KS-16C folded unexpectedly

[IainC]

Hi all

Long time lurker and owner of a KS-16 for the last couple of months. I've put about 200km on the wheel with no issues, but today disaster struck. I was doing the same route as I have many dozens of times, a short downhill and then a gentle uphill climb. On the latter, I was going about 7-8mph and the wheel dumped me - no warning, no beeps, no sign that anything was wrong. Luckily I was wearing wrist protectors, and a helmet, but my left knee (unprotected) is trashed (as is the jacket I was wearing on both elbows).

Battery is/was at 85%, temp at 35C, both according to wheellog (which unfortunately was not set to record a log file). The wheel is a KS-16C 850wh. When I picked myself and the wheel up, the power was off, but a quick press on the power button restored power and balancing, and all looks OK with the behaviour here.

So, ideas on what may have happened would be welcome- I was not asking for any acceleration, no significant bumps to go over, just a slow constant speed. Not sure I can trust this wheel from here on in...

[Planetpapi]

Sorry about the fall. I have to commend you for wearing the safety gear though. I don't own a KingSong so I can't really offer any insight. But the point is proved more than few times that..."No wheel is safe". I ride my Ninebot but I NEVER trusted it. One fall and you lose your confidence and from then on wards  it's always on back of your mind and your riding style changes slightly, at least for me. KingSongs are reputable but shit can happen to any wheel. Let's see if the experts can figure it out here.

[JumpMaster]

I've had this happen when I accelerate too hard on a hill. Usually from a standing start or from very low speeds. It seems to be to protect the 40A fuse. I'm very careful when accelerating up hill!

[US69]

31 minutes ago, IainC said:

Hi all

Long time lurker and owner of a KS-16 for the last couple of months. I've put about 200km on the wheel with no issues, but today disaster struck. I was doing the same route as I have many dozens of times, a short downhill and then a gentle uphill climb. On the latter, I was going about 7-8mph and the wheel dumped me - no warning, no beeps, no sign that anything was wrong. Luckily I was wearing wrist protectors, and a helmet, but my left knee (unprotected) is trashed (as is the jacket I was wearing on both elbows).

Battery is/was at 85%, temp at 35C, both according to wheellog (which unfortunately was not set to record a log file). The wheel is a KS-16C 850wh. When I picked myself and the wheel up, the power was off, but a quick press on the power button restored power and balancing, and all looks OK with the behaviour here.

So, ideas on what may have happened would be welcome- I was not asking for any acceleration, no significant bumps to go over, just a slow constant speed. Not sure I can trust this wheel from here on in...

i would go and check the Motherboard for any defects...by myself or through distibutor!

does not sound like an cutoff because of to mich power needed...perhaps something wrong with the board? capacitor broken?

Did your wheel before show any signs of sluggish reaction? not promptly doing accelerate or brake?

 

[IainC]

Thanks for the replies guys. Really nothing to suggest I was 'overdemanding' from the wheel- just trundling along (fortunately slowly) without asking for any sudden acceleration. One of the reasons I went for the Kingsong was because they're supposed to be pretty good at not crapping out unexpectedly, but it seems not the case with mine. The fact that there was no warning, and the wheel switched back on with no fuss (and seems absolutely fine, no instability or sluggishness) suggests that there's no major fault, but I may take it apart and check the capacitors and connections. I had hoped with this latest revision of the wheel things like dodgy capacitor legs would have been fixed anyway :-)

A lesson for everyone though- don't skip the protective gear, even for a short ride. I think I was lucky in some ways- my wrist guards took the brunt of the fall, and I was lucky that the wheel rolling into the road didn't coincide with any cars coming! Knee and elbow guards for me from here on in!

My next few rides when I'm healed will be very careful and slow- but one other point to note, even if you think you can 'run off' a wheel at low speeds, the suddenness of a cut-off really doesn't make that a realistic escape manoevre.

Happy riding to everyone on this forum, and a quick thanks to all for everything I have learnt before even my first post

[Frode]

Just out of curiosity: Have you lately been going down a relatively steep hill with full battery and relatively high speed, or have you been braking relatively hard? (relatively steep means here like 8%+, relatively fast means ca. 10 mph/15 kph, relatively hard breaking means like emergency breaking)

[IainC]

Nope, my ride started with battery at 85%, and the gentle incline I went down first was just that, so no reason to think that regenerative charging was the culprit here. May just be one of these mysteries, but I'll be very cautious for my next few rides!

[Hunka Hunka Burning Love]

Could you tell us your weight and was there any head wind?  I remember reading about someone else going up a hill, and he experienced the same thing... I can't recall whether it was a KS16 or not.  He was about 258 pounds I think in weight.  I wonder if some cells might not be putting out enough current from the pack or the BMS might be faulty?  Check the capacitor legs as Kingsong69 mentioned.

[Cloud]

Check if motor connections are in any way loose, or any other connectors loose....

check if there is any burning smell inside the wheel

it wasnt raining by any chance, was it?

did this happen when you just started going uphill, or after a little while?

[OliverH]

The connector is very good and tight on the KS16. This should be checked but I can't imagine that there's an issue. As I just have a KS16 from a guy of our riding group on my work bench I saw two issues:

- harness comes out of the axle and makes a hard turn. A coil spring should protect the cable here. I saw the same stuff on FireWheels where the cable got damaged by this spring coils

- the motor harness is routed between the battery and the case on one location, the cable of the three phases can become very hot 

What Firmware is installed? 

Was audible connection switched on? Any note occurred?

Was front/ head light on?

[Hunka Hunka Burning Love]

I found the other thread I was talking about.  It was @Frode 's KS16 that face-planted him while going up a hill.

 

[IainC]

Wow, lots of great replies, thanks all. My weight is 85kg, no rain, firmware 1.22, light was off, sounds (speech) on - but nothing made any noise, no significant wind. The 'hill' is really just a gentle incline, no more than 3-4% gradient. I'm pretty familiar with the smell of burning components (years tinkering with electronics) and there's no smell of burning, but I will take the wheel apart and have a look. Back at home, I switched the wheel on and jiggled it around fairly vigourously, no cut-offs, but you never know whether there's a loose wire or similar.

Will keep everyone posted on investigations, I'm kind of hoping I do find a fault (that I can fix) as otherwise I'm going to be very nervous using this wheel in the future...

[Hunka Hunka Burning Love]

If it's still under warranty, you might want to contact your dealer to see what they think of this happening.  It could be a component going out of spec, but it's not giving any overt signs of failure (eg. burnt casing, eroded capacitor leg, etc).  They might suggest swapping out the control board just to be on the safe side.  Has the wheel functioned normally since (eg. same distance range, hill climbing ability)?

[Greg Spalding]

I'll bet Iain has yet to put it to the test yet at all... I think everyone of us would be leery about writing. The pleasure factor of being on one of these electronic unicycles would certainly go down dramatically if one had to worry about the power being suddenly cut at any possible moment

[IainC]

Quick update from me- took the wheel apart this evening, no signs of anything wrong at all. All connectors tight, no sign of any issues with any components on the board (by the way, the big capacitor is now glued to the fan and effectively immobilised), no pinched wiring or any signs of bad cable routing. Cable from the hub is now spring protected as it goes round the sharp angle.

Shame in some ways as I was hoping to find something obvious. A very ginger ride round the house revealed absolutely normal behaviour, no odd noises. I'm going to speak to my dealer tomorrow to see if they have any ideas.

Thanks again for all the input here guys

[Greg Spalding]

51 minutes ago, IainC said:

Quick update from me- took the wheel apart this evening, no signs of anything wrong at all. All connectors tight, no sign of any issues with any components on the board (by the way, the big capacitor is now glued to the fan and effectively immobilised), no pinched wiring or any signs of bad cable routing. Cable from the hub is now spring protected as it goes round the sharp angle.

Shame in some ways as I was hoping to find something obvious. A very ginger ride round the house revealed absolutely normal behaviour, no odd noises. I'm going to speak to my dealer tomorrow to see if they have any ideas.

Thanks again for all the input here guys

in a perfect world they would take BACK that wheel and give you a new one, promising to fully report back about what was wrong because the manufacturer wanted to know.

that would NEVER happen in the world of EUCs, sadly... but, i we can dream

[HermanTheGerman]

Hi All,

please don't burn me if I say something which you might not want to hear, but just a thought.

When I was first pointed to WheelLog, I installed the app (since 1.6) and used it first a few weeks after I got my new KS16. At the first usage my KS16 folded suddenly once when I wanted to start a ride.  I thought it could be the new app, possibly transferring some frames which the KS16 is not happy about, deinstalled it, and then everything went well.

Then jumpmaster came out with 1.7, I gave it a new try and since then everything worked fine, I never had a problem again (rode since then approx. 500 km).

So my question is, because the communication with the wheels is of course reverse-engineered via traces and the like:

Could it be, that some frame or command is sent via bluetooth to the wheel, which unintentionally does something bad ?

Don't forget that via the normal KS app you can set several things (lights, calibration and so on), maybe even more is possible, which is not documented (like turning the wheel off).

I don't know, what does jumpmaster say, could that be possible, that a bad communication between wheelLog and the KS16 could lead to such problems ?

Thanks, HtG

[HEC]

3 hours ago, HermanTheGerman said:

Could it be, that some frame or command is sent via bluetooth to the wheel, which unintentionally does something bad ?

In theory it's not impossible though @JumpMaster is using only two commands in the direction towards the wheel - horn and requesting the wheel details (serial number etc.) otherwise he's just subscribing to standard "broadcast" services which are only for receiving data.

[JumpMaster]

2 hours ago, HEC said:

In theory it's not impossible though @JumpMaster is using only two commands in the direction towards the wheel - horn and requesting the wheel details (serial number etc.) otherwise he's just subscribing to standard "broadcast" services which are only for receiving data.

That's exactly right.  There are three commands though.  Two to request data which is why the wheel beeps twice when the app connects.  Plus the horn.

3 hours ago, HermanTheGerman said:

So my question is, because the communication with the wheels is of course reverse-engineered via traces and the like:

Could it be, that some frame or command is sent via bluetooth to the wheel, which unintentionally does something bad ?

The code I've used to from a decompiled version of the KingSong app.  So when a command is sent to the wheel it is EXACTLY the same as the KingSong app.  Their app is based on the Google BluetoothLE example (it still includes the heart rate monitoring code) as is mine.  My code is, in my opinion, an improvement and less likely to cause issues then their app.  At the heart of the apps neither does anything particularity clever.  Both tell the Android OS to send commands to the wheel.  I think it is more likely that there are bugs in different phones or versions of Android that could cause an issue then there is in either app.

If there's a bug causing this, it's most likely in the Wheel. Whether its triggered by the Bluetooth commands is very unlikely but as both WheelLog and KingSong use the same standard BluetoothLE commands it won't be caused by either app.

[Greg Spalding]

6 minutes ago, JumpMaster said:

That's exactly right.  There are three commands though.  Two to request data which is why the wheel beeps twice when the app connects.  Plus the horn.

The code I've used to from a decompiled version of the KingSong app.  So when a command is sent to the wheel it is EXACTLY the same as the KingSong app.  Their app is based on the Google BluetoothLE example (it still includes the heart rate monitoring code) as is mine.  My code is, in my opinion, an improvement and less likely to cause issues then their app.  At the heart of the apps neither does anything particularity clever.  Both tell the Android OS to send commands to the wheel.  I think it is more likely that there are bugs in different phones or versions of Android that could cause an issue then there is in either app.

If there's a bug causing this, it's most likely in the Wheel. Whether its triggered by the Bluetooth commands is very unlikely but as both WheelLog and KingSong use the same standard BluetoothLE commands it won't be caused by either app.

Always nice to have you posting here and teaching us

thanks for the clarification and education

[HermanTheGerman]

On 6.10.2016 at 6:57 PM, JumpMaster said:

That's exactly right.  There are three commands though.  Two to request data which is why the wheel beeps twice when the app connects.  Plus the horn.

The code I've used to from a decompiled version of the KingSong app.  So when a command is sent to the wheel it is EXACTLY the same as the KingSong app.  Their app is based on the Google BluetoothLE example (it still includes the heart rate monitoring code) as is mine.  My code is, in my opinion, an improvement and less likely to cause issues then their app.  At the heart of the apps neither does anything particularity clever.  Both tell the Android OS to send commands to the wheel.  I think it is more likely that there are bugs in different phones or versions of Android that could cause an issue then there is in either app.

If there's a bug causing this, it's most likely in the Wheel. Whether its triggered by the Bluetooth commands is very unlikely but as both WheelLog and KingSong use the same standard BluetoothLE commands it won't be caused by either app.

Thanks for the info !

It's good to know some details.

Cheers, HtG