KS16X Cutoff/Overlean - 29-30mph

[Aneta]

5 minutes ago, Chriull said:

So beside curiosity i'm not really sure if such a v2 model makes "sense" here for EUCs. We don't even know the coil resistances, accuracy/factor of reported current values, etc...

... and no test rigs like they have to veryfy and "calibrate" the model.

So i stay with my small collection of overlean logs to guess some simple approximated limits...

For us, Simulator is a great illustrative tool, not precise calculator. But general trends we deduce from it are quite useful for us.

[Chriull]

52 minutes ago, Aneta said:

More Re: concaveness - at the bottom of the Simulator page, there's this piece:

"

The actual model in SimulatorV2 is substantially more complicated, taking into account the commutations that happen on a regular basis as a function of the speed and number of poles of the hub motor, and determining the resulting current waveforms that are produced when this is applied to the inductive motor windings.

"

It's the negative quadratic term caused by inductance that makes the torque curve concave.

I don't see how the negative quadratic term could be caused by inductance? Could be an approximation for the di/dt for the voltage drop? But since it's with omega² it would be reasonable to describe the inertia of the wheels mass which depends on d omega/dt?

But still these all would be dynamic effects for burden changes, and as i understand it such a max torque over speed diagramms describes static states.

 

[Aneta]

6 minutes ago, Chriull said:

I don't see how the negative quadratic term could be caused by inductance? Could be an approximation for the di/dt for the voltage drop? But since it's with omega² it would be reasonable to describe the inertia of the wheels mass which depends on d omega/dt?

But still these all would be dynamic effects for burden changes, and as i understand it such a max torque over speed diagramms describes static states.

 

I guess this is just an empirical approximation to account for distorted waveforms.

[RockyTop]

It can certainly be explained.

However, I do believe that we would have to enter quite a few new, ugly factors. ( pole spacing, number of poles , electrical inertia, electrical and mechanical harmonics and reactions with frequencies) You know?! Ugly stuff. 

[DjPanJan]

So when I summarize it for people like me mentally less powerful.
When I give myself a maximum speed of 40kmh / 25 miles so my ks16X does not hit my chin on the ground?

[RockyTop]

8 minutes ago, DjPanJan said:

So when I summarize it for people like me mentally less powerful.
When I give myself a maximum speed of 40kmh / 25 miles so my ks16X does not hit my chin on the ground?

Yes, That will work. Most people stay below 25mph anyway. 

Edited November 30, 2019 by RockyTop

[EUC Custom Power-Pads]

Yes, but only with full battery. 

Battery 100 % = 50 km/h max. - set tilt back 40 km/h = 10 km/h buffer

Battery 33 % = 45 km/h max. - set tilt back 40 km/h = 5 km/h buffer

Battery 25 % = 40 km/h max. - set tilt back 40 km/h = 0 km/h buffer

Just an example from one of my trips. The speed limit changes continuously due to factors such as gradient, wind, rider weight and personal driving style. Not only the speed is decisive, but also the still available engine power. 

 

Edited November 30, 2019 by buell47

[skautas2003]

is there any results from this case analysis?

[who_the]

On 11/19/2019 at 11:21 AM, redfoxdude said:

Hey everyone,

On the way to work, I experienced an overlean on my KS16X, firmware v1.07. ... according to my EUC world tour data, I was going around 29-30mph on a slight downhill, when I just leaned slightly too hard. I had probably about a second from when it started the triple beep warning to when I was on the ground. I did capture it on my helmet cam, here. For the record, I weigh around 190 pounds with gear.

Glad you were geared up and that you're OK. Thanks for sharing your data to start this useful and enlightening discussion.

A friend on FB shared a Wheellog screenshot after a no-load lift test of his 16X. Maxed out at 38 mph. That seems insanely low for a supposed ~31 mph wheel.

Could a no-load lift test on a wheel be used to "calibrate" a function in EUC World that could be used to calculate an approximate safety threshold in real-time? Maybe if a riding weight parameter is stored as well? @Seba

[Chriull]

On 11/30/2019 at 11:40 PM, who_the said:

A friend on FB shared a Wheellog screenshot after a no-load lift test of his 16X. Maxed out at 38 mph. That seems insanely low for a supposed ~31 mph wheel.

Could it be that he measured at low battery voltage? Then the value would be quite normal:

	kv (v/km/h)	v empty battery (km/h)	v full battery (km/h)	Max Tiltback Speed (km/h)	factor empty battery	factor full battery
KS16B	1,46	36,16	46,03	30	1,21	1,53
KS16S	1,32	40,00	50,91	35	1,14	1,45
?KS16X?	1,10	60,00	76,36	50	1,20	1,53

Just the KS16S would be with a little lower than the others. If one calculates the KS16X with the same lower factor as the KS16S:

	kv (v/km/h)	v empty battery (km/h)	v full battery (km/h)	Max Tiltback Speed (km/h)	factor empty battery	factor full battery
KS16B	1,46	36,16	46,03	30	1,21	1,53
KS16S	1,32	40,00	50,91	35	1,14	1,45
?KS16X?	1,16	57,14	72,73	50	1,14	1,45

Just if he got the 38 mph ~ 60 km/h with full batteries there is something wrong/very strange.

PS.: At ~30% charge the tilt back speed gets lowered by firmware - so in reality the "factor empty battery" is never reached...

Idid this sheet with 3.3V for the empty cell and 4.2 for the full cell - but for the KS16X low battery is somewhere about ?3.15V? instead of 3.3V as with the KS16B/S...

Quote

Could a no-load lift test on a wheel be used to "calibrate" a function in EUC World that could be used to calculate an approximate safety threshold in real-time? 

That's how GW has their 3rd alarm implemented. It's linear with battery voltage/charge. So for the 84V Monster for example at 100%==84V at 55 km/h and at 10% charge down at 43 km/h.

As lift cut off speed is linear with battery voltage/charge that is exactly this implemented.

On 11/30/2019 at 11:40 PM, who_the said:

Maybe if a riding weight parameter is stored as well? @Seba

Riding weight is just one parameter for a burden, as also wind, inclination, etc is. But one could choose the factor between alarm and battery voltage so everyone can get his personal safety margin.

To really regard the burden one could additionally lower this alarm speed by the reported current (with again some factor) - this would then reproduce more or less the max torque over speed limit of the wheel!

Edited September 9, 2020 by Chriull
ks16x Lift cut off speed should be 67 km/h, not 76 km/h...

[Mike Sacristan]

I just tested with MSX, 18XL and 16X.

18XL 
Battery 76.7v
Speed 62.8 kmh

16X
Battery 78.8v
Speed 61.7 kmh

MSX
Battery 78.03V
Speed 71.4 kmh

Which makes sense as the MSX is at least 10 kmh faster.

Edited December 1, 2019 by Mike Sacristan
changed decimal sign

[Chriull]

1 hour ago, Mike Sacristan said:

I just tested with MSX, 18XL and 16X.

18XL 
Battery 76.7v
Speed 62.8 kmh

16X
Battery 78.8v
Speed 61.7 kmh

MSX
Battery 78.03V
Speed 71.4 kmh

Which makes sense as the MSX is at least 10 kmh faster.

Cell Voltage (V)		Empty	10%	30%	Full
GW/KS16B/S		3,30	3,39	3,57	4,20
KS16X		3,15	3,26	3,47	4,20
KS18XL		3,00	3,12	3,36	4,20
Battery Voltage 16s (V)	KS16B/S	52,80	54,24	57,12	67,20
Battery Voltage 20s (V)	GW	66,00	67,80	71,40	84,00
Battery Voltage 20s (V)	KS16X	63,00	65,10	69,30	84,00
Battery Voltage 20s (V)	KS18XL	60,00	62,40	67,20	84,00
	kv ((km/h)/V)	lift cut off speed 10% charge (km/h)	lift cut off speed 30% charge (km/h)	lift cut off speed full battery (km/h)	Max Tiltback/3rd Alarm Speed (km/h)	factor at 10% charge	factor at 30% charge	factor full battery
KS16B	0,68	37,15	39,12	46,03	30		1,30	1,53
KS16S	0,76	41,09	43,27	50,91	35		1,24	1,45
KS16X	0,78	50,97	54,26	65,77	50		1,09	1,32
KS18XL	0,82	51,09	55,02	68,78	50		1,10	1,38
MSX	0,92	62,04	65,33	76,86	55			1,40
MSX	0,92	62,04	65,33	76,86	45,67		1,43
MSX	0,92	62,04	65,33	76,86	43	1,44

 

OK - Put your numbers into the sheet. Afair KS16X had 3,15V(63V low pack voltage) as low cell voltage, the KS18XL 3.0V(60V low pack voltage)?

I also made a column for 30% charge - were the KS wheels start speed throttling, so any safety margin factor with the higher speed limits would be senseless for empty batteries.

For the MSX 84V (GW) i entered the values for the 3rd Level Speed Alarm - as one sees, here one has a constant safety margin factor over battery level!

kv i changed from V/(km/h) to (km/h)/V as imho the normal units are rpm/V for kv?

 

If my numbers are right, the KS16X/18XL lift cut-off speeds come dangerously near to max tilt-back speed at 30% charge!

Ps.: this factor = Lift cut off speed at this battery voltage / max tiltback-3rd alarm speed

Edit: First Changes to the table: Added 10% charge, as this is the charge for the 43 km/h 3rd alarm for the MSX 84V

Edited December 1, 2019 by Chriull

[Chriull]

3 hours ago, Chriull said:

To really regard the burden one could additionally lower this alarm speed by the reported current (with again some factor) - this would then reproduce more or less the max torque over speed limit of the wheel!

Since i had an wheellog log of a MSX-84V ride (ending with an overlean) i added the 3rd level alarm to the charts. In the current over speed diagram are the two limit lines in black and grey for the max current (torque) over speed for the highest on lowest battery voltage in the log and the dark red and red line for the 3rd level alarm at these voltages.

Ps.: the time axis labels are a bit unfortionate - they mean "day hour:minute"...

[RockyTop]

Interesting. However most of this would assume that the motors, Motor controllers, application and programs are the same. Even with these results one unit could be much stronger with a heavy load close to cut out. The biggest determining factor could easily be the motor controller when it comes to thrust at close to cut out. The motor controller  and software would also make a big difference when it comes to recovering from a near loss of control.
 

Yet if the motors are the same and you are giving it everything it tends to even out. Given the more powerful boards of the MSX , the MSX can get closer to givings everything that the battery has. The battery then becomes the weak point. 

Edited December 1, 2019 by RockyTop

[Mono]

@Chriull, I don't understand the meaning of kv. I could have a 20V system which performs identical to a 100V system that would now have completely different values for kv, am I right? Shouldn't kv at least be normalized by the number of cells in series?

[RockyTop]

It is not taking parallels into the equation. Although the MSX and 18XL do have the same battery configuration. 

km/h  per cell would be more useful. 

Edited December 1, 2019 by RockyTop

[Chriull]

 

1 hour ago, Mono said:

@Chriull, I don't understand the meaning of kv. I could have a 20V system which performs identical to a 100V system that would now have completely different values for kv, am I right? Shouldn't kv at least be normalized by the number of cells in series?

 

1 hour ago, RockyTop said:

It is not taking parallels into the equation. Although the MSX and 18XL do have the same battery configuration. 

km/h  per cell would be more useful. 

kv is just in this list as helper for the calculations - it's just the motor constant showing the speed/rotations per Volt.

The "result" of this table is the factors between (battery voltage depending) lift cut-off speed and maximum riding speed (Tiltback for KS, 3rd speed alarm for GW).

This factor shows about a safety margin.

So if one looks at the KS16X row the factor at 30% charge is 1,09. Telling us that by tiltback 50 km/h would be still allowed, but with this battery voltage the lift cut-off speed is only 9% higher at 54,26 km/h! That's about no safety margin and one should never reach this 50 km/h tiltback at 30% charge - one will just faceplant before.

The KS16S had at 30% still ~24% reserve, the MSX has evertime ~40% reserve.

And this reserve "is only true" in a no load situation - that's why i added the MSX current(torque) over speed diagramm with the 3rd alarm shown and the current(torque) over speed limit.

 

[Mono]

Concluding from these data: in some utopian future, the maximal speed will be limited to half of the no-load speed. That should make overleans very rare and well deserved ceremonies. (It would also mean that the power output of the motor is monotonously increasing over the entire range of rideable speeds, nice).

Edited December 1, 2019 by Mono

[who_the]

8 hours ago, Chriull said:

22 hours ago, who_the said:

A friend on FB shared a Wheellog screenshot after a no-load lift test of his 16X. Maxed out at 38 mph. That seems insanely low for a supposed ~31 mph wheel.

Could it be that he measured at low battery voltage? Then the value would be quite normal:

He tested with 85% battery.

[Aneta]

9 hours ago, Chriull said:

So if one looks at the KS16X row the factor at 30% charge is 1,09. Telling us that by tiltback 50 km/h would be still allowed, but with this battery voltage the lift cut-off speed is only 9% higher at 54,26 km/h! That's about no safety margin and one should never reach this 50 km/h tiltback at 30% charge - one will just faceplant before.

This is a brilliant analysis and vividly demonstrates that every EUC rider must understand these basic things - "or else".

What is absolutely shocking is that none of these things are explained by manufacturers themselves, WE must discover them. No EUC comes with a manual that explains them. Some have brief instructions on a sticker on the pedals, like stand straight; some have a couple of paragraphs in the printed manual. NONE explains basic EUC safety things like max thrust vs. speed; thrust margin; effects of incline, acceleration, and wind; reduction of thrust with falling voltage; etc. If this is not called barbaric or neanderthal, I don't know what to call it.

[Chriull]

On 11/30/2019 at 11:40 PM, who_the said:

A friend on FB shared a Wheellog screenshot after a no-load lift test of his 16X. Maxed out at 38 mph. That seems insanely low for a supposed ~31 mph wheel.

 

20 hours ago, Mike Sacristan said:

I just tested

...

16X
Battery 78.8v
Speed 61.7 kmh

 

 

12 hours ago, who_the said:

He tested with 85% battery.

That fits quite well together! Although it's hard to get the charge percentage/voltages together - i'm not really sure if 0% is now 3V, 3.15V or something in this range by now for the KS16X... Afair it was 3.15V?

So it's understandable why KS introduced the speed reduction for lower battery charges - the margin is just too low. (if my numbers are right and there is no typo hiding in my sheet...  )

4 hours ago, Aneta said:

NONE explains basic EUC safety things like max thrust vs. speed; thrust margin; effects of incline, acceleration, and wind; reduction of thrust with falling voltage; etc.

This forum would be a great place to start with such a summary - i've thought of doing so since i posted my "anatomy of an overlean" topic. But it's a bit hard for me to write easy understandable "prose/lyrics" with simple concise graphs instead of tables, formulas and graphs with some conclusions that are "hard" to follow... ;(

13 hours ago, Mono said:

Concluding from these data: in some utopian future, the maximal speed will be limited to half of the no-load speed. That should make overleans very rare and well deserved ceremonies. (It would also mean that the power output of the motor is monotonously increasing over the entire range of rideable speeds, nice).

Will be a great future! But imho GW with the 3rd alarm is almost there - just the beeping is imo easy to be overheard at higher speeds. The margin as seen in the graph above for the MSX 84V is really nice (from the dark-red 3rd alarm to the black "torque" limit and from the light-red 3rd alarm to the corresponding grey "torque" limit). If they now would make the 3rd alarm motor current dependend it would be about perfect. And customizable, and optionaly as tiltback, ...

[Mike Sacristan]

Tested 16X again now at 74.5V.
58.3 kmh.
At that voltage I get soft tiltback at 42-44 kmh and can press up to 45 kmh but I have gotten 4 beep 88% alarms so I respect the soft tiltback.

When I was riding just now I hit 45.4kmh on a straight at 75.33V. No beeps and could hardly feel soft tiltback.
-2C and 3m/s wind.

I ride a lot with an 18XL rider. He says his 18XL does soft tiltback earlier and earlier as the ride progresses. I have never seen him do 50kmh at below half battery but I have seen him capped to 42kmh at slightly above 30%. Once he overleaned the wheel at 60% battery and 49.09 kmh. And this was on a softly inclined road. He was hanging hard on the tiltback.

16X 1.05FW and 18XL behaved similarly from what I could see.
16X 1.07FW keeps 2-3kmh lower speed as a buffer compared to FW 1.05.

I think the soft tiltback implementation on the 16X is good enough. Especially in FW 1.07 with the 4 beep 88% alarm. I hit the 88% alarm today while accelerating to full speed. Soft tiltback and directly after 88% alarm. The top speed achieved was lower than the top speed I would have achieved with a slower acceleration.

I wonder if the Kingsong no-load lift speed is simply limited so that it does not spin beyond X speed.
Gotway seems to simply let it spin more.

After riding quite a few kmh on the MSX and 16X I can say that the 80% alarm for the MSX and the soft tiltback for the 16X work similarly. After the soft tiltback comes the 88% alarm. So reasonably the 16X soft tiltback should be at ~80%.

This doesn't mean I have reserves to go 25% faster though. That last power gets eaten up quickly and we need it for other stuff... like bumps, wind, inclines, etc.

When viewing the speed reduction thread and especially the graphs for the 18XL I was also very concerned regarding the voltage vs the still top speed of 50kmh. 

I did a FW 1.07 test on the 16X a couple of weeks ago. I got it down to 63.0V and was still doing a little over 20kmh.
According to the manual 63.0V is 0%. And 83V is 100%.
This corresponds with previous discussion regarding voltage cutoff for the wheels. 3.3V for Gotway, 3.15V for 16X and 3.0V for 18XL/L.

 

 

[Chriull]

1 hour ago, Mike Sacristan said:

Tested 16X again now at 74.5V.
58.3 kmh.
At that voltage I get soft tiltback at 42-44 kmh and can press up to 45 kmh but I have gotten 4 beep 88% alarms so I respect the soft tiltback.

When I was riding just now I hit 45.4kmh on a straight at 75.33V. No beeps and could hardly feel soft tiltback.
-2C and 3m/s wind.

So the KS16X has no fixed tiltback speed of 50 km/h from 100% downto 30% charge?!

It has some reductions already inbetween, as it seems. Your voltages seem to be about a bit above 50% ((84+63)/2=73,5) and tiltback is already a bit lower than the 50 km/h.

Does someone know the actual speed reductions from the KS16X? Then i could update my above posted sheets - and they would not look as dangerous anymore....

.... sorry i did not really follow all discussions about this ...

[redfoxdude]

4 hours ago, Chriull said:

So the KS16X has no fixed tiltback speed of 50 km/h from 100% downto 30% charge?!

It has some reductions already inbetween, as it seems. Your voltages seem to be about a bit above 50% ((84+63)/2=73,5) and tiltback is already a bit lower than the 50 km/h.

Does someone know the actual speed reductions from the KS16X? Then i could update my above posted sheets - and they would not look as dangerous anymore....

.... sorry i did not really follow all discussions about this ...

I don't recall the actual reductions, but the tiltback speed of 50 km/h is supposed to be fixed until ~70V I believe, and then it starts reducing proportionally with the voltage. I think v1.06 started reducing the tiltback speed at 72V, and v1.05 at 67V.

Going back and carefully playing around with the KS16X with ~90% battery, I feel like I could very easily overpower it even at ~38 km/h on flat ground. I can reproduce @chrisjunlee's ankle flicks, and it seriously feels like there is hardly any torque/thrust margin. It just gives like a hinge.

This really is a super fun wheel, but yeah, I'm not buying King Song's safety claims, not as the wheel currently is. The 84V Nikola has more torque/thrust margin, and I feel much more safe and secure on my 100V Nikola +.

Edited December 2, 2019 by redfoxdude

[Mike Sacristan]

I did some pretty thorough testing here: