TomOnWheels Posted July 7, 2018 Author Share Posted July 7, 2018 1 hour ago, Nikolai Pohodenko said: I suggest that everyone disable tension sensors and use the button until the problem is fixed. It seems, sensors generate too many on/off events and controller cannot handle this high frequency control input. Maybe... At last it makes sense... Especially that most issues are happening when using trolley. Also I can't get how possibly KS did not managed it in firmware. With the last update (1.0.7) the sensor really react in different way. Link to comment Share on other sites More sharing options...
tim716 Posted July 7, 2018 Share Posted July 7, 2018 3 hours ago, Nikolai Pohodenko said: I suggest that everyone disable tension sensors and use the button until the problem is fixed. It seems, sensors generate too many on/off events and controller cannot handle this high frequency control input. I also see this sensor as a very probable cause of the problem. But there's no guarantee that disabling the sensor in the app is enough. The controller might still process its signals and disable just the motor deactivation. I guess the only 100% reliable way is to physically disconnect the sensor from the controller... Link to comment Share on other sites More sharing options...
TomOnWheels Posted July 9, 2018 Author Share Posted July 9, 2018 So here is update to my initial post that started this topic. Today I've reconnected the battery, and surprise, it works ! So it looks like this was a firmware crash instead of mosfets death, as I was thinking initially. So the story could just ended here, but I was worrying about the motor and the motherboard (especially mosfets) being damaged by the heat... First of all I noticed that the battery loose around 20% after the wheel stop working. This is not so much, especially when you think about all the heat that the wheel produced before I've disconnected the battery. So I'm pretty sure that the battery is not damaged (it was not even very warm at the time of issue). Moreover @tinawong told me that below 130°C there should be no damage to the motor. But I noticed a smell of burned components on the controller, so finally took the decision to change it and send back the original controller back to kingsong for investigation. For now my wheel is working as expected with the new controller. So all in all, if it happens to you: 1) disconnect the battery as soon as possible 2) wait for the wheel to become cold 3) reconnect the battery. It should then work as expected. Also until this is cleared by kingsong, it's probably a good idea to disable tension sensor as suggested by @Nikolai Pohodenko. Link to comment Share on other sites More sharing options...
esaj Posted July 12, 2018 Share Posted July 12, 2018 On 7/9/2018 at 11:26 PM, TomOnWheels said: So here is update to my initial post that started this topic. Today I've reconnected the battery, and surprise, it works ! So it looks like this was a firmware crash instead of mosfets death, as I was thinking initially. So the story could just ended here, but I was worrying about the motor and the motherboard (especially mosfets) being damaged by the heat... First of all I noticed that the battery loose around 20% after the wheel stop working. This is not so much, especially when you think about all the heat that the wheel produced before I've disconnected the battery. So I'm pretty sure that the battery is not damaged (it was not even very warm at the time of issue). Moreover @tinawong told me that below 130°C there should be no damage to the motor. But I noticed a smell of burned components on the controller, so finally took the decision to change it and send back the original controller back to kingsong for investigation. For now my wheel is working as expected with the new controller. So all in all, if it happens to you: 1) disconnect the battery as soon as possible 2) wait for the wheel to become cold 3) reconnect the battery. It should then work as expected. Also until this is cleared by kingsong, it's probably a good idea to disable tension sensor as suggested by @Nikolai Pohodenko. While the firmware bug is a really bad thing, wow, that's some robust mosfets and motor there. If it indeed is a firmware crash (like it looks), it has left one high- and one low-side mosfet open, running full stall current through the 'fets and the motor (as much current as the battery can give shorted over the fets and the motor) for an hour for that one guy, and everything still works Link to comment Share on other sites More sharing options...
Lutalo Posted July 12, 2018 Share Posted July 12, 2018 On 7/7/2018 at 3:56 PM, Nikolai Pohodenko said: I suggest that everyone disable tension sensors and use the button until the problem is fixed. It seems, sensors generate too many on/off events and controller cannot handle this high frequency control input. Will someone like @esaj or some other knowledgeable person please post instructions on how to do this? I am expecting an 18L delivery soon (as are others). I would like to avoid any issues. Link to comment Share on other sites More sharing options...
ntz Posted July 12, 2018 Share Posted July 12, 2018 Is there any simple way to do a "factory reset" on the KS18L from v1.07 to an older version? Link to comment Share on other sites More sharing options...
TomOnWheels Posted July 12, 2018 Author Share Posted July 12, 2018 6 hours ago, esaj said: While the firmware bug is a really bad thing, wow, that's some robust mosfets and motor there. If it indeed is a firmware crash (like it looks), it has left one high- and one low-side mosfet open, running full stall current through the 'fets and the motor (as much current as the battery can give shorted over the fets and the motor) for an hour for that one guy, and everything still works Indeed, the electronics seems to be very robust ! Also I noticed that battery is redundant. If you disconnect on of them it continue working... 1 hour ago, ntz said: Is there any simple way to do a "factory reset" on the KS18L from v1.07 to an older version? Don't think so... Link to comment Share on other sites More sharing options...
esaj Posted July 12, 2018 Share Posted July 12, 2018 3 hours ago, ntz said: Is there any simple way to do a "factory reset" on the KS18L from v1.07 to an older version? This. I think the manufacturers should give the users an option the fallback to an older revision of the firmware if they see it necessary, sometimes the new versions add unnecessary limitations (for experienced riders) or if they find out a bug (regression) that wasn't on an earlier version, users should have to option to return to older version. Link to comment Share on other sites More sharing options...
Seba Posted July 17, 2018 Share Posted July 17, 2018 On 7/12/2018 at 4:36 PM, esaj said: While the firmware bug is a really bad thing, wow, that's some robust mosfets and motor there. If it indeed is a firmware crash (like it looks), it has left one high- and one low-side mosfet open, running full stall current through the 'fets and the motor (as much current as the battery can give shorted over the fets and the motor) for an hour for that one guy, and everything still works I suppose that current flowing thru motor windings and bridge MOSFETs is somehow limited to a value safe enough not to fry the controller board, even when MCU stops working. This assumption can be confirmed by the fact that the fuses are not blown even during prolonged failure conditions. So quick battery disconnection allows to avoid any damages to the motor, batteries and controller board. And this is a good thing. The bad thing is that we have to be prepared to quickly disconnect the battery in case of "trolley problem", because if windings will get too hot, it's possible to permanently damage the windings insulation and magnets. Remember - rare earth magnets used in EUCs can partially loose their magnetic properties when heated above 80°C. Unfortunately, this is irreversible Link to comment Share on other sites More sharing options...
US69 Posted July 17, 2018 Share Posted July 17, 2018 On 7/12/2018 at 10:12 PM, ntz said: Is there any simple way to do a "factory reset" on the KS18L from v1.07 to an older version? This is not possible, for several reasons. Also, even if you might think the Wheel has become "worse" with 1.07...that's not the case. Before 1.07 (1.03 to 1.06) some more boards fried because of a an LED Lightring Loop, causing the fry. KS was able to reproduce this fault and release 1.07 for this case…(And some more Things becoming better)…. The Actual fries on Firmware 1.07 are still a Kind of mystery….but very, very, seldom cases. Worldwide i know from 2-3, maximum 4 cases: Let me Quote @tinawong Statement for the actual Standing on this: For the wheel freeze on V1.07, I have contacted with the french guy, our french dealer will send back the control board to us these two days for analyze. also, our team is imitating the same situation to see whether we can find the issues, till now, we failed to get the same case. I will update to you once we have new progress. As you see, they are working hard to find out what still is going wrong…. I for my part have to admit that they are very open About this problme and dont try to hide anything. Thats not obviously or self-evident in this "Branche"!!! Link to comment Share on other sites More sharing options...
TomOnWheels Posted July 17, 2018 Author Share Posted July 17, 2018 1 hour ago, Seba said: I suppose that current flowing thru motor windings and bridge MOSFETs is somehow limited to a value safe enough not to fry the controller board, even when MCU stops working. This assumption can be confirmed by the fact that the fuses are not blown even during prolonged failure conditions. So quick battery disconnection allows to avoid any damages to the motor, batteries and controller board. And this is a good thing. The bad thing is that we have to be prepared to quickly disconnect the battery in case of "trolley problem", because if windings will get too hot, it's possible to permanently damage the windings insulation and magnets. Remember - rare earth magnets used in EUCs can partially loose their magnetic properties when heated above 80°C. Unfortunately, this is irreversible a person from KS said that it can survive 130°C (did not said for how long). Also the class of neodymiums they are using is fortunately quite resistant... Link to comment Share on other sites More sharing options...
esaj Posted July 17, 2018 Share Posted July 17, 2018 2 hours ago, Seba said: I suppose that current flowing thru motor windings and bridge MOSFETs is somehow limited to a value safe enough not to fry the controller board, even when MCU stops working. This assumption can be confirmed by the fact that the fuses are not blown even during prolonged failure conditions. So quick battery disconnection allows to avoid any damages to the motor, batteries and controller board. And this is a good thing. The bad thing is that we have to be prepared to quickly disconnect the battery in case of "trolley problem", because if windings will get too hot, it's possible to permanently damage the windings insulation and magnets. Remember - rare earth magnets used in EUCs can partially loose their magnetic properties when heated above 80°C. Unfortunately, this is irreversible Haven't checked the datasheets (I assume they're still using STM32F103?), but it could be that the PWM output is still pulsing at whatever duty cycle it was left at when the firmware crashed (ie. while the mainloop is in infinite loop / stuck, the timer interrupts are still firing). The internal resistance of batteries, cabling, connectors, fuses, mosfets and the motor phases themselves are likely very low, so you're probably right that it should have blown the fuses if the mosfets were left fully open, but other than the mosfets pulsing, I can't think what could limit the current further without frying itself in the process. Still, there's gotta be quite a lot of current flowing to heat up even the batteries themselves (someone described them as burning hot to touch? Or was it just the mosfets?)... KS16S uses 2 x 40A fuses (one per each 16S2P), don't know about the KS18L. Link to comment Share on other sites More sharing options...
TomOnWheels Posted July 17, 2018 Author Share Posted July 17, 2018 1 minute ago, esaj said: Haven't checked the datasheets (I assume they're still using STM32F103?), but it could be that the PWM output is still pulsing at whatever duty cycle it was left at when the firmware crashed. The internal resistance of batteries, cabling, connectors, fuses, mosfets and the motor phases themselves are likely very low, so you're probably right that it should have blown the fuses if the mosfets were left fully open, but other than the mosfets pulsing, I can't think what could limit the current further without frying itself in the process. Still, there's gotta be quite a lot of current flowing to heat up even the batteries themselves (someone described them as burning hot to touch? Or was it just the mosfets?)... KS16S uses 2 x 40A fuses (one per each 16S2P), don't know about the KS18L. Batteries was not even warm... I think that they get little bit warm because of the motor that was become more and more hot... Also my battery was around 95% full when that happens and when I reconnected wires it was around 65%. It tooks around 2h for me to move to the location where I finally disconnected battery cables. Link to comment Share on other sites More sharing options...
esaj Posted July 17, 2018 Share Posted July 17, 2018 2 minutes ago, TomOnWheels said: Batteries was not even warm... I think that they get little bit warm because of the motor that was become more and more hot... Also my battery was around 95% full when that happens and when I reconnected wires it was around 65%. It tooks around 2h for me to move to the location where I finally disconnected battery cables. Right, I was referring to this report: The user Asu6nik walked his KS18L by the trolley inside a shopping mall. After some 20 metres, the wheel stopped, locked in one position and resisted turning, it did not respond to button presses. After one hour, he was able to take it apart and disconnect the batteries. By that time, the whole case became very hot, the batteries inside were very hot and were making a quiet hissing sound that could be heard by pressing one's ear against the battery case, the sound resembled a boiling teapot. After letting everything cool off and reconnecting the batteries, Asu6nik had the wheel ridable again. The charge level dropped from 85% to 40% in the process. While the batteries survived, likely they did lose at least some capacity permanently after such an ordeal..? Link to comment Share on other sites More sharing options...
TomOnWheels Posted July 17, 2018 Author Share Posted July 17, 2018 2 minutes ago, esaj said: Right, I was referring to this report: The user Asu6nik walked his KS18L by the trolley inside a shopping mall. After some 20 metres, the wheel stopped, locked in one position and resisted turning, it did not respond to button presses. After one hour, he was able to take it apart and disconnect the batteries. By that time, the whole case became very hot, the batteries inside were very hot and were making a quiet hissing sound that could be heard by pressing one's ear against the battery case, the sound resembled a boiling teapot. After letting everything cool off and reconnecting the batteries, Asu6nik had the wheel ridable again. The charge level dropped from 85% to 40% in the process. While the batteries survived, likely they did lose at least some capacity permanently after such an ordeal..? Hmmm I think the noise was probably the fan... Also there is the high pitch noise of power circuit... The motor become quickly very hot (some people in Poland said it was smelling the tire burning...) and this heat up all the wheel batteries included. But not sure the heat is coming directly from the battery. Link to comment Share on other sites More sharing options...
Xoltri Posted July 17, 2018 Share Posted July 17, 2018 With the amount of current necessary to heat up the motor enough to cause the tire to smell like it was burning, surely the batteries would get hot also. Link to comment Share on other sites More sharing options...
hal2000 Posted July 17, 2018 Share Posted July 17, 2018 Just to clarify the tyre smell. It was not the tyre burning smell as if you set it on fire. It rather resembled the smell you get from brand new straight from the box tyre, meaning it was heated up enough by extremely hot engine and rim to evoke this smell/stink. There was no smoke nor melting however. I did not check the temperature of the batteries because I did not open the wheel. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.