King Song 14C Balance Issue

[FreeRide]

Winter arrived before I could get more than a few hours on my 14C, but now I want to get back in the Saddle and try to learn this thing.  I turn it on today after it has been sitting for a few months and it will not balance.  Fully charged, it turns one anthem slowly tilts one direction then starts beeping goes further and then reverses direction all slowly like it is trying to find equilibrium.  the wheel has have no big bumps or crashes and is essentially new.  I can connect to it with the iOS App, and control the light, but it will not self balance.  I thought I may have entered a test mode or something, but it seem more like a failure.  I'm really disappointed, I sent a message to Jason and am waiting to see if he has any ideas, but I wanted to ride today as I have time and this is a real bummer, so I thought I would check if anyone here has any ideas on a hard reset for this wheel or anything?

Thanks, waiting to join the EUC club, but it's not been easy.

 

Thanks!

[Marty Backe]

I would try and calibrate it. That's probably not the problem but it's easy enough to try.

[FreeRide]

Thank's Marty.  I tried calibration, and it seems to calibrate fine, all the right beeps.  When I turn it on though same as before starts up and then slowly tilts forward then backward trying to do something.  Wheel had power and will spin up if lifted.

As it is trying to balance it emits sequences of three quick beeps.

Ironically, the last thing I did was put bumpers on this thing.

[Marty Backe]

Looks like you've tried everything. Hopefully these failure symptoms have been seen before and Jason or someone else will know the solution.

It's hard to believe the electronics would fail while the wheel has been laying dormant. Maybe it's a bad cable-to-control board connection. With temperature and humidity changes mechanical parts can move over time, and if a connection was marginal to begin with, maybe an 'open' can develop. Just throwing some ideas out there.

Your predicament is why I started collecting wheels. I hated the thought of being 'grounded' with no wheel to ride if I were to have a failure.

Best of luck.

[FreeRide]

Yup, that is what makes it frustrating for me since it was just sitting indoors, what the heck happened.  I keep thinking because of this it must be something easy... but... I'm already thinking of ordering another one as these issues could take some time to resolve, and Spring and Summer only last so long.  

[DaveThomasPilot]

I had/have exactly the same symptoms.  After sitting for months, the wheel just tilts back and forth.

In my case, the reason was that the wheel thinks the battery voltage is too high.  I had to leave it powered on for days until the voltage decayed low enough that the wheel didn't think it was too high.

Check out my threads, in the first one I start talking about the problem about 1/2 way through...

 

 

In my case, the wheel thinks the battery voltage is 2 - 3 volts higher than it really is.  So, it does the tilt forward/back thing after a full charge. 

You can buy a Charge Doctor to partially charge the wheel so you avoid the overvoltage.  But, then you have to decide how long/far/to what voltage you can ride safely.  That's what the second post about charge/discharge curves is all about.

Pm for a phone number if you want to understand better.

 

[DaveThomasPilot]

29 minutes ago, Marty Backe said:

It's hard to believe the electronics would fail while the wheel has been laying dormant. Maybe it's a bad cable-to-control board connection. With temperature and humidity changes mechanical parts can move over time, and if a connection was marginal to begin with, maybe an 'open' can develop. Just throwing some ideas out there.

 

I've seen three instances of these failure symptoms on this forum (besides mine).   I'd love to get a schematic so I could figure out how they are measuring voltage.  Surely, they are not using a resistor divider to scale the voltages with high resistance values and failing to account for input bias current...  The parasitic current would drift over time and cause this type of issue.

But, that would be pretty naive design.

[Marty Backe]

8 minutes ago, DaveThomasPilot said:

I had/have exactly the same symptoms.  After sitting for months, the wheel just tilts back and forth.

In my case, the reason was that the wheel thinks the battery voltage is too high.  I had to leave it powered on for days until the voltage decayed low enough that the wheel didn't think it was too high.

Check out my threads, in the first one I start talking about the problem about 1/2 way through...

 

 

In my case, the wheel thinks the battery voltage is 2 - 3 volts higher than it really is.  So, it does the tilt forward/back thing after a full charge. 

You can buy a Charge Doctor to partially charge the wheel so you avoid the overvoltage.  But, then you have to decide how long/far/to what voltage you can ride safely.  That's what the second post about charge/discharge curves is all about.

Pm for a phone number if you want to understand better.

 

Forum to the rescue. Be sure to let us know if @DaveThomasPilot is correct with regards to your wheel. It may be helpful for people in the future.

[FreeRide]

Thanks so much!  Might make sense in my case, because of course after sitting for months I put it on the charger so it would be ready.   It was only on the charger for an hour before the green light showed.  I will try to drain the battery a bit, and report back if it works.

[Marty Backe]

4 minutes ago, FreeRide said:

Thanks so much!  Might make sense in my case, because of course after sitting for months I put it on the charger so it would be ready.   It was only on the charger for an hour before the green light showed.  I will try to drain the battery a bit, and report back if it works.

You can put some tape over the light sensor to force the light on for an hour or so.

[DaveThomasPilot]

It took my three days to discharge enough to start riding, with the lights on the whole time.

[FreeRide]

Many Thanks @DaveThomasPilot!!!  

You were right i put it under load, and after a short time it stopped beeping while loaded, and then even when resting.  Looks like parts tolerance, their measurement A-to-D must have some low quality analog parts or something.  I would not expect drift in the resisters, but in a capacitor totally possible.  Or that aftermarket charger put the cells too high, or the device is setup so the cells should be undercharged a bit to prolong life.  If I believe the voltage from the KingSong iOS App (and I don't) 67.7 is too high and 67.5 is probably OK.  

I'm so happy!!!  my 3.5' test track got a good workout. I presume the extra dirt I got on the device did not cure the issue so to me you have nailed it.  That symptom means that the device thinks the battery voltage is too high!

[FreeRide]

1 hour ago, Marty Backe said:

You can put some tape over the light sensor to force the light on for an hour or so.

I did something a bit more dangerous, but a lot more effective (quicker).  

[Marty Backe]

14 minutes ago, FreeRide said:

Many Thanks @DaveThomasPilot!!!  

You were right i put it under load, and after a short time it stopped beeping while loaded, and then even when resting.  Looks like parts tolerance, their measurement A-to-D must have some low quality analog parts or something.  I would not expect drift in the resisters, but in a capacitor totally possible.  Or that aftermarket charger put the cells too high, or the device is setup so the cells should be undercharged a bit to prolong life.  If I believe the voltage from the KingSong iOS App (and I don't) 64.7 is too high and 64.5 is probably OK.  

I'm so happy!!!  my 3.5' test track got a good workout. I presume the extra dirt I got on the device did not cure the issue so to me you have nailed it.  That symptom means that the device thinks the battery voltage is too high!

Yeah! Now get out and have fun with that wheel. I really enjoy mine and ride it almost every day.

[FreeRide]

I sure hope I can learn to ride it/  The Segway MiniPro was too easy, but this is a bit of a challenge.  Still I made some progress today even when it was broken.  I think I need to check the air pressure again too.  I hope I'll be able to ride mine everyday and get at least one more wheel.  I'm very happy now. 

[Jason McNeil]

Getting spares 14C boards has been a nightmare: I ordered 4x of these a couple months ago, sent one of these out to @DaveThomasPilot to replace his over-voltage unit & the heatsink was slightly bent. He was able to knock it back into shape but after installing it, the blasted thing didn't work. Board #2 reported the voltage correction, but the battery indicator was defective. I suspect KS stopped making these some time ago, & some lazy ***** in the Parts department took these from the reject pile & sent them out. 

It's curious that both @FreeRide & Dave's problems occurred after being kept in storage for several months—they were fine initially. Is it possible the internal Voltmeter drifts if left unused storage? Might just be a coincidence.   

I feel bad about this, Dave's been wonderfully patient (issue has been going on for over a month now) & we're still awaiting KS to send some working boards. To go on a bit of a tangent, KS has some serious problems with shoddiness in with spare parts handling, on one occasion I ordered 5x 840Wh battery packs for the KS18 (these are not cheap), & the imbeciles throw the lot without any padding or protection into a soft cardboard box with a bunch metal handles—how fortunate it didn't start a fire on the ship!! Only three of five packs were in box, KS refused to do anything about it... Grrh

[Marty Backe]

4 minutes ago, Jason McNeil said:

Getting spares 14C boards has been a nightmare: I ordered 4x of these a couple months ago, sent one of these out to @DaveThomasPilot to replace his over-voltage unit & the heatsink was slightly bent. He was able to knock it back into shape but after installing it, the blasted thing didn't work. Board #2 reported the voltage correction, but the battery indicator was defective. I suspect KS stopped making these some time ago, & some lazy ***** in the Parts department took these from the reject pile & sent them out. 

I feel bad about this, Dave's been wonderfully patient (issue has been going on for over a month now) & we're still awaiting KS to send some working boards. To go on a bit of a tangent, KS has some serious problems with shoddiness in with spare parts handling, on one occasion I ordered 5x 840Wh battery packs for the KS18 (these are not cheap), & the imbeciles throw the lot without any padding or protection into a soft cardboard box with a bunch metal handles—how fortunate it didn't start a fire on the ship!! Only three of five packs were in box, KS refused to do anything about it... Grrh

Folks who order direct from these Chinese companies should know the risk that they are taking. If this is an example of how they treat their dealers you can imagine what non-service you would get as an individual.

I sure hope my KS14C (from you Jason) continues to work because I know that if it fails it'll be going in the trash. I hope your Gotway experience ends up being better.

[FreeRide]

So it's been confirmed @Jason McNeilthat it is the device that is in error, and not the charger?  Seems more likely there would be drift in the charger.  Talk about frustrating.  This is a great new product, we need some quality manufacturers.  It's too bad that Ninebot took a left turn they were on the right path.  I'm probably going to adjust the power supply in the mean time.  I have no intention of opening up that beast of a wheel and trying to replace parts.  This seems like a warranty issue, but I'm also not interested in shipping back that heavy beast.  Live and learn.  I may need to open the wheel so I can measure the battery voltage, but if I have to adjust the supply I can also just go by how much distance I get, I got it with the large pack which should be more than I need.  All this presumes I'll be able to ride it further than my driveway.  Eventually someone will probably figure out how to fix the board if that is the core of the issue.

[DaveThomasPilot]

Jason first sent me a replacement charger when the issue first happened last fall.  That was after the wheel had been idle after the initial purchase for about 6 months (I broke my back in an aircraft accident and couldn't ride the wheel for several months). 

First he sent a slow charger.  That didn't overcharge and we thought the problem was solved.  I rode on that charge, but my feet hurt so bad if motivated me (along with other things) to loose weight.  I've lost over 30 pounds so far and wanted to start riding again last month.  That's when I noticed the tilting again. I compared the voltage reported by the app to the output voltage of the charger and noticed the inconsistency.

Then, I bought the charger doctor and found the voltage delta to be still worse (over 3 volts!).

So, something is drifting.  No component drifts that fast unless it's over-stressed in the design (3X rated power dissipation in a precision resistor, for example).  But, I can see the design using very high resistor values to scale the relatively high battery voltage and failing to recognize the impact of stray, leakage, and input bias currents on a high impedance nets voltage.

 

 

[FreeRide]

Thanks Dave.  So the standard charger initially solved the problem, but then it returned?  I read that Jason had sent you a standard slow charger, but never found the results so thanks for the update. I'd rather use my DMM than the Charge Dr. for voltage measurements especially ones I want to measure in tight tolerances although the Dr. would be fine if I confirmed it's measurements first, which I think I read you did.  Still to be accurate I need to have the voltage right at the battery and I'll have to open the unit for that.   Seems like since these things were powered off there should be little reason for component drift unless it is something like a very leaky capacitor, nothing else you mentioned would have such drift if it was powered off.   There could be external factors, such as temperature and input voltage to the charger.   Both of these seem not likely as more problems would likely be showing up with more people.   Thanks again for you're help, you really saved the day! 

[DaveThomasPilot]

Yes, I slow charged once and rode immediately after the charge and had no issue. 

Best I can recall, I rode on that charge for miles, but still got tilt-back going down hills.  But, my feet really hurt a lot and it just wasn't fun riding more than a few minutes.

I fully charged it using the fast charger Jason sent me.  I don't recall if I tried riding immediately after that charge, but I did check to make sure it wasn't doing the tilt back thing immediately after being fully charged with the fast charger.

Then, it sat for several months.   When I powered it on again, I got beeping and tilt-backs and had to leave it on for days before I could get on it.

Before the Charge Doctor arrived (I ordered it from France before I knew Jason distributes them in the US), I added wires inside the fast charger to measure its output voltage.  I expected the charger voltage to be higher than what the wheel reported, since it's on the anode side of a protection diode, but it was lower.  Which said something was clearly wrong.

When I got the ChargeDoctor, I did open up the wheel so it measured the battery voltage.  I did that for the discharge curve, but didn't bother for the charge curve (I would have needed to build another connector adapter).

The reason to use the Charge Doctor is to have it automatically cut-off the charge at a specified voltage.  You have to figure out what voltage (or current) will do that for your specific wheel (and hope it doesn't drift a lot).  Absolute accuracy isn't really important--you just need to determine the cut-off voltage needed to prevent tiltiback.

Seems like since these things were powered off there should be little reason for component drift unless it is something like a very leaky capacitor

Aluminum electrolytic caps are leaky by nature.  They also dry out over time.  But, I don't know if they could get significantly more leaky sitting in a relatively hot garage.  The analog circuity designer should not have a circuit where the leakage current of a cap is important to the voltage measurement accuracy, but it seems like it must be something like that going on.

u mentioned would have such drift if it was powered off.  

Both of these seem not likely as more problems would likely be showing up with more people.

I think I've seen four reports on this forum with the same symptoms.  I'll try to go back through and find them.

[DaveThomasPilot]

Both of these seem not likely as more problems would likely be showing up with more people.

Also, if many people are using the ChargeDoctor and terminating the charge early to preserve battery life, they wouldn't know if this issue existed in their wheel or not.  But, the low voltage tilt-back warning might not occur in time to prevent an accident.  At least, that is my worry.

[FreeRide]

Yes, i want to know the exact voltage on the battery to get a definitive answer if it is battery charge or measurement.  The original charger may have ben designed to undercharge for a reason.  Also if i have to undercharge it to have the wheel not complain, i want to know buy how much i'm undercharging it.  The last bit of charge is not important, but if the measurement is on voltage and it is off by a couple volts that would be significant in some cases.  Unfortunately a discharge is really required to get an accurate state of charge, but of course I'm sure no wheel manufacturer discharges their batteries to the recommended LVC of the particular cells in the pack since that would be too big of a safety issue.  So the real capacity that can be discharged from the pack under use is probably going to be quite a bit below advertised in any case.  The wheels require a known power level rather than voltage level so their important discharge curve is power, not the typical V or mAh.  At least with Lithium you can get a indication of charge state based on OCV within a certain range and not adjusting for the details of the specific cells.

Occam's Razor tells me it is most likely the charger, but I'd really like to know for future purchases. 

[esaj]

On 5/9/2017 at 10:47 PM, FreeRide said:

The last bit of charge is not important, but if the measurement is on voltage and it is off by a couple volts that would be significant in some cases.

Actually, my understanding is that on most wheels, the BMS balances the cells only by bypassing them once they reach full voltage, so if not charged all the way to the "brim" every once in a while, the cell voltages might start to drift, and eventually one or more of them could actually discharge too far. This is one of the example circuits I found:

No idea if something similar is used in wheel BMSs. Maybe the higher quality wheels use active balancing.

It would also seem that at least most wheels detect the remaining battery charge simply based on battery voltage, although maybe newer ones do something more precise. On older wheels, the battery gauges on the wheels themselves would go up and down with braking and accelerating, once the battery was somewhat depleted. On my Firewheel, which had a numerical 0-99% (two digits only) -battery gauge, it would be pretty hilarious, 0% going up hill, 60% going down and jumping in between those values while riding straight, depending on acceleration/braking  

 

On 5/9/2017 at 2:44 PM, FreeRide said:

 Seems like since these things were powered off there should be little reason for component drift unless it is something like a very leaky capacitor, nothing else you mentioned would have such drift if it was powered off.  

I don't know how it really works, but I think there's at least some part of the (mainboard) circuitry powered at all times (when the battery's connected, maybe if it uses really small current it could run off of just the capacitors even without battery for a while). The reason I've come to that conclusion is that the power switch couldn't handle it if the currents the wheels use would run right through it, instead it seems to give a signal to the mainboard to start up the wheel, and (at least on most wheels) doesn't power down if the wheel is moving (or has some alternate functions, like turning lights on/off with different length of holding the button down etc). Or then there's some separate circuitry that only powers up the actual CPU etc. on the board on first press, and after that, the CPU would take control of the power down...  I don't know, I'm just guessing and "thinking out loud" really   But, if the CPU was indeed powered and "doing things" even when the wheel seems otherwise powered off, I could imagine something like an overflow could occur over a long period with high voltage readings (if it would indeed do the measurements even when otherwise off)... couple that with something like "calibration/offset" values calculated based on some long-term average or whatever, and stored in EEPROM/Flash and you could get some very "interesting" bugs like this one.

A leaky capacitor is probably a lot more realistic, though   I don't know if they use a resistor divider, but I'd guess high-voltage ADC's would be pretty expensive (if such even exist?) and to my knowledge you can't stick 67+V straight into the CPU ADC-pins...     If anyone ever bothers to trace down the board and draw a schematic, I'd be interested (although not sure how much of it I would really understand, if there's lots of some analog-magic ). I've been planning to do that with a couple of (dead) Firewheel boards I have, but haven't gotten around to it so far (it does't help that they're very likely multilayered).

[FreeRide]

Yes, like you say there is certainly always some parasitic drain because the powerswitch is an electronic switch, but typically that is very low and usually only in a very small part of the circuit, well away from I would suspect measurement stuff, but we'd need to know more about where the problem lies.  I'm still betting it is the charger, because even though I left it sitting, before using it I plugged it into the charger.  I probably should have just used it as the drain was probably negligible, so at least there was an opportunity for it to get over charged.  We'll see if we start hearing about more instances of this problem, we might be able to narrow it down.  I need to get my training in before I rip it open.