King Song S22 motor stator slippage issue: more severe than expected

[supercurio]

Updates

As the story is developing, I will update this first post periodically but I recommend checking the recent posts in the thread. I am posting progress updates and everyone else contributes as well sharing their findings.
Last update Sep 26.

TL;DR

At the time of writing, all King Song S22 units in the hands of customers worldwide can be affected by a critical motor issue which is not easy to identify, and can result in a cutout in any riding scenario.

King Song hopes to have fixed the issues in motors produced from end of June/July onwards.

Dealers are doing their best to repair broken wheels but costs are getting out of hand.

There is possibly a test which tell if your unit is affected or not, but. more evaluation is needed to confirm its validity and reliability.
Update Sept 9: the test was able to break at least 10 defective motors, it appears effective, but might not be conclusive.

The issue can occur without aggressive riding or jumps. A new rider with S22 as first wheel had a cutout 2s after the loud scraping sound started.

My opinion

King Song completed a motor recall in China - for the same reason (educated guess). I believe it needs to happen worldwide eventually, in case the test is not 100% reliable.

King Song must make a public statement on the issue to educate riders on the problem, in order to reduce the risk of serious injury.

Description of the problem

A manufacturing defect in an unspecified amount of King Song S22 motors leads to its electro magnets slipping over the stator base which supports them instead of remaining in place.

This rotation has two effects:

• The hall sensor and connection between phase wire and windings become misaligned with the motor hub.
• An audible “clunk” sound when it happens (working theory at the moment: needs confirmation)

What it is not

This is not the knocking sound.

Knocking sound is a period knock caused by poor manufacturing tolerances of the motor side covers.

Bearings are not sufficiently held in place by the friction of their press fit with the side covers, so they move and shift as the wheel rotates, creating this sound. It is also an issue requiring attention but it is not as critical.

More details in this other post.

How to identify your motor production date

Described in this separate post:

Inside defective motors

Motor 1 stator rotated extensively, in the braking direction.

The temperature sensor got pulled away from its location (black wire); all the phase wires on the other side got severed.

This second photo shows how much rotation occurred, from where it should be on the left to where it landed.

 

Motor 2 (my own S22)

The temperature sensor (black wire) also got dislodged as well as some phase got disconnected from the windings.

Better view on one of the severed phase wire:

The phase wire created a short circuit on the side cover, which burned the mainboard (captured on video). We can see the traces of the short here:

King Song’s response so far

King Song replaced S22 motors before they exhibited issues in their Chinese domestic market in August.

For worldwide customers, the approach has been to handle issues case by case, only for motors which started to exhibit sign of failure when being ridden. The signs identified are:

• bad sounds (different sounds for 3 different known issues)
• motor tremors and vibrations at low or no speed (not officially recognized)

King Song has not made any public statement regarding the various types of ongoing motor issues.

Problems with the current approach

Impractical costs for dealers

While King Song promises to take care of every S22 motor breaking down and has been, all dealers have a very hard time convincing them to pay for air shipping for the replacement parts. Despite it is a requirement to fix customers’ wheels in a timely fashion.

Since most dealers were unable to get shipping costs covered, several already spend thousands of dollars in shipping. All repairs and shipping wheels back and forth are also at the expense of the dealers, with King Song covering none of the costs caused by them shipping defective products.

Safety risk to riders

Once the motor stator slipped, phase wires can get stripped which will lead to a short.

A short results in an instantaneous cutout. It means a crash in most cases.

Stress tests to identify a defective motor

Spin test

Notice how the motor emits loud “clunk” sound on some of the rotation changes. This is suspected to be the result of the stator slippage in question.

Regardless if the “clunk” sound is related to slippage, this stress test is likely to introduce slippage on defective motors.

When running the same test on motors recently produced (post supplier fix), no clunk sound occurred.

Update Sept 9:
This test will be even more effective by giving more time for the wheel to spin faster before changing direction.

 

Update Sept 26:
This test is now fully deprecated, for two main reasons:

• It tends to burn the mainboard as collateral damage in case of motor failure.
• Passing this test does not guarantee that the motor will remain free of defect later on.
  At least one wheel had a cutout during a customer ride, due to hall sensor error caused by stator slippage, after passing the extended version of this test "with flying colors" by the distributor. More details in this post: 
• The "clunk" sound during this test are not considered relevant anymore.

 

 

Spin test results

• Clunk sound might indicate that the motor is affected, but it is too early to tell for sure.
• The absence of clunk sound are not absolute proof that the motor is not affected. It could mean that:
  • The tester was not able to apply enough force or didn’t get the right timing
  • The motor can survive spin tests but might not handle the real world stress of a jump difficult landing

Limitation of this test for later observation: the stator might slip back in position thanks to the back and forth nature, and some of the silicon holding the temperature sensor might prevent slippage, holding the motor just enough.

Landing test

Without actually taking the risk of jumping with a bad landing, here is another proposed test:

• Free spinning at high enough RPM by lifting the wheel, then dropping it to the ground so the tire gets grip back, stopping rotation nearly instantly.
• ⚠️ It might be dangerous to execute: if the wheel jumps backwards or forward before stabilizing. However that might be enough to introduce slippage in only one direction.
• Reserve this test for outdoor, it'll leave mad skid marks
• This test is best suited to check later by opening the motor.

 

Update Sept 26:
The landing test is now the only known valid test, it is expected that an incoming King Song statement will describe the details on how to run it properly. It is confirmed to be able to force the issue to occur rapidly, with a clean failure which can be observed by the wheel as Error 17 (hall sensor error)

Watch out for a public announcement with a video demonstrating the proper form to run that one safely.

Opening the motor

By opening the motor before and after either of these tests, it is possible to determine if the stator slipped over the motor stator base.

Here is how:

1. Open the motor and set a mark on both the stator and its base which is connected to the hub, indicating the reference stator position.
2. Put the motor and wheel back together and complete a few rounds of the tests
3. Open the motor again and see if there was any stator slippage, using the mark as reference. Slippage might initially be of sub-mm distance, so the mark must be accurate enough.

Tests are destructive

Note that both these tests can be destructive for wheels with defective motors, whereas the issue might take longer to appear with normal riding.

These will be harmless for wheels with motors and mainboards free of defect.

If one of the tests result in any damage, although it is sad to destroy equipment , it also means it likely saved the rider from a later crash or even an expensive parts replacements after the warranty expired.

Recognize bad motor sounds

Update Sept 9: confirmation that failure can happen without prior scraping sound

Previous reports indicated that the issue occurred in the real world after audible scraping sounds of varying direction which could be noticed.
One of eWheels spin test highlighted that stator slippage, resulting in short-circuit of the phase wires can happen so quick that there was is no time for evolution from audible scraping to short.

As a result, listening for bad motor sounds is not a valid strategy to avoid sudden cut outs because of this issue.
The only solutions are passing without any abnormal sound the stress test demonstrated below (might not be 100% conclusive) or motor replacement to a July+ production unit.

More details here:

Motor clunks

It is suspected, but not fully confirmed yet that motor clunk sounds occur when the stator slips over the stator base.

 

Update February 26:

While the clunk sounds are still an area of concern in terms of how the mainboard controls the current sent to the motor, learnings from the ongoing investigation invalidated the link between between this specific sound and the phenomenon of stator slippage.

 

From Adam (Wrong Way) going up stairs:

https://youtube.com/clip/Ugkx4ryKlODDx8nZ7P8aQVJ6zfcKVJzHh0o_
https://youtube.com/clip/UgkxlP6V7s7TFEMb7KP_xXSAbcRZDZ7hwI8G

From Brian Norris on a heavy landing (clunk suspected):

 

Scraping sounds

Light, temporary, more on one side

Scraping sounds are likely to start by minor rubbing sounds coming from the motor. They are easy to dismiss as something else.

It is likely that this early scraping sound might get stronger and weaker depending on which side the wheel is leaning towards, like in turns or when carving.

The scraping sound is also likely to disappear at any point.

However if any scraping sound occurred, it is already indicative of serious internal motor damage. The motor status will range from:

• Temperature sensor detached (severity: low)
• Zip ties holding phase wires broken (severity: high)
• Phase wires coating scraping against the motor cover (severity: critical)

Riding must stop immediately at that point, and the motor should be either opened to diagnose the issue or replaced.

Louder, continuous

Later on, after disappearing, the scraping sound might come back for any reason. It happened on my own unit. It became louder than the initial scraping sound, and not knowing what it was I kept riding back home, only 1 km away (at walking speeds)

This continuous, louder state indicates that phase wires are scraping against the motor cover, which will certainly lead to a cutout and possibly a short once the coating worn off.

Now that we know what’s going on: don’t do like shown here and stop riding immediately. The wheel can cut-out at any point.

 

 

Cutout demonstration

When trying to make a video of the scraping sound for warranty purposes and before the issues was understood, I accidentally captured the motor killing the mainboard in this short video.

The result was unexpected.

 

Cutout can occur without scraping sounds

 

 

Areas of uncertainly

This document describes direct observations and conclusions based on them, however there are a few elements which are deductions or

educated guesses like:

• The nature of King Song domestic recall: we can’t tell for sure which problems were addressed. It means we don’t know if the stator slippage problem was indeed at the center of this recall, since there are other known issues like the knocking sound linked to bearing alignment. The reason is secrecy and general lack of transparency regarding the Chinese voluntary recall.
  Update Sep 26:
  Although this is still a contentious topic, we now have a good understanding on why motors were replaced in China, as well as the changes to the motors. Details 1, Details 2
• Generally, we don't know yet if the stator slipping problem has been identified, understood and fixed by King Song’s motor supplier
  Update Sep 26:
  We know it is the case now.
• There is a possible that the root cause for the NYC fire was stator slippage, pulling on the phase wires and shorting them (like shown on Motor 1 pictures) - instead of a mainboard fault at the time. It is impossible to tell without looking at this motor. Fortunately, the BMS bug which lead to the escalation into a fire has been fixed.

Credits

Special thanks to Voltride who understood the nature of the issue first and took all the pictures illustrating the motor defects and MyEWheel for testing newly received units.

Thanks to King Song customer representative for some very useful information helping to understand the context. I hope for a continued fruitful conversation.

Props to everyone in the EUC safety group and telegram group created for the occasion. If you are a distributor, please get in touch to become part of the group of dealers working together on solutions for this.

Edited September 26, 2022 by supercurio

[UniVehje]

Now the question is: are all the early motors defective and those still working just haven’t been stressed enough, yet.

[supercurio]

@UniVehje yes tough question, I guess there's several scenarios.

• QC problem leading to the stator slippage due to something like glue missing on some motors (wild uneducated guess), or bad quality materials that made it in the supply.
• Molds for parts with the wrong tolerances, so what's expected to fit and hold slips instead, on all units until the mold was fixed.
• Only one of the molds had the wrong tolerances, so a set % of motors are defective until it was removed from production.

Edited September 1, 2022 by supercurio

[Tawpie]

@supercurio is eWheels part of your distributor group? Curious what they'll do with this information now that they have 300 units set for F'dEx pickup tomorrow.

[supercurio]

@Tawpie eWheels is not part of the group chatting together on a regular basis, but yes I gave @Jason McNeil a heads up as soon as we made the discovery, and sent him the completed draft yesterday.
He confirmed that he's in now in touch with KS on this specific issue. I hope eWheels shipment comes with motors with the fix for the stator slippage, but I'm not sure.

To be continued.

Edited September 1, 2022 by supercurio

[Goonman]

In all my time mucking around with ebikes and hub motors generally I have never seen this or imagined it was even possible. 

[supercurio]

@Goonman indeed, before posting I searched for this problem being documented elsewhere but was not able to find anything. Not really the kind of innovation we were hoping for.

By the way, if anybody has a better name for this I'm all ears - I had to come up with something.

[Forwardnbak]

I have seen you working on this. It is outstanding you took the time to organise and chase all this up. 

Congrats Francois! really concise info that will help riders and dealers.

 

 

[Rich Sam]

@supercurio, that may be the most thorough root cause analysis I have seen yet on PEVs!  Lovin it!  Its a real service to empower the public with this knowledge instead staying mums the word.  Big hats off to you!  I was extreemly curious what was causing the S22s in the Seattle area to be dropping like flies.

 

[supercurio]

Appreciated @Rich Sam, glad you enjoyed the read!

From what I gathered, the S22s in Seattle broke from separate issues: the mainboard failing without the phase wires being physically shorted.
There's no shortage of concurrent issues on the S22 for now.

Edited September 1, 2022 by supercurio

[alcatraz]

If they offered some official self repair videos/kits maybe someone would start to offer a service on the same continent, and move towards becoming an authorized service place.

I bet some buyers or resellers would rather fix it themselves than send it all back.

[supercurio]

@alcatraz it might be possible indeed to locally repair all motors before they destroy themselves 

I just watched this video which highlights how the stator is attached to its base. And it's with glue indeed, but stators can also remain in place on their base thanks to a friction fit.

I still don't know if the S22 motor stator slips because of insufficient glue or friction fit, but it has to be one or the other.
In case removing the stator of these dimensions doesn't require impractical forces which would damage it in the processes then yes they could be repaired, essentially by glueing the stator to its base correctly.

 

[NSFW]

Fantastic post. Thanks for the detailed write-up.

[mikko]

@supercurio Did you have pictures from the other side of the motor? There's the black plastic ring-thing that has tabs going under each coil, which has moved out of place with the stator. Is it that plastic that tries to hold it in place or why does it have those other wider tabs that extend inwards towards the center?

[mhpr262]

Holy moly, this is much worse than the cutout issue of the Inmotion V12. Almost 300 years of building electric motors and we STILL havent figured it out?

 

EDIT: if the stator really held on the hub base by glue and/or friction only that is really poor engineering. I have no engineering qualifications whatsoever, just a bunch of experience from building RC cars and planes but it is obviuos to me that there should have been another, mechanical means of holding the stator in place, like small, intermeshing teeth like on gears, or a couple of large "nubs" on the inside of the hole in the stator and corresponding "holes" in the base. This would have prevented any slippage, and with cast parts it would have had no additional cost. Or even just a couple of TIG weld spots, its not like you have to separate stator and hub base on a daily basis.

Edited September 2, 2022 by mhpr262

[Chriull]

Unspecified "I warned you" post and follow up posts move to 

to keep this topic clear for further information.

Plus topic is pinned, so it's easily found!

[supercurio]

1 hour ago, mikko said:

@supercurio Did you have pictures from the other side of the motor? There's the black plastic ring-thing that has tabs going under each coil, which has moved out of place with the stator. Is it that plastic that tries to hold it in place or why does it have those other wider tabs that extend inwards towards the center?

Sorry I don't have pictures from the other side of the motor, the motors were not open on the phase wire side.

But that's a very good point @UniVehje also noted in a discussion on Telegram. It's useful to look at this black ring as in the existing picture, even if the rotation is not obvious, the ring tabs which are supposed to be aligned with the spokes show fairly precisely how much rotation occurred.

Using this as reference, anybody opening their motor can quickly check if the stator slipped without adding a mark as prior reference 👍

The rotation can be seen easily in this close-up picture from the original set.

[UniVehje]

And for reference, here's a picture of a good motor (mine) with the tabs lining with the spokes. The tabs can be seen better on the side where the motor cable doesn't come out from (at least on mine, in the picture it has been moved). 

[supercurio]

@mhpr262 it sounds like my post make its point convincingly, however I don't think this specific problem is as bad as the V12 original boards cutouts. Or at least, it's different, mainly:

In most or at least some of the cases, there will be warning signs with motor scraping sounds. With the V12, there were no warnings whatsoever.
Although these signs can be missed, or maybe be absent in some case if the failure is more sudden like during a hard braking or bad landing instance.

There were scraping sounds in all document cases so far which is a positive.

The negative is that the S22 also has an issue with boards dying suddenly, so both compounded is not great.

 

2 hours ago, mhpr262 said:

EDIT: if the stator really held on the hub base by glue and/or friction only that is really poor engineering. I have no engineering qualifications whatsoever, just a bunch of experience from building RC cars and planes but it is obviuos to me that there should have been another, mechanical means of holding the stator in place, like small, intermeshing teeth like on gears, or a couple of large "nubs" on the inside of the hole in the stator and corresponding "holes" in the base. This would have prevented any slippage, and with cast parts it would have had no additional cost. Or even just a couple of TIG weld spots, its not like you have to separate stator and hub base on a daily basis.

Maybe someone building EV motors can share knowledge on how it is typically (successfully done) to prevent stator slippage.
It would be nice to know, especially to check the validity of the solution adopted in the fixed motors and be confident it doesn't happen again.

Edited September 2, 2022 by supercurio

[supercurio]

Progress update:

King Song started to informed dealers that there are working on a test to identify bad motors.
I suggested them to investigate the spin test showed above and confirm if the clunk sounds are indicative of rotor slippage.

V12 history taught us that stress tests are not always guaranteed to catch all defective units, as some V12 HS boards kept cutting out after passing the stress test.
But it might work in this case: worth a shot.

[supercurio]

Another case reported on the reddit thread, reseller: EUCO, country: US, new rider (first wheel)

Notable that the cutout can happen only a couple seconds after the louder rubbing sound appears.

It's because the louder rubbing sound are due to the phase wires rubbing on the side covers. When that happens it's quickly game over.

[Planemo]

18 hours ago, Goonman said:

In all my time mucking around with ebikes and hub motors generally I have never seen this or imagined it was even possible. 

Same here, never heard of this problem on ebikes. That said, most ebikes using Direct Drive hubs are not used off road and are certainly not likely to experience the same sudden directional forces that a jumping/heavy braking whilst skipping EUC will. Personally I'd like to see the stators pinned to the hub in at least 3 places to sort this problem.

[supercurio]

@Planemo important to note that the issue can occur without the aggressive riding, jumping an experienced rider would do, or running the stress test.

One of the case is a new rider who learned to ride on the S22 as first wheel.
It would be comforting to think it can only happens if you push the wheel really hard, but it doesn't seem to be the case.

[nitecrawlr]

eWheels Update:

"""

Good morning, 

 

With the S22 arrivals last evening, we're preparing to ship out the orders this afternoon.  

 

For full-disclosure, in the past 24hrs, there's been a recent forum posting about a potential motor issue on the S22. Since there's several thousand S22s in circulation, there's a decent body of statistical reliability data. According to King Song, this fault condition affects between 5-15% of the Wheels shipped between June & current. 

https://forum.electricunicycle.org/topic/29877-king-song-s22-motor-stator-slippage-issue-more-severe-than-expected/

 

In anticipation of any faults of this nature, we've flown in a supply of controllers, & the motors will be arriving in the next couple weeks. 

 

If you would prefer holding off on shipping out your S22 today, until the dust settles, & a clearer picture is formed, or for the next round in about 5-6 weeks, please let us know & we'll have your order processed accordingly. 

 

Sorry again about this troubled product release. We're committed to supporting you. however you choose to proceed. 

 

Thank you,

 

Jason 

 

""""

 

[supercurio]

Excellent transparency and very good message to address the issue from eWheels 👏
I noticed @Alien Rides also started to answer customer questions on Facebook on the topic.

It's great to see sellers both communicating about it and taking measures ahead of time.

Next, I would like to see King Song make a public statement to educate customers about the risks and teach them what to look for in terms of early signs.