SAFETY ADVISORY: Faulty charger may lead to risk of 16X battery fire

[EvoKnvl]

So I'm confused now.  Is this ALL chargers supplying more than 5A, or is this an isolated case of 1 charger?  Because it seems like the specific ewheels charger in the email had some weird issues going on with it.  Versus the title leads me to believe charging over 5A may burn your 16x to the ground. 

[Marty Backe]

49 minutes ago, Mark Lee said:

My 16X came with 2.5 Amp charger.

You sure about that? Look at your charger closer. The input current is a maximum of 2.5 amps but the output current is 1.5 amps max.

[EvoKnvl]

29 minutes ago, alex5454 said:

read back of charge...OUTPUT 1.5A 84 V....

 

I believe there are some vendors that are offering a 2.5A charger with the 16x, vs the 1.5A

https://tec-toyz.com/products/kingsong-ks16x-1554wh-ver-2-electric-unicycle-preorder

[xorbe]

My 16X eWheels 5A rapid charger shows 0.00V until it's actually charging the 16X [ie, so I can't confirm 84v before charging].  It is not like my other rapid chargers that show peak voltage while idle (0.00A).  And I have to click the button several times sometimes.  Or unplug the charger from the mains, wait, and plug the charger back in.  It behaves totally different than my other ones.

 

About the charging to 88.2v ... one cell too many.  Random firmware bit-flip failure?  No checksum/ecc apparently.

Edited October 15, 2019 by xorbe

[Marty Backe]

1 minute ago, EvoKnvl said:

So I'm confused now.  Is this ALL chargers supplying more than 5A, or is this an isolated case of 1 charger?  Because it seems like the specific ewheels charger in the email had some weird issues going on with it.  Versus the title leads me to believe charging over 5A may burn your 16x to the ground. 

From the description of the problem I'm pretty sure it's an output voltage that is too high which is causing the problem. It's not a problem with the amperes.

[Mark Lee]

29 minutes ago, Marty Backe said:

You sure about that? Look at your charger closer. The input current is a maximum of 2.5 amps but the output current is 1.5 amps max.

You are right, as usual thx. So KS charger 1.5 amps + 3 amps ( 2 x 1.5 amp GW charger with charge Doctor) = 4.5 amps.  Close enough I guess.

Edited October 15, 2019 by Mark Lee

[JZT-Colorado]

Could it be that the rapid charger just isn’t communicating properly with the BMS?

Here’s a clue:

I noticed with my own 16X that when I set my rapid charger to charge to only 80%, it kept charging all the way to 90%+... at which point I just unplugged it as I didn’t want to charge to 100%.  But if it kept charging past 100%, it would probably lead to a similar result. 
 

Hmmm...

Edited October 16, 2019 by JZT-Colorado

[Obah]

This information is not true!!! about KS16X....

Why will it be a Kingsong and not Goyway in this knock down campaign?

A faulty charger is the case, kingsong has nothing to figure out with charger producers... Neither do charger producers with kingsong (a bad charger should be replaced simple) 

5A fast charger on KS18X 84v is same as every other 84v wheel regardless of name, size, speed or capacity.

KS16X is currently the best 16series wheel on planet earth, it is well engineered and nicely built and holds a position in the top 3 wheels.

 

[FlightRisk]

My 16x is dead. Only maybe 20-30 miles total, had a successful first charge with standard charger couple weeks ago. Topped it off with the rapid charger yesterday (set at 90%). It oddly continually double-beeped every 10-15 minutes before it was done which took about an hour (battery had +/- 40% juice). Went to fire it up and it promptly died. No sign of life since. Tried to revive with standard charger, no bueno :-(

On the phone with eWheels today, they sounded very overwhelmed since a 16x caught fire after a rapid-charge attempt. Use standard charger til this gets hammered out.

BTW, Jason at eWheels was super helpful. I didn't expect such a rapid response and had a great chat with him as he picked my brain trying to figure out what's up. Kudos, Jason!

[xorbe]

11 minutes ago, FlightRisk said:

My 16x is dead. Only maybe 20-30 miles total, had a successful first charge with standard charger couple weeks ago. Topped it off with the rapid charger yesterday (set at 90%). It oddly continually double-beeped every 10-15 minutes before it was done which took about an hour (battery had +/- 40% juice). Went to fire it up and it promptly died. No sign of life since. Tried to revive with standard charger, no bueno :-(

On the phone with eWheels today, they sounded very overwhelmed since a 16x caught fire after a rapid-charge attempt. Use standard charger til this gets hammered out.

BTW, Jason at eWheels was super helpful. I didn't expect such a rapid response and had a great chat with him as he picked my brain trying to figure out what's up. Kudos, Jason!

Wow, is there a way to get a voltage reading on your battery pack for safety's concern?  Mine's at 43 miles here in Mt View.

Edited October 16, 2019 by xorbe

[meepmeepmayer]

1 hour ago, JZT-Colorado said:

Could it be that the rapid charger just isn’t communicating properly with the BMS?

These chargers are cheap and dumb. They increase the voltage, limited by a max current, until they're at 84V. Then they stay at 84V (or a little more to account for losses on the way) to top off the battery. I don't believe there is any communication between wheel and charger, or even a possible communication channel.

So, it seems the problem here is that both of these failures happen at the same time:

• The defective charger increasing the voltage above 84V and not stopping there.
• The wheel (board or BMS) not detecting the overvoltage and shutting down the charging automatically, even if the overvolted charger is plugged in.

With the result that there is an overvoltage which can cause the battery to catch fire.

Or maybe it is some different mechanism that caused the one fire, don't take my word for it.

1 hour ago, JZT-Colorado said:

I noticed with my own 16X that when I set my rapid charger to charge to only 80%, it kept charging all the way to 90%+... at which point I just unplugged it as I didn’t want to charge to 100%.  But if it kept charging past 100%, it would probably lead to a similar result.

What voltage the charger considers "80%" and what voltage the wheel considers "80%" can differ. There were reports of setting the fast charger to 80%, which charged the wheel above 80%, so this might be normal. At least if the charger stops charging at some point.

[FlightRisk]

28 minutes ago, xorbe said:

Wow, is there a way to get a voltage reading on your battery pack for safety's concern?  Mine's at 43 miles here in Mt View.

I'll just send it back at this point, too worried about it cutting out while on it. Of course that would depend on it actually working, no signs of life!! Definitely stay off the rapid charger.

Email to me today from Jason at eWheels:

"IMPORTANT URGENT Safety Advisory on the 16X & 5A Rapid-charger!

 

We have just received an alarming report of a 16X catching on fire after being used with a defective charger. Under the circumstances, it was extremely fortunate that the Wheel was outside at the time, where the damage was relatively minor. 

 

Here is a summary of the facts:

1. After a few weeks of normal use, the Customer noticed that when the charger was plugged into the wall (but not the Wheel), the charger's LCD readout was showing '1 . ' on the Voltage readout (see image below), instead of the normal ~84v
2. The charger was emitting a low frequency clicking sound
3. Pressing the charger's reset button kicked off the charge cycle, where the voltage had continued to climb exceeding 88v, as represented on the charger's LCD display
4. 15 minutes later, the rapid-charger turned off, but on turning on the Wheel, the Customer observed the voltage (reported through WheelLog) completely discharging in seconds 
5. He then plugged in the standard charger, about one hour later the Wheel caught on fire 

What is being done & recommendation: 

This charger is being sent to an independent laboratory for analysis, I am also coordinating for the King Song 16X battery to be sent in the laboratory for over-voltage testing.

 

Until we receive the lab results back & have a complete understanding of exactly what went wrong, it is strongly recommended not to continue using the rapid-charger.

 

WARNING: if you do continue using the rapid-charger, please double-check that the voltage readout displays ~84v before plugging into the Wheel & keep an eye on the charging progress.

    

For context, the charger supplier, JaiRui, has produced hundreds-of-thousands of chargers, we have seen many cases of the charger failing in a multitude of ways, but invariably it has been in a harmless under-voltage state. This is the first & only such case of an over-voltage condition, but considering the extreme severity & intensity of a battery fire, other chargers from this production batch may also be effected. 

 

We will continue to update you as more information becomes available. "

[mrelwood]

6 hours ago, Marty Backe said:

Is this really true? So you can pump 100-volts into a 84-volt wheel with no problems? Hmmmm.

The BMS should stop the charge to protect the battery from too much overvoltage. But the BMS limits of course have to be higher than what a healthy charger will put out, so it would still cause a slight overvoltage.

As seen in this case, the BMS may indeed fail on recognizing the overvoltage, so I wouldn’t push it. Also, a 100V charger would not switch to the CV stage of the charge, meaning the battery charge would drop to somewhere around 80-90% when the BMS (should) stops the charging.

 

4 hours ago, Obah said:

Why will it be a Kingsong and not Goyway in this knock down campaign?

A wheel and it’s manufacturer gets targeted if they have issues, simple as that.

4 hours ago, Obah said:

A faulty charger is the case, kingsong has nothing to figure out with charger producers...

The BMS in the 16X and all wheels is equipped with an overvoltage protection that should stop the charging if the charger goes haywire like it did in the case of the fire. The BMS on the 16X failed to do this, so it was faulty as well.

 

4 hours ago, Obah said:

KS16X is currently the best 16series wheel on planet earth, it is well engineered and nicely built and holds a position in the top 3 wheels.

I’m glad that you love your wheel as much as you do, but that doesn’t make it the best for others.

[Seba]

7 hours ago, meepmeepmayer said:

Feel free to make the title less alarmistic. Didn't know what to call it

Exactly. Title is at least misleading.

 

6 hours ago, photorph said:

to be on the safe side for now avoid using the 5A rapid charger. 

You should clearly state both in the post title and in text that you meant exact charger type sold by eWheels. The problem you are posting about has nothing to do with the charging current and it could very well occur with a damaged 1.5A stock charger. In this forum there is a whole bunch of people from around the world who use a variety of chargers with a 5A charging current.

Definitely there is no risk of KS-16X catching fire with using 5 A charger that is not faulty. This thread describes a case in which a 5A quick charger purchased from eWheels most likely had an incorrect, too high output voltage, which resulted in a state of overcharge of the battery and, consequently, a fire. It's however a mystery to me why BMS didn't detected cell overcharge condition and consequently didn't disconnected charging port.

 

7 hours ago, buell47 said:

But actually the charger should only be switched on after it has been plugged into the EUC. So first wheel then wall.

This is no longer the case with current wheels and chargers. You can safely turn your charger to the wall, verify it's working correctly (in this case by checking voltage readout) and then connect it to the EUC. This advice was actual in the ancient times of lead acid batteries and "transformer/rectifier" unregulated and unprotected chargers  With modern, current limited CC/CV chargers there is no risk of arcing on the charging port side. Arcing may however still occur when connecting to the wall socket because there are large capacitors on the primary charger circuit side. 

[EvoKnvl]

1 hour ago, Seba said:

Exactly. Title is at least misleading.

 

You should clearly state both in the post title and in text that you meant exact charger type sold by eWheels. The problem you are posting about has nothing to do with the charging current and it could very well occur with a damaged 1.5A stock charger. In this forum there is a whole bunch of people from around the world who use a variety of chargers with a 5A charging current.

Definitely there is no risk of KS-16X catching fire with using 5 A charger that is not faulty. This thread describes a case in which a 5A quick charger purchased from eWheels most likely had an incorrect, too high output voltage, which resulted in a state of overcharge of the battery and, consequently, a fire. It's however a mystery to me why BMS didn't detected cell overcharge condition and consequently didn't disconnected charging port.

 

This is no longer the case with current wheels and chargers. You can safely turn your charger to the wall, verify it's working correctly (in this case by checking voltage readout) and then connect it to the EUC. This advice was actual in the ancient times of lead acid batteries and "transformer/rectifier" unregulated and unprotected chargers  With modern, current limited CC/CV chargers there is no risk of arcing on the charging port side. Arcing may however still occur when connecting to the wall socket because there are large capacitors on the primary charger circuit side. 

Agreed, but not surprised...

[mrelwood]

I edited the thread title to hopefully describe the issue better. Please correct me if I have understood the issue incorrectly.

[Obah]

3 hours ago, mrelwood said:

The BMS in the 16X and all wheels is equipped with an overvoltage protection that should stop the charging if the charger goes haywire like it did in the case of the fire. The BMS on the 16X failed to do this, so it was faulty as well.

have you ever wonder what causes BMS to fail on wheels? 

First is the charging state; overvoltage + over current...(input) If the power from your electricity could damage chargers and other electrical appliances, how much more a detective charger to a  BMS which is just there to manage.

Example: When you plug in an 84v charger to a 60v battery pack it gets heated in minutes without been fully charge, this i think is the case. BMS on the KS16X didn't fail rather the defected charger gave a higher input (overvoltage + over current)(over 84v + over 5A) the BMS are programamed to shut down overvoltage and not current.

Again, battery got over heated and BMS didn't detect it because battery was not fully charged at that point to cut off overvoltage, this is common with hoverboards, etc. avoid cheap and poorly built chargers.

Slow charge= slow discharge/healthy battery life

Fast charge= fast discharge/reduces the battery life span

 

Edited October 16, 2019 by Obah
Correct spelling

[Seba]

4 minutes ago, Obah said:

When you plug in an 84v charger to a 60v battery pack it gets heated in minutes without been fully charge

No  When you plug 84 V charger to 67.2 V battery pack it will start charging as usual (of course if the battery is discharged). Why? Because every Li-Ion charger is just a power supply with limited output current and voltage. The only difference is that when battery will get fully charged, charger will still be in "constant-current" mode and will start to overcharge battery. Now BMS should activate overvoltage protection by turning off MOSFET transistor in charging path. Question is why this protection didn't kicked in in this particular case.

[mrelwood]

@Obah, I’m having a hard time following your points and reasonings, but I’m pretty sure the overvoltage detector doesn’t care about the charging current, since it is supposed to work even without the charger being connected. A simple voltmeter at the battery mains, and if the voltage is too high -> disconnect.

Ps. You might want to hide one of your double posts.

[Chriull]

12 hours ago, meepmeepmayer said:

Sounds like the situation isn't 100% clear right now. Usually the BMS should prevent overcharging, shouldn't it?

 

11 hours ago, Marty Backe said:

Is this really true? So you can pump 100-volts into a 84-volt wheel with no problems? Hmmmm.

 

11 hours ago, meepmeepmayer said:

That was a real question. I don't know if the BMS is supposed to switch off the charging if the input voltage is too high.

Nowadays EUC BMS monitor each single cell (pair/triple - the paralleled ones) for overvoltage (some 4.2xV). If the threshold is reached the charger input is cut off with Mosfet(s).

One idea came to my mind, as @photorph mentioned the overflow of the charger output voltage display, which could mean more than 99V. In such a case this voltage could be above the specification of the mosfet (drain source breakdown voltage) damaging him.

But the fire happened with the normal stock charger afterwards there (without overvoltage) some bad/steongly mismatched cells would be needed to cause troubles. ... Or just some loose contacts/bad components or whatever. The possibilities would give space for endless speculationd - so i'm looking forward to the result of the lab analysis.

[Seba]

This thread inspired me with the idea to add an additional overvoltage alarm to WheelLog. Such an alarm will be useful not only when going downhill with a fully charged battery, but also during charging. The application can connect over BT to King Song wheel during charging and allow you to monitor the charging process.

[Obah]

3 hours ago, mrelwood said:

@Obah, I’m having a hard time following your points and reasonings, but I’m pretty sure the overvoltage detector doesn’t care about the charging current, since it is supposed to work even without the charger being connected. A simple voltmeter at the battery mains, and if the voltage is too high -> disconnect.

Ps. You might want to hide one of your double posts.

@mrelwood...3 questions you might love to take personal..

1. What is the role of a mosfet on a BMS board?

2. When a BMS fails what actually got damaged on the board?

2.what can possibly damage a BMS or its mofset?

Riding experience/mod idea on a unicycle is not enough electrical/electronics knowledge 

Edited October 16, 2019 by Obah
Edit

[Chriull]

19 hours ago, Seba said:

This thread inspired me with the idea to add an additional overvoltage alarm to WheelLog. Such an alarm will be useful not only when going downhill with a fully charged battery,

Yes.

Quote

but also during charging.

Only limited - just if all cells are fine and balanced. So this could lead to a wrong wrong feeling of safety :(. Imo the only solution is the usage of BMS which also monitor single cell undervoltage and report this to the mainboard/app (and in no case like the very first generations did cut off the discharge port...)

Quote

The application can connect over BT to King Song wheel during charging and allow you to monitor the charging process.

But maybe also some "abonrmality" can be detected by wheellog with not fully charging wheels. But there are ?three? cases:

- the charger does not deliver enough max voltage, so it starts the CV phase at too low voltage. Could be seen by the constant current decreasing at not maximum voltage.

- the main board reporting a wrong voltage. Mabe some correction factor can/should be implemented if the users verifies the charger voltage? Or just some "margin" is introduced?

- Charger output and mainboard is behaving fine, but some cells are dead. I have no idea how this will behave? The voltage will stay at some too low voltage - interesting would be the "current behaviour". If it still stays like in the CC phase or show some other anomalies?

Edit: This last option will just charge with some CC up until the first cell reaches the overvoltage cutoff threshold - so some premature charging end (thanks to https://forum.electricunicycle.org/topic/16045-safety-advisory-faulty-charger-may-lead-to-risk-of-16x-battery-fire/?do=findComment&comment=273279

)

I'll try to monitor some charge cycles and compare them to https://batteryuniversity.com/learn/article/charging_lithium_ion_batteries and https://batteryuniversity.com/learn/article/bu_409a_why_do_old_li_ion_batteries_take_long_to_charge.

So if one makes such logs from time to time and compares the curves battery state should be "easily" determinable? With other members input (enough charging wheellog data file - especially from "abnormal" battery pack/charger combinations) maybe a nice battery pack analysator can be evolved?

Edited October 17, 2019 by Chriull

[Obah]

4 hours ago, Seba said:

No  When you plug 84 V charger to 67.2 V battery pack it will start charging as usual (of course if the battery is discharged). Why? Because every Li-Ion charger is just a power supply with limited output current and voltage. The only difference is that when battery will get fully charged, charger will still be in "constant-current" mode and will start to overcharge battery. Now BMS should activate overvoltage protection by turning off MOSFET transistor in charging path. Question is why this protection didn't kicked in in this particular case.

Yes🤔 same lion batteries with limited voltage and current are the ones burning out motor on a wheel and frying control board?

Maybe engineers didn't calculate properly before built?

 

[RetroThruster]

13 hours ago, JZT-Colorado said:

Could it be that the rapid charger just isn’t communicating properly with the BMS?

Here’s a clue:

I noticed with my own 16X that when I set my rapid charger to charge to only 80%, it kept charging all the way to 90%+... at which point I just unplugged it as I didn’t want to charge to 100%.  But if it kept charging past 100%, it would probably lead to a similar result. 
 

Hmmm...

Mine did the same thing, I had it set to stop at 80% but when I checked WheelLog it stated that it was at 92%, at this point, I stopped charging, I'm a little fuzzy on what the charger was reading at at that time but it seems like it was just barely over 80% but had stayed there for quite some time, I wonder if it's a combination of the charger charging at 88v and the BMS not recognizing the issue and stopping it AND the charger not stopping at the selected percentage, if at all.  

Edited October 16, 2019 by RetroThruster