KS16S stops charging at 76%

[fritzlakritz]

Annyone an idea why my new KS16S stops charging at 76%

The KS is new and i made two charges since now but think this ist not a normal behaviour.

I checked the charger - the output voltage ist 67,2 Volt.

[Keith]

17 minutes ago, fritzlakritz said:

Annyone an idea why my new KS16S stops charging at 76%

I checked the charger - the output voltage ist 67,2 Volt.

The only thing that can be measured on a battery easily is voltage and 67.2 volts is 100%. So you are going to have to explain what you mean a bit more clearly, because , if the charger is reaching 67.2 volts, then it is charging to 100%. Do you mean the LED's on the wheel or the reading in the app is showing 76% in which case the problem is every bit as likely to be an incorrect reading from them as it is the charger not doing its job.

Basic charger functionality is:

• restrict maximum current to a fixed stated value - 2 Amps, 4 Amps or whatever it says on the charger.
• voltage slowly rises
• voltage hits 67.2 volts 
• restrict voltage to no more than 67.2 volts
• current slowly drops 
• When current has reduced to a trickle (typically 150-200mA) turn the red charging LED to green

The charger never stops charging, the current simply gets lower and lower as the battery gets closer to full voltage. The point where the LED goes green is simply an arbitrary point where the battery is considered to be pretty much full. It will go green if any sort of fault condition reduces the current to below its set point of 200mA or so.

[fritzlakritz]

Sorry Keith, you´re right - my description is a litle bit imprecisely.

So here some additionaly informations:

1) Charging is finished, the charger shows the green LED
2) Connected via App it shows 77% charged and 46km milage
3) 7 to 8 LED on the Wheel shows the green light
4) checking Battery Voltage in the App. It is 62,4 Volt
5) checking the Charger it has 67,2 Volt output voltage

I think there could be 2 reasons for that:

1) The charger has a malfunction under load
2) The App has a malfunction and shows wrong informations

Any Idea how to check the Voltage of the Wheel without App?

 

Edited July 6, 2017 by fritzlakritz

[US69]

I have another idea, which i heard can be a reason for such a behaviour:

The batteries ARE fully loaded...but the "voltage reader" on the main board might have a defect!

as your charger seams to put out correct voltage, this could be the reason!

 

there is only one way to check: measure the batteries themself after a full charge!

As this may ruin your warranty, if opening the wheel,.i would contact the seller.....

Edited July 6, 2017 by KingSong69

[Keith]

2 hours ago, fritzlakritz said:

I think there could be 2 reasons for that:

1) The charger has a malfunction under load - possible but fairly unlikely
2) The App has a malfunction and shows wrong informations - very possible 

3) The battery has some faulty cells and cannot hold 67V - ought to result in the charger never going green.

62.4 volts would indeed be around 76% charge. Is the wheel charging up to that level OK? I.e. Is the charge light going red at all? Has the wheel been discharged to a lower voltage and then charged back up to that level?

If the answer to all of the above is "yes" that would suggest the charger is working OK.

If the wheel has never been below 76% since this problem occurred and the charger is not going red at all then the problem is either with the charger or the connection between charger and battery.

As @KingSong69 has said, you need to check the actual voltage on the battery to ascertain if this is just false info from the app. However these batteries are capable of discharging arc welder level currents so great care is needed as any short will likely have blobs of molten metal flying about. Contacting the seller would be the safer option.

Edited July 6, 2017 by Keith

[Chriull]

According to the previous posters either the charger goes down in voltage under load ( strange and unlikely since the load is reduced the higher the voltage/charge gets - but still a possibility) or some cells are degraded. So, if possible a return to the reseller!

if not, measure the different cell voltages (you need to open the plastic shrink wrap of the battery pack ) and be quite cautious doing this - every short can easily lead to desaster!

edit: i am not sure about @Keith s statement "3) The battery has some faulty cells and cannot hold 67V - ought to result in the charger never going green.)

the charger acts in this range as current source and if the degraded cells have a higher internal resistance this could lead to the charger showing the green led - but i am not as much into knowing the behavior of degraded cells to give any final statement regarding this...

Edited July 6, 2017 by Chriull

[Chriull]

Or just as @zlymexexperienced in 

the bms should kick in the overvoltage protection and stop the charge.

[Christoph Zens]

14 hours ago, Chriull said:

the charger acts in this range as current source and if the degraded cells have a higher internal resistance this could lead to the charger showing the green led - but i am not as much into knowing the behavior of degraded cells to give any final statement regarding this...

According to my experience, bad cells usually go bad in a way that they have increased internal resistance. I don't know the exact current where the light turns green, but I would guess something around 50 to 100mA. Since the difference between the charger voltage and the battery voltage seems to be 4.8V, and assuming that the LED is green at this point and current is down to 100mA, this would mean a total series resistance of the battery pack of 48 Ohms. With the battery in this state, he could not ride very far before falling off (undervoltage shutdown), so I don't think that this is his problem. Also, this wheel has 16sx4p cells, so one or two bad cells could not increase series resistance like this. There would have to be at least one bad cell in every 16s line, and every line would have to have 192 Ohms resistance on average. Unlikely. I would say the voltage measurement on the board is broken (cold solder joint). I would not dare to ride in this state, better return it and have it fixed. The main board on the 16S seems to communicate back to the BMS, so I would think that in case the problem gets worse, and measured voltage drops dramatically, the board could suddenly shut down the BMS.

[fritzlakritz]

Thanks all to the comments. Actually i am in touch with the seller. Will give Feedback soon.

[DaveThomasPilot]

Several have reported (including me) problems with KingSong control boards having a inaccurate reading of the battery voltage. 

I started a couple of threads--there's some detail towards the end (page three) of this one:

 

Someone (esaj?) posted a schematic of the voltage scaler used on a "generic" design.  The large value of resistors used in that design would certainly contribute to voltage measurement inaccuracy due to parasitic leakage currents and the input current of the chip that does the analog to digital conversion of the scaled battery voltage.   I have no clue if KingSong used a "less naive" design for the voltage scaling.

But, I'm thinking, the root cause is a poor circuit design for the voltage scaling.   The on-chip ADC conversion accuracy is much better than the errors we'd notice in EUC designs.

But, depending on the magnitude of the the chip's input current (it will vary widely chip to chip), the voltage can read too high, too low, or correctly if large value resistors are used to do the necessary voltage scaling.

My replacement board is reading low, but that's better than reading too high. 

If the control board thinks the battery voltage is higher than it actually is, you can't ride the wheel after a full charge, even on level ground (that was my problem on the first board).

If the control board thinks the battery voltage is lower than it actually is, the predicted range will be lower, and perhaps safety things that kick in at lower voltages are premature.  But, I don't ride anywhere near long enough for that to be an issue (I'm the weakest link, not the wheel!)

Edited July 8, 2017 by DaveThomasPilot
scalar changed to scaler

[dmethvin]

I think I have the high-reading problem on my 14C as well, but since I never charge fully (I use a Charge Doctor) it isn't generally a problem. The only time it appears is when I ride down this one hill close to my house after a full charge, I get the overcharge tilt-back.

[DaveThomasPilot]

21 hours ago, dmethvin said:

I think I have the high-reading problem on my 14C as well, but since I never charge fully (I use a Charge Doctor) it isn't generally a problem. The only time it appears is when I ride down this one hill close to my house after a full charge, I get the overcharge tilt-back.

You'll occasionally need to fully charge the battery so the cells get balanced.  Then it can take a long time (days) for the battery to discharge enough to ride even on a level surface.   At least that was the situation I was in until I got the control board replaced.

The battery drain from the lights and control board when the wheel is just sitting is tiny compared to something like a 640 Watt hour battery.  So, I used three, 300 watt light bulbs to discharge the battery enough to ride after a full charge.

[fritzlakritz]

Is there any possibility checking the battery voltage without opening the wheel? There is no voltage at the plug measurable - why?

[DaveThomasPilot]

There is a diode between the charge connector and the battery.  So, current will only flow into the battery, not out.

So, you have to open the wheel to measure the battery at the connector.

Before I did that, I opened the charger and measured its output voltage. 

It should be higher than the battery voltage the app reports (by how much dependson the charge current), but the app reported battery voltage two volts higher than the output of the charger. 

[dmethvin]

2 minutes ago, DaveThomasPilot said:

There is a diode between the charge connector and the battery.  So, current will only flow into the battery, not out.

Which is funny because my Firewheel lacks that diode. The charging port is "live" with the battery voltage so you have to be careful about shorting it! If there was no diode the bug would become a feature.

3 minutes ago, DaveThomasPilot said:

Before I did that, I opened the charger and measured its output voltage. It should be higher than the battery voltage the app reports (by how much depends on the charge current), but the app reported battery voltage two volts higher than the output of the charger. 

Did anyone ever find out which chip is measuring the voltage? I wonder if we could just throw a few diodes in front of that to get the voltage to read lower.

[DaveThomasPilot]

My control board has some kind of coating on it that would make it difficult to hack.

But, if that could be scraped off somehow, there is probably a resistor divider somewhere that scales the battery voltage down to something the chip can measure.   The resistors could be replaced with lower value resistors while maintaining the appropriate ratio for the needed scaling.

The resistors might need to be physically bigger than what's on the board now, since their power dissipation will be higher when the resistance is lower.  But, I've done much "hackier" things than that.

But, I 've never seen a schematic of the KS control board, let a lone a layout diagram.  All I have seen was the schematic for a generic design that esaj posted.

If we could find the resistor divider that does the scaling, and scrape enough of the coating off to measure the voltage that the chip measures, we could at least confirm that's the issue.

[z4tchmo]

I wonder how common this issue is.  I had just changed my mind towards opting for a KS16S instead of an ACM due to all the issues Gotways have been having. If this is relatively common, that'd be pretty disappointing.  Granted, not nearly as disappointing as the wheel giving out or going nuts on you mid-ride due to firmware issues or bad hardware. 

[US69]

I would say thats NOT very common...the biggest 16s community sits in Singapur- "the wheelies" facebook group....

No mention of such problems over there....

[Christoph Zens]

4 hours ago, z4tchmo said:

I wonder how common this issue is.  I had just changed my mind towards opting for a KS16S instead of an ACM due to all the issues Gotways have been having. If this is relatively common, that'd be pretty disappointing.  Granted, not nearly as disappointing as the wheel giving out or going nuts on you mid-ride due to firmware issues or bad hardware. 

My KS16S reads accurately. At least to the point that it reports 100% charge after the chargers LED turned green. I did not compare the App reading to actual voltage at the battery, because as long as I don't have any problems, I'm not going to open it. It's not like you need to open a KS16 to rework the motor cabling or flash a new firmware before you can do any serious riding. 

[DaveThomasPilot]

My original board reads the battery voltage too high.  The replacement board is too low.

So, I'm thinking it's not rare...

If it's on the low side, I don't think most people would ever notice.   If it's too high, then you'll only notice with a full charge, and not notice if you use Charge Doctor.

 

[corwin]

You're lucky that you can tell you're charing to 76%. I just gone mine today but I can't read Chinese so I have no idea if it even works. There are two buttons and if I press them some lights flash. Other than the red light on the charger I don't see an indicator. You said something about an app?

[corwin]

I managed to get the app working. It shows 100%  Any change you can test the range to see how far "76%" will take you?

 

[Joker10]

I had the exact same problem.  It ended up being a bad cell in one of the battery packs.  I bought mine used from someone who just couldn’t learn it.  He purchased it from Jason McNeil at Ewheels.  I contacted Jason and he helped me troubleshoot it, mailed the parts needed to make the repair quickly, and got me all fixed up.  How can you beat that customer service.  The wheel was still under warranty and Jason covered the wheel for the second owner, yet another reason to purchase all your wheels from  Ewheels.  Once replaced the wheel charged right up to 100%.  My best range on flat ground with no music or headlight with the bad battery was 20 miles and about 18% battery left.  I just got back from a 13 mile ride with 750’ of ascent and had 68-70% battery left.  Impressive wheel.  I can’t recommend Ewheels enough.  Excellent company and excellent customer service.  

[ERKO929]

On 7/20/2018 at 4:29 AM, Joker10 said:

I contacted Jason and he helped me troubleshoot it, mailed the parts needed to make the repair quickly, and got me all fixed up. 

What parts where needed to make the repair? What did you do to the wheel?

[Joker10]

On 9/8/2020 at 5:59 PM, ERKO929 said:

What parts where needed to make the repair? What did you do to the wheel?

See my post from July 19.  It explained it but it needed up being a bad cell in one of the batteries.