KS16X Cutoff/Overlean - 29-30mph

[redfoxdude]

Hey everyone,

On the way to work, I experienced an overlean on my KS16X, firmware v1.07. I do think this is somewhat related to the previously documented reduced torque at higher speeds, and I certainly hope it could be improved with firmware updates. According to the log, I wasn't even close to the max power output, so maybe King Song can work on tweaking their algorithm to allow more power output at higher speeds. I have had a feeling that their work at reducing high speed oscillations may be related to this reduced power output, but that's mere speculation...

That said, 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.

Needless to say - good protective gear is essential! I walked away with only minor scrapes and bruises. I wear a motorcycle jacket, knee pads, and gloves, all with CE level 1 protection. Worth every penny. For now, I've reduced my warning and tiltback speeds a bit to give me more margin. I don't recall even feeling the tiltback.

Below is a capture from the log - the data point right before the crash.

Edited November 20, 2019 by redfoxdude
Added my weight!

[Chriull]

3 minutes ago, redfoxdude said:

On the way to work, I experienced an overlean on my KS16X, firmware v1.07. I do think this is somewhat related to the previously documented reduced torque at higher speeds, and I certainly hope it could be improved with firmware updates.

Every E motor has reducing torque with increasing speed.

There is nothing that can be done against it. One can just shift limits by choosing other battery/motor combination

5 minutes ago, redfoxdude said:

According to the log, I wasn't even close to the max power output, so maybe King Song can work on tweaking their algorithm to allow more power output at higher speeds.

Maximum power is only available at half of maximum no load speeds.

For overleans (not enough torque at higher speeds) the beforementioned torque over speed limit of e motors is relevant.

For more reading/details:

 

Great you made it without severe injuries!

[Aneta]

15 minutes ago, redfoxdude said:

I do think this is somewhat related to the previously documented reduced torque at higher speeds, and I certainly hope it could be improved with firmware updates.

It's not somewhat related, it's 100% absolutely related to EUC Physics 101 that has been discussed quite intensely recently. See e.g. this and links from there:

The manufacturer cannot do anything to circumvent the laws of physics.

Edited November 19, 2019 by Aneta

[Aneta]

17 minutes ago, redfoxdude said:

I don't recall even feeling the tiltback.

When speed is too close to "terminal", the wheel physically cannot do a tiltback, because for that it needs to spin faster under the same load to get ahead of the rider, but it can't, because the thrust margin is already pegged at near zero. Fully expected.

[redfoxdude]

@Chriull Yes, the back emf certainly reduces the available torque as speed increases. I would have hoped that King Song selected a battery/motor combination that yields enough torque at the desired top speed to maintain a good safety margin.

@Aneta I think you misunderstand. I feel that the hardware was not actually operating at or quite near its limit, and that, if the firmware commanded, there could have been more thrust. Utilizing more of the wheel's capable power is not circumventing the laws of physics. Thanks for your concern for my well being, though.

Edited November 19, 2019 by redfoxdude

[Aneta]

1 minute ago, redfoxdude said:

I am confident that the hardware was not actually operating at or quite near its limit, and that, if the firmware commanded, there could have been more thrust.

What exactly is this confidence backed up with, in technical terms (like log data)?

[fryman]

Sorry this happen and I'm glad you are okay.  You are doing the right thing by lowering your tilt-back and warning and I would suggest 26mph.

[Chriull]

16 minutes ago, redfoxdude said:

I would have hoped that King Song selected a battery/motor combination that yields enough torque at the desired top speed to maintain a good safety margin.

Choosing a certain tiltback speed is a compromise - there always will be cases it cancels too "early" and cases one overleans without tiltback ...

Time will come (hopefully) when more sophisticated warnings will be implemented.

16 minutes ago, redfoxdude said:

I feel that the hardware was not actually operating at or quite near its limit, and that, if the firmware commanded, there could have been more thrust.

How to feel such if its near the limit? To get a real picture one should draw a current over speed diagram.

[xorbe]

Dang, sudden beeping heart attack and dropped you on the floor.  Glad you're okay!  My 16X w/1.05 is configured to last alarm/tiltback at 25 mph iirc.  What version control board you have ... I hear that you were using BT music.

Edited November 19, 2019 by xorbe

[skautas2003]

Could you write you weight please? And if possible with and without gear?

Is everything ok with your wheel after crash is it functioning ok?

[Aneta]

51 minutes ago, redfoxdude said:

I would have hoped that King Song selected a battery/motor combination that yields enough torque at the desired top speed to maintain a good safety margin.

For given motor/controller/battery, that's just not arbitrarily possible (unless the manufacturer is intentionally throttling the system down). Back EMF is only determined by speed, nothing can be done about it. Battery voltage is whatever it is, nothing can be done about it. Windings resistance is fixed, nothing can be done about it. Then the max phase current at 100% duty cycle is I_max = (V_batt - V_emf)/R_windings and determines the max thrust the motor can output at this speed. The air drag is determined by your body position and velocity, nothing can be done about it (other than "chooching"/ducking to reduce Cd). The difference is the available thrust margin, and nothing can be done to make it higher.

The only way is to have a wider motor, with wider magnets/stator, to have proportionally more torque across the same speed range. But then riders will be approaching the terminal speed closer as they feel safer, doing more aggressive accelerations closer to the edge... and the end result will be the same. It's a well known phenomenon in extreme sports - "risk homeostasis": https://www.google.com/search?q=risk+homeostasis&oq=risk+homeostasis

If the ice on partially frozen river grows thicker, risky people will simply go farther towards the edge and snap the ice where it has the same thickness as before.

The only cure is self-discipline, period.

[redfoxdude]

1 hour ago, skautas2003 said:

Could you write you weight please? And if possible with and without gear?

Is everything ok with your wheel after crash is it functioning ok?

Ah, good point. I'm ~176lb, probably ~190lb with gear. Everything seems to be good! This was less than halfway into my commute - RIP. Haha, but I made it the rest of the way just fine, keeping to less than 26 or 27 mph or so.

@Aneta I am actually an electrical engineer, but your post is nice and informative for anyone with curiosity to learn more! The comment you quoted is exactly to your points, that I had hoped King Song's motor and battery selection for the KS16X would yield a high enough "thrust margin" to safely sustain the advertised "maximum cruising speed." Again, the point that I am trying to make, is that I have been wondering if the system is throttled down, and would be capable of a higher safety margin with further tuning in the firmware. Wouldn't that be nifty? I was well within the operating design limits of the wheel, and it's on King Song to continue to improve their product! This is partially for folks on the forum to see, but also, this is for King Song to see, in hopes that they would act to improve.

That said, something that I do quite like about Gotway's 100V wheels is their safety margin when you are riding ≤ 31mph. If King Song made wheels with the same capability, but kept their usual top speed of ~31mph, the safety margin would be awesome. Folks would probably complain that they are artificially limiting it and blah blah blah, but I think that would be stellar. If the Nikola+ was more comfortable and nimble, I would choose it any day over the KS16X, since I don't like to ride past 30mph pretty much ever. Maybe I just need to try the medium mode again? It's really noticeably more fatiguing to ride the Nikola+ for me. Partially the shape/width, but maybe the stiff pedals are contributing more than I thought. But, that's off topic.

Edited November 19, 2019 by redfoxdude

[redfoxdude]

1 hour ago, xorbe said:

Dang, sudden beeping heart attack and dropped you on the floor.  Glad you're okay!  My 16X w/1.05 is configured to last alarm/tiltback at 25 mph iirc.  What version control board you have ... I hear that you were using BT music.

Thank you! It was exciting, haha! It's the latest I think, 1.5? I replaced it during the investigation into the pedal flick thing. Haven't had any of the bluetooth related symptoms, I don't think. 

And haha, that music playing caused YouTube to warn me about the copyrighted content in my video. 

[Mono]

2 hours ago, redfoxdude said:

According to the log, I wasn't even close to the max power output

What is the max power output of the 16X at 31mph?

[Aneta]

19 minutes ago, redfoxdude said:

I was well within the operating design limits of the wheel

What, in numbers, are the operating design limits of the wheel at this speed? And what was the margin over the limit you were trying to achieve?

[Aneta]

By the way, the numbers for current and power as shown in OP seem to be suspiciously low. As can be seen from Motor Sumulator, the mechanical power at speeds ~50kph is around 2kW, while battery power is in the upper 2000's, closer to 3kW:

https://www.ebikes.ca/tools/simulator.html?batt=cust_84_0.2_20&cont=cust_100_200_0.03_V&wheel=17i&frame=cust_1_0.01&hp=0&blue=Lbs&motor=M3540

10 amps/800W seems to be ridiculously low.

I've already raised this concern about possible bug in 16X telemetry here:

but people don't seem to care.

[RockyTop]

@redfoxdude I am glad you are OK. Thanks for giving us this information. 

One more question. What temperature was it outside?

I had a similar cut out on a MSX. In my case the battery was low (35-40%)  I weigh 220 , I was going up a slight hill for an extended amount of time at 30 mph and the final straw was that it was 35 f degrees outside. I don't believe the battery could keep up with the demand at that temperature. 

Edited November 19, 2019 by RockyTop

[Mono]

8 minutes ago, Aneta said:

but people don't seem to care.

it's because they just multiply by pi themselves 

[Aneta]

7 minutes ago, RockyTop said:

I weigh 220 , I was going up a slight hill for an extended amount of time at 30 mph

What was % grade on that hill? (GPS coords would be sufficient - one can easily measure grades in Google Earth)

[RockyTop]

8 minutes ago, Aneta said:

What was % grade on that hill? (GPS coords would be sufficient - one can easily measure grades in Google Earth)

An average of 3%  for a mile. 

As I said, everything was going against me at the time. I never heard a beep ....... but I had my ears covered because of the cold. Now I pay attention to all of the factors and I got a much loader beeper flush with but outside the wheel. I am not sure that it ever did beep but I know that I would not have heard it anyway. 

[Aneta]

21 minutes ago, RockyTop said:

An average of 3%  for a mile. 

As I said, everything was going against me at the time. I never heard a beep ....... but I had my ears covered because of the cold. Now I pay attention to all of the factors and I got a much loader beeper flush with but outside the wheel. I am not sure that it ever did beep but I know that I would not have heard it anyway. 

Well, 3% is nothing to sneeze at at these speeds, because for 100kg total weight that's 3 kilos of extra thrust needed which the motor may not have at these speeds.

Also, if 3% is the average slope, the road can easily have double or triple of that in some spots, it won't be even noticeable to the eye (i.e. 3% is only a rise of less than half-inch in one foot forward).

Edited November 19, 2019 by Aneta

[Jason McNeil]

2 hours ago, redfoxdude said:

hoped King Song's motor and battery selection for the KS16X would yield a high enough "thrust margin" to safely sustain the advertised "maximum cruising speed." Again, the point that I am trying to make, is that I have been wondering if the system is throttled down, and would be capable of a higher safety margin with further tuning in the firmware. Wouldn't that be nifty? I was well within the operating design limits of the wheel

Thanks for sharing this information. When you sent the email, with all this wonderful telemetry & direct video evidence, I thought it had to be shared for everyone to see the very real risk potential of this narrow power safety band.

Either KS need to get their firmware bulletproofed, where there is a sufficient >5-10kph speed/rpm margin before falling off the torque cliff, or they have to reduce the max speed of the Wheel to 45kph. There's no extenuating circumstances here: you're on v1.07, the battery level is >80%, temperature is not a factor, nor are you a sumo wrestler pushing the 320lb weight class. Guess this necessitates another safety advisory tomorrow   

Edited November 20, 2019 by Jason McNeil

[Chriull]

15 hours ago, Aneta said:

Well, 3% is nothing to sneeze at at these speeds, because for 100kg total weight that's 3 kilos of extra thrust needed which the motor may not have at these speeds.

At a flat road there is just thrust needed for acceleration friction, air drag and self balancing. Zero thrust needed to lift some weight.

So there is no "three kilos extra" for an 100kg rider at a 3% slope. "Just" the thrust is needed to lift the 100kg + the wheel against gravity at a certain rate, which is determined by the slope percent and speed.

Edited November 20, 2019 by Chriull

[redfoxdude]

1 hour ago, Aneta said:

10 amps/800W seems to be ridiculously low.

I agree that it seems very low. I neglected to mention, however, that I was on a slight downhill. Still, seems low.

1 hour ago, RockyTop said:

@redfoxdude I am glad you are OK. Thanks for giving us this information. 

One more question. What temperature was it outside?

Thank you. Protective gear is well worth the money, that's for sure! Ah, another good question! It was probably something like 50 or 60 degrees Fahrenheit, so I don't think that was the issue.

22 minutes ago, Jason McNeil said:

Thanks for sharing this information. When you sent the email, with all this wonderful telemetry & first-hand video evidence, I thought it had to be shared for everyone to see the very real risk potential of this narrow power safety band.

Either KS need to get their firmware bulletproofed, where there is a sufficient >5-10kph speed/rpm margin before falling off the torque cliff, or they have to reduce the max speed of the Wheel to 45kph. There's no extenuating circumstances here: you're on v1.07, the battery level is >80%, temperature is not a factor, nor are you a sumo wrestler pushing the 320lb weight class. Guess this necessitates another safety advisory tomorrow   

Absolutely! If King Song wants to continue to claim 50kph is the max speed, something has to give. Again, thank you for always advocating for us with the manufacturers, and your responsiveness with important issues like this!

[Mike Sacristan]

4 hours ago, redfoxdude said:

Hey everyone,

On the way to work, I experienced an overlean on my KS16X, firmware v1.07. I do think this is somewhat related to the previously documented reduced torque at higher speeds, and I certainly hope it could be improved with firmware updates. According to the log, I wasn't even close to the max power output, so maybe King Song can work on tweaking their algorithm to allow more power output at higher speeds. I have had a feeling that their work at reducing high speed oscillations may be related to this reduced power output, but that's mere speculation...

That said, according to my EUC world tour data, I was going around 29-30mph, 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.

 

Damn man. Sorry to hear. Glad you made it out 100% okay and hope your wheel is fine too.

I saw that you were at around 80% battery. ~79v. At that battery level I can just reach 48kmh with my riding weight of 69kg. Then I am on soft tiltback. I have also pressed medium-aggressively at this speed to provoke the 4 beep / 88% alarm. 

I see that your watt numbers at the beginning of the ride are a bit higher which looks more normal.

I feel that the 16X constantly throttles as we are riding and I wonder if it throttling in the middle of the acceleration/lean that you were doing.

Even at very low voltage when I was playing around I never felt close to an overlean but I am a pretty small guy. This is very worrying though... and that was a nasty fall that just came out of nowhere. You should try running your 16X down to low voltage say for instance 20% battery or 67v. At this battery level you will be limited at 30-34kmh and maybe then you could look closer at the tiltback / soft tiltback / hard tiltback and make sure that it is working correctly. If you haven't already. Seems like you didn't have much margin before the overlean and the whole point of 1.07 is keeping power in reserve to avoid this kind of scenario.

Good job taking it like a champ! Looked for a muscle-arm emoji but couldn't find one! You know what I mean though.