!!! Firmware or hardware problems with KS18L, wheels locking up while trolleying

[winter]

 

Edited August 18, 2018 by nte
added additional information

[em1barns]

2 minutes ago, atph said:

Same here and that's not conforting.

My wild guess is that things could go horribly wrong when it reached 100%... in standard usage like waiting stand still with trolley open, I did not expect to reach such a high CPU usage.

[esaj]

2 hours ago, em1barns said:

What surprised me (may be unrelated, maybe not), is that when taking my wheel on trolley, and not moving, CPU shows 80+ % utilization (in app > More > Details).

Any idea what happens when it reaches 100%?

Don't know how they measure the utilization, but if there's "more to do" than what the processor has time for, it would start to miss doing things "timely", if it was driving the motor, this could be a bad thing (if it starts missing switching the motor phases or such). The MCUs (microcontroller units) used in thesee are a bit different than you "normal" computers (well, technically you could say they're not even computers) in that they don't run an "operating system" like your computer or phone, but rather the code runs "directly on hardware". Kinda hard to explain without going into a lot of technical terms...  For example, the Bluetooth data might become more intermittent or get garbled all together.

[winter]

 

Edited August 18, 2018 by nte

[Kael]

Fairly simple to explain: the CPUs have to run in real time, with minimal overhead, hence no operating system. If they have input to process, then stuff like Bluetooth might not be given the time it needs. 

[Shad0z]

1 hour ago, hal2000 said:

Did they know what to do? And if yes, did it happen again? What mileage do they have on their wheels?

They got it a week before it happened. They both knew to disconnect batteries. And nothing happened. But they got sent new boards from ks

[TomOnWheels]

2 hours ago, atph said:

Same here and that's not conforting.

Simply you are confusing a standard PC where 80% CPU is a lot, but for a RTS system this is quite normal and should be stable during usage... I'm even asking myself why they are giving that details because it should not move a lot... Also the way a standard CPU is working is quite different from how a micro-controller... Nothing surprising here.

[US69]

2 hours ago, Shad0z said:

They got it a week before it happened. They both knew to disconnect batteries. And nothing happened. But they got sent new boards from ks

4 hours ago, Shad0z said:

my freinds ordered together. 3 in total, 2 of them had the issue. But all 3 of them had turned off the lift sensor feature because they dont trust it. so if its a lift sensor issue how can it happen when they all turned it off? o_O

 

Just to get things right and clear up the statistics that KS, @tinawong ,me and the resellers from all over the world are having and collecting!

Where did they order, at which reseller? When please  did they get new boards from „KS“? As this is i bit irritating....as the first board where this happens not even reached KS factory? 

And btw. i am a bit irritated that your friends....going through all you other posts, just 3 of them jumped over to KS? While Tesla, Msuper X just got such a better perfomance....as you always stated?.......

Edited July 20, 2018 by US69

[esaj]

53 minutes ago, TomOnWheels said:

Simply you are confusing a standard PC where 80% CPU is a lot, but for a RTS system this is quite normal and should be stable during usage... I'm even asking myself why they are giving that details because it should not move a lot... Also the way a standard CPU is working is quite different from how a micro-controller... Nothing surprising here.

This is why I was wondering how they even measure it... for a "normal" operating system scheduler, it's far more trivial to calculate the free "time slots"/processes waiting for CPU time, but what do you do with a real-time system? All I could think of was that they've set a hard limit (1ms, 100µs or whatever, "real-time hard agent" or whatever the correct technical term is, with a strict time limit, vs. a "soft agent" which has some target time, but no actual "penalty" for missing it) for a single main loop, and compare how much time is left "free" at the end of it. If the processing would fall back a lot while driving the motor, that would be a bad thing (it could start missing phase changes for the motor or dropping dead-time insertions to too low, causing a bridge shoot-through), but I'd hope they'd dealt with this "intelligently" (ie. all the critical functions are interrupt timer based and the rest can fall behind / be skipped if need be). Not that I've ever really worked with any time-critical embedded systems, my background's in "traditional" software running on top of OS's.

Edited July 20, 2018 by esaj

[US69]

Just to -perhaps- cool down things a bit....

Collecting stats from United States, Russia, France, and rest of Europe, where the main activities of selling the 18 L happend.

There are now....dont hammer me down on one or two numbers....7-8 known facts of this „trolleying than Lock“ behavoir!

As you can imagine, the poeple just dont report on FB or here in the Forum, they also contact there reseller, at least as long as they have an „legal“ one and did not do a grey import.  All these cases, i reported as much as possible infos to KS to adjust there testing procedures  to found out the real reason behind this! Any , i will call it „this is spreading like nuts“ reports without any facts dont help at all....and Sorry, i am here abit longer to see what might be a real report or a „i hang myself on to this train“.....

I have personally contact to my FRIENDS from KS Thailand and Thomas Hoon/wheelies , which sold about 30plus 18L wheels and did not have ONE of this behaviour!

To get it right: 7-8....or even 8-10....are exact these numbers to much! KS is working hard time to reproduce this fault....unfortunately until now without any luck....but believe me, they are working hard on it! 

Any kind of „assumption“...like CPU usage ...hmmh perhaps might help...perhaps might get others only confused, as this might mean just nothing.....KS has taken just a lot of infos new into their app system...like CPU usage, motor temperature, pressure sensors, hall sensors....

I for my part can only report they have been thousand percent honest to me ....and if they would have any clue whats going on...they allready would have told us!

I personally asked all my friends...Ks Thailand, Thomas from the wheelies, all my German 18L riders to first to shut down the handle sensor.....no report of lock from there, not before my advice, not after!

So my main point is: Let KS analyse there thing! they ARE ON IT!... I dont like comparisons...but i personally cant remember such an open handling of such a fault in EUC world before....can you?

@tinawong

 

 

Edited July 20, 2018 by US69

[tim716]

Another 18L is dead today in Russia. Just two kilometers from new. Obviously it's not "locked why trolleying" but some new issue I haven't heard of before.

The owner rode two kilometers from the store to a subway station, then turned the wheel off and trolleyed it around in the subway in a TURNED OFF state. When he tried to turn the EUC back on, it behaved like this:

https://www.youtube.com/watch?v=mmMRzJvNOks

The firmware version is unknown, but unlikely 1.07, since he didn't have much chance to update. He didn't report any heating up. The trolley sensor was ON. 

Multiple turns on and off, reconnecting batteries, firmware update to 1.07 (he can connect to the whell and update the firmware) didn;t help.

Any comments or suggestions?

 

[US69]

9 minutes ago, tim716 said:

Another 18L is dead today in Russia. Just two kilometers from new. Obviously it's not "locked why trolleying" but some new issue I haven't heard of before.

The owner rode two kilometers from the store to a subway station, then turned the wheel off and trolleyed it around in the subway in a TURNED OFF state. When he tried to turn the EUC back on, it behaved like this:

https?/www.youtube.com/watch?v=mmMRzJvNOks

The firmware version is unknown, but unlikely 1.07, since he didn't have much chance to update. He didn't report any heating up. The trolley sensor was ON. 

Multiple turns on and off, reconnecting batteries, firmware update to 1.07 (he can connect to the whell and update the firmware) didn;t help.

Any comments or suggestions?

 

youtube link doesnt work on me....perhaps they can keep contacting me per FB message...maybe i can help!

 

[winter]

 

Edited August 18, 2018 by nte

[sb3en]

13 minutes ago, tim716 said:

Another 18L is dead today in Russia. Just two kilometers from new. Obviously it's not "locked why trolleying" but some new issue I haven't heard of before.

The owner rode two kilometers from the store to a subway station, then turned the wheel off and trolleyed it around in the subway in a TURNED OFF state. When he tried to turn the EUC back on, it behaved like this:

https://www.youtube.com/watch?v=mmMRzJvNOks

The firmware version is unknown, but unlikely 1.07, since he didn't have much chance to update. He didn't report any heating up. The trolley sensor was ON. 

Multiple turns on and off, reconnecting batteries, firmware update to 1.07 (he can connect to the whell and update the firmware) didn;t help.

Any comments or suggestions?

 

open up your king song app and unlock your king song, that happened to my king song. The king song app has a (lock the wheel function) it sounds to me like you locked by accident.

 

 

REPLY

[US69]

1 minute ago, nte said:

 

 tell this guy to contact me through FB ...Ulf Scheidsteger....maybe i can help.....firmware reset by android engeneer app!

but hey ...i am just. a human...maybe i get a sleep ?

[Shad0z]

1 hour ago, US69 said:

 

Just to get things right and clear up the statistics that KS, @tinawong ,me and the resellers from all over the world are having and collecting!

Where did they order, at which reseller? When please  did they get new boards from „KS“? As this is i bit irritating....as the first board where this happens not even reached KS factory? 

And btw. i am a bit irritated that your friends....going through all you other posts, just 3 of them jumped over to KS? While Tesla, Msuper X just got such a better perfomance....as you always stated?.......

i think they ordered straight from kingsong. thats what they told me 

the 3 of them had a ks16, ks18A and an msuper v3 before they go their 18L

the video i posted in the video thread of some of my other freinds one of them bought the msv3 from the guy who got an 18L while the other had an acm2 and another a tesla like me

[esaj]

53 minutes ago, US69 said:

So my main point is: Let KS analyse there thing! they ARE ON IT!... I dont like comparisons...but i personally cant remember such an open handling of such a fault in EUC world before....can you?

You do have a point in your post, but for this I can say at least Ninebot One firmware catastrophe of 2015, Ninebot P, melting Gotway connectors 2016/was it part of 2017, Gotway oscillation-firmware issue of 2017, Gotway callback of (was it) MTen in 2017 come to mind... All of these were picked apart on the forums for a long time.  This is no different in that sense.

Edited July 20, 2018 by esaj

[US69]

5 minutes ago, esaj said:

You do have a point in your post, but for this I can say at least Ninebot One firmware catastrophe of 2015, melting Gotway connectors 2016/was it part of 2017, Gotway oscillation-firmware issue of 2017 come to mind... All of these were picked a part on the forums for a long time.  This is no different in that sense.

I would say the big difference is that -for example the oscillating issue...or the melting issue of connectors....nobody from the company even admitted it, or never ever a open statement was released, or?

Edited July 20, 2018 by US69

[esaj]

14 minutes ago, US69 said:

I would say the big difference is that -for example the oscillating issue...or the melting issue of connectors....nobody from the company even admitted it, or never ever a open statement was released, or?

Yes, this is much more "minor" issue comparing to anything that can cause the wheel to cut out or otherwise fault while someone's standing on top of it riding. I have nothing but utmost respect for King Song, when it comes to technology (which I only understand a small part of in the end) and PR. I ride both KS16S and KS16B, neither of which has ever faulted on me in any way, both are great wheels. I guess I'm a border-line fanboy. I have no vendetta against King Song, I pick at the GW-fanboys from time to time, but have no real hatred or hard feelings against any manufacturer. 

But, then comes the whole "being a moderator"-issue: I personally feel that when there's an issue with any wheel, even my "beloved" KingSongs, that I have the responsibility to let people know that there's an issue and make up their own conclusions from that. I can not shrug and say, "well, it seems like a minor issue touching a minority of shipped wheels, let's just forget about it, most people shouldn't ever encounter it, and at least mostly it seems they can get away just with disconnecting and reconnecting the batteries". It's about neutrality, at least some form of objectiveness (no person is truly totally objective ever, though), and credibility. All the moderators have some favorite wheel / manufacturer, but we shouldn't let that get in the way when it comes to stuff like this.

I hope KS can reproduce (or otherwise figure out) the issue as soon as possible, fix it, and keep on building great wheels.

As for reporting new cases and taking guesses about what's causing the fault, it's not meant to hinder the company or tarnish their name. I'm personally curious of what causes this behavior, and people will always talk. The CPU/MPU-stuff I wrote above actually has (from my point of view) nothing to do with the issue, I was just curious and giving my own input on what I know about embedded systems...

Edited July 20, 2018 by esaj

[Chriull]

1 minute ago, esaj said:

Yes, this is much more "minor" issue comparing to anything that can cause the wheel to cut out or otherwise fault while someone's standing on top of it riding.

Yes and no - battery heating up and not beeing able to turn the wheel off could lead to the most serious hazard. But it did not till now, fortionately! And as it seems it does not cause high speed incidents!

I also like my ks16s, heard much about GW issues, but never had personal experiences with them. Just with ninebot one e+ before - and luckily did not get anything fried with their firmware issues in these times...

Anyhow - also beeing a mod and liking kingsong very much (without too much personal comparison experience) i would not call this issue minor, although nothing grave was reported till now and it seems to just happen very seldomly.

[tim716]

38 minutes ago, sb3en said:

open up your king song app and unlock your king song, that happened to my king song. The king song app has a (lock the wheel function) it sounds to me like you locked by accident.

 

 

REPLY

Thank you very much! Looks like your advice helped to fix his EUC! It works now. Only, it wasn't an accidental lock, but the wheel locked itself somehow. By the owner's words, he had to lock the wheel from the app, then unlock it, then turn the trolley sensor off, then back on. Only then it worked. The chance is that the uncalibrated sensor caused this. Looks like the unboxing procedure above should become a mandatory standard for now...

[sb3en]

2 minutes ago, tim716 said:

Thank you very much! Looks like your advice helped to fix his EUC! It works now. Only, it wasn't an accidental lock, but the wheel locked itself somehow. By the owner's words, he had to lock the wheel from the app, then unlock it, then turn the trolley sensor off, then back on. Only then it worked. The chance is that the uncalibrated sensor caused this. Looks like the unboxing procedure above should become a mandatory standard for now...

You welcome !, I sense your relief, what ever wheel one has, ride till you drop, its bad enough we are gawked at by some people in a negative way, we shouldnt fight among ourselves for non issues.

[esaj]

42 minutes ago, Chriull said:

Yes and no - battery heating up and not beeing able to turn the wheel off could lead to the most serious hazard. But it did not till now, fortionately! And as it seems it does not cause high speed incidents!

Right, should have mentioned this, this is why I originally posted the entire thread & pinned it. At that point, pretty much the only information was that the wheels had locked up, and actually, at that point no-one had yet tried to reconnect the batteries, so the locked wheel was assumed to be caused by fried mosfets. Things still heating up alarmed me that the circuit is still live, and pushing current through the mosfets and motor phases, causing the battery to heat up even more.

There are (minor and usually single) issues reported all the time, but my more or less "personal policy" has been that topics about them aren't made / pinned unless it seems that the issue's more widespread and/or that it can cause more significant damage (to rider or other people, or property). No one wants their house to be potentially burned down by the wheel or things blowing up in their face, or the wheel to cut out on them (the last one not being an issue in this case, at least to my knowledge). 

"Luckily", here it seems (personally drawn up conclusion from what I've heard of the issue so far) that it's "just" the firmware somehow getting stuck in an infinite loop / crashing, while the mosfets are still left pulsing and keeping motor phases powered, locking up the wheel as the magnets are being energized and causing (at least somewhat) significant current draw the battery, that gradually heats up the mosfets and the motor. On that train of thought, likely the amount of heat produced varies depending at how high duty cycle the mosfets are left at, and there's still potential that if it's high enough bad things can happen. But this is all just a personal theory. I don't know for sure, so far it's not even certain whether this is actually caused by the handle sensor or not.

It's not about "brand X has had worse issues" or such, it's a single issue on single wheel model of a single manufacturer, and I think things like these should get attention. At least the riders reading these forums do have the information of what to do, should this ever happen to them. Disconnect the battery wires as soon as possible, but even disconnecting the motor should help, since then no current can flow. Based on what we know now, disconnecting the batteries, letting things cool down and reconnecting the batteries, the wheel should function again just fine.

Edited July 20, 2018 by esaj

[ricqvo]

So, when my new 18L will arrive on monday, should I turn that sensor off after updating to 1.07 and calibrating it or not?

I hope it wont happen to me... I dont need that sensor to use at all... I just need to ride or trolley around..

[esaj]

2 minutes ago, ricqvo said:

So, when my new 18L will arrive on monday, should I turn that sensor off after updating to 1.07 and calibrating it or not?

I hope it wont happen to me... I dont need that sensor to use at all... I just need to ride or trolley around..

What nte suggested about the unboxing procedure doesn't sound like a bad idea to me:

5 hours ago, nte said:

I'm thinking this might become the 18L unboxing procedure for those that don't want to deal with this unexpectedly. According to @Shad0z it could still lock up with the sensor off. 

1. Unbox 18L
2. Remove covers
3. Power on
4. Update firmware
5. Calibrate lift sensor
6. Turn off lift sensor
7. Trolley around the house for 30+ minutes