries changes.

Additionaly one has resistances by the connectors, soldering joints, wires and protection mosfets in the BMS - but these values should be quite lower than the internal battery resistance. (very roughly <10%)

Temperature, charge state (%) and (quite) previous load situations inflict the internal resistance of the batteries.

You have about a bit more than twice the voltage variation?

Reasons for this could be:

- that you pushed your wheel just a bit more than twice as hard than i did. Double load -> double current -> double voltage drop.

- ?your batteries where already quite empty and had by this higher changes?

- the batteries were quite cold

- KS used other batteries in your wheel with higher resistance/the same batteries of the new batches have higher resistances.

- Some connections/etc in your wheel are bad and have a way to high resistance - Normaly they should be roughly below 10% of the battery internal resistance.

- There also could be interferences at your controll board, so the measured voltages not only show the real voltage but +/- some distortion?

- or any combination of points above (plus some points that did not came into my mind by now...)

You also have the 828Wh model? Do there exist models with lower capacities? If so lower capacities (less battery cells in parallel) have a higher internal resistance and lead to more voltage variation.

If the wheel is turned off it is normal that the battery cells revover - my last log stopped with 58,1V and now when i turned the wheel on it showed 59,44V. So a raise in voltage of ~2%. They raise more, the higher the load was before turning off and there is presumably also some influence by the overall charge state (%) and presumably also by temperature changes.

[Freetzee]

18 minutes ago, Chriull said:

If the wheel is turned off it is normal that the battery cells revover - my last log stopped with 58,1V and now when i turned the wheel on it showed 59,44V. So a raise in voltage of ~2%. They raise more, the higher the load was before turning off and there is presumably also some influence by the overall charge state (%) and presumably also by temperature changes.

Just one quick comment, a voltage variation between 58.1V and 59.44V is not 2% in capacity (wich is the %batt shown in the app), it's a 10% battery capacity difference.

Voltage varying between 52.8V (0%) and 67.2V (100%)

So, yes I and apparently others see the same variation of battery capacity, nothing alarming here.

My experience (KS16S) is that when I start riding the battery capacity immediately drops 10%, and that after a stop it regains up to those 10%

[Chriull]

1 hour ago, Freetzee said:

Just one quick comment, a voltage variation between 58.1V and 59.44V is not 2% in capacity (wich is the %batt shown in the app), it's a 10% battery capacity difference.

Voltage varying between 52.8V (0%) and 67.2V (100%)

So, yes I and apparently others see the same variation of battery capacity, nothing alarming here.

My experience (KS16S) is that when I start riding the battery capacity immediately drops 10%, and that after a stop it regains up to those 10%

Ups, yes - i did not read the original post from @Zanonicarefully enough. I took the % values as variation from the voltage and not as the capacity % shown in the app...

With 66V being 100% and 50V being 0% (taken from wheellog source code - should hopefully be more or less what the kingsong app shows, too) my 1V change while pushing the wheel with the hand should be about 6% capactity change (100/(66-50) * 1). So about the same as your readings from the app, @Zanoni - just normal voltage drop by the load applied.

Maybe for your understanding, how the charge percentage is calculated by the kingsong (and wheellog, darknesbot, ...) app: It is directly calculated from the battery voltage! Like written above, for 50V it shows 0, for 66V and above 100% and inbetween just a linear interpolation. charge % = (battery_voltage-50)*100/(66-50).

So the charge percentage shown is only about right, if one lets the batteries rest for a couple of hours! Everytime there is any load on the batteries they show something lower than the real capacity left:

http://batteryuniversity.com/learn/article/how_to_measure_state_of_charge

"The most blatant error of the voltage-based SoC occurs when disturbing a battery with a charge or discharge. The resulting agitation distorts the voltage and it no longer represents a correct SoC reference. To get accurate readings, the battery needs to rest in the open circuit state for at least four hours; battery manufacturers recommend 24 hours for lead acid. This makes the voltage-based SoC method impractical for a battery in active duty."

[Keith]

1 hour ago, Freetzee said:

My experience (KS16S) is that when I start riding the battery capacity immediately drops 10%, and that after a stop it regains up to those 10%

Just to be pedantic here, your capacity hasn’t dropped your voltage has. As @Chriull has said, the voltage dropping under load, and of course proportionally to load, is a direct product of the total internal resistance in the power train. On top of this batteries do chemically recovery a little when left to rest for a time.

The problem, or at least the symptom, that you are seeing is that voltage is being used to indicate capacity and, as you can now see it is a fairly crude measure.

since voltage drop due to capacity use isn’t as linear as one would hope and also varies with load (i.e. how much current is drawn) it is fairly pointless  to put a percentage figure on it. It will be much worst if the battery is close to empty - when it matters - and least when the battery is half full - when it doesn’t much. When the battery is completely full the first drop is very steep, however I suspect most algorithms see something like 4.1 volts as 100% to hide that. The below graph is for  small 800mAh cell at different temperatures and quite low discharge rate. If instead different output loads were used then a higher current at constant temperature would look very similar to the lower temperature curves here.  

 

[Zanoni]

Thanks so much for the explanation.  I actually had no idea that this is how they measured capacity in these batteries.  It sounds like there's probably nothing too unique about my situation then.