INMOTION V12

[Chester Copperpot]

Ok, wow, I now have a whole new twist on the cutout situation. To summarize what I have previously posted, mine cut out in mid December while walking off a small step.

As others have experienced, it would not turn on unless it was plugged into the charger. The wheel would not spin freely by hand. Ewheels said it is likely blown mostefs and is sending me a new board. 

I had previously sent my logs to IM. Last week I decided to email IM as suggested by Cecily to let them know to associate my log with my issues. They responded this morning asking for a video.

When I went to make a new video, it turned on and balanced like normal!!! I pushed it around the garage and ran it backwards and forwards and it acted like nothing had ever been wrong with it!!! That’s all I could do because I had to head to work.

Now I don’t know what this means. I most certainly don’t trust it. It tried turning it on probably 50 times in the past month in different scenarios and it never worked when not plugged in.

I’d found that if you put it in travel mode you could turn it on while not plugged in but as soon as you turn off travel mode and it tried to apply power to the motor it would instantly cut off.

All of those issues had been previously reported to Ewheels with video, so I’m not crazy, lol. 

[supercurio]

If a small bump, including at walking speed with or without carrying a rider and reaction to tilt is all that's needed for the board to fail:

Here's proposed stress test that's safe:

1. Grab your running V12 and lift it without disabling the motor
2. Tilt the wheel forward or backwards to get the motor spinning using your second hand.
3. Yank the wheel abruptly to the reverse tilt angle: that'll force the wheel to slow down then start turning in the reverse direction.
4. Repeat at various tilt angles, speed and intensities.

This will simulate some of the peak power demands experienced with bumps and going onto/off a curb.

As long as you get a good grip on the wheel itself, this test is not dangerous. Be sure to not do it for too long so your arms could get tired to the point the spinning wheel could slip out of your hands.

[yoos]

12 minutes ago, supercurio said:

This will simulate some of the peak power demands

I am not sure this will work properly since this only recreates sudden tilts but there is no load. I assume that current is not a function of just tilt alone but more complicated than that. I would speculate that the very quick change in tilt produced by a curb makes the controller panic (the algorithm might extrapolate and think the wheel is about to tilt even further) and push an insane current as a response.

Still, the test is a good start and should be done. [Better yet, recreate the real cutout situation and do some curbs ]

[Grimm10]

2 hours ago, Chester Copperpot said:

Ok, wow, I now have a whole new twist on the cutout situation. To summarize what I have previously posted, mine cut out in mid December while walking off a small step.

As others have experienced, it would not turn on unless it was plugged into the charger. The wheel would not spin freely by hand. Ewheels said it is likely blown mostefs and is sending me a new board. 

I had previously sent my logs to IM. Last week I decided to email IM as suggested by Cecily to let them know to associate my log with my issues. They responded this morning asking for a video.

When I went to make a new video, it turned on and balanced like normal!!! I pushed it around the garage and ran it backwards and forwards and it acted like nothing had ever been wrong with it!!! That’s all I could do because I had to head to work.

Now I don’t know what this means. I most certainly don’t trust it. It tried turning it on probably 50 times in the past month in different scenarios and it never worked when not plugged in.

I’d found that if you put it in travel mode you could turn it on while not plugged in but as soon as you turn off travel mode and it tried to apply power to the motor it would instantly cut off.

All of those issues had been previously reported to Ewheels with video, so I’m not crazy, lol. 

Maybe inmotion pushed through an update to the wheel?

[Khazik]

7 minutes ago, yoos said:

[Better yet, recreate the real cutout situation and do some curbs ]

^ This, lol.

Dress for the slide, not the ride.™

[supercurio]

42 minutes ago, yoos said:

I am not sure this will work properly since this only recreates sudden tilts but there is no load. I assume that current is not a function of just tilt alone but more complicated than that. I would speculate that the very quick change in tilt produced by a curb makes the controller panic (the algorithm might extrapolate and think the wheel is about to tilt even further) and push an insane current as a response.

Still, the test is a good start and should be done. [Better yet, recreate the real cutout situation and do some curbs ]

Hehe yeah no guarantee. I would try this test over and over if I had a V12 at hand now.

In terms of load, it's essentially only the kinetic energy from the wheel inertia (motor + tire weight) making it a flywheel.

I noticed that on Sherman, it's possible to overpower the motor that way, when changing direction: the power demands in such test can still be high. At the same time, on Sherman you have a really strong grip to do so with the front and rear handles, it might not be as easy to replicate with a V12.

Edited January 4, 2022 by supercurio

[Paul g]

7 hours ago, The Brahan Seer said:

From reading this thread it sounds like people are wanting EUC manufacturers with Quality Assurance Standards eg ISO9001. The advantage of this is obvious but the downside is cost. If IM are to send out reps, do recalls etc etc to remedy the issues then that's great as long as we are prepared to pay for it. Ultimately we know we are taking a risk with this product especially first batches. If the alternative is businesses shut down and go bust how would you feel?

??? Makes no sense what I read here

[terlikaa]

From Facebook:

"Hi All.
This is an email sent by eWheels in regards to the cut outs:
I had a telephone call with Inmotion this morning about this situation. Here's the facts that we know of so far: 
- The root cause is with the Infineon MOSFET chips, during manufacturing, not all of them may be up to spec
- So far there's been about a 10% failure rate on these controllers. In nearly all these cases are under no-load spin, going over a curb at low speed, or taking off from the charger & all within the first few weeks of taking possession.... 
- Inmotion's procedure for stress-testing these controllers is going to be spinning up, with the lift button disabled, then reversing the motor direction. If it survives this test, then they're confident that the boards are going to hold up. 

We're going to be sending out a mass communication update on this tomorrow, once they have a video clip for this stress-testing procedure. 

In summary, if the V12 has survived so far, it really should not be among those with the predisposition to FET failure."

[meepmeepmayer]

2 minutes ago, terlikaa said:

The root cause is with the Infineon MOSFET chips, during manufacturing, not all of them may be up to spec

Oh wow, you can trust nobody. Though I wonder if these mosfets are simply too small or have a too narrow margin for a wheel like the V12.

Great that Inmotion has a result and communicates it, though!

Edited January 4, 2022 by meepmeepmayer

[supercurio]

5 minutes ago, terlikaa said:

- Inmotion's procedure for stress-testing these controllers is going to be spinning up, with the lift button disabled, then reversing the motor direction. If it survives this test, then they're confident that the boards are going to hold up. 

We're going to be sending out a mass communication update on this tomorrow, once they have a video clip for this stress-testing procedure. 

Oh! It looks like the testing procedure will be the same as what I suggested 2h ago then.

[Paul g]

1 minute ago, meepmeepmayer said:

Oh wow, you can trust nobody. Though I wonder if these mosfets are simply too small or have a too narrow margin for a wheel like the V12.

Great that Inmotion has a result and communicates it, though!

I doubt the mosfets are faulty, from what we read on electronics forums their simply not the proper mosfets for the job. If not like that, than a less-than-ideal quality wheel producer like Begode would not put in their wheels 25% margin mosfets for no reason. And if some of the 100 V mosfets are a little more resistant and pass the 100,8V test on V12 is pure luck, not because their manufacturer guaranties for it.

[ESB]

19 minutes ago, terlikaa said:

In summary, if the V12 has survived so far, it really should not be among those with the predisposition to FET failure."

Hmm. Not very reassuring given what the person who went to ER last week said (200 miles. All good until it wasn’t).

Curious to see the instructions in the video. Props to ewheels for sending out mass email.

[Paul g]

Just now, ESB said:

Hmm. Not very reassuring given what the person who went to ER last week said (200 miles. All good until it wasn’t).

Curious to see the instructions in the video. Props to ewheels for sending out mass email.

Would not trust it either, it’s just a matter of time.

[supercurio]

Are the MOSFETs supposed to degrade with use if stressed with voltage and currents beyond their specs and close to their maximum capacity?

I'm thinking about: 100V MOSFETs, 100.8V wheel (beyond that when braking) passing the stress test then failing 1000km later due to degradation.

In case it's a thing.

Ping @RagingGrandpa 

[Rawnei]

Ok so if that's true it means there's no easy fix and only an updated board revision with beefier components would be a real fix, I'm not sure about the free-spin test being sufficient since which has already been pointed out the cut-outs have happened to people who already had a decent amount of distance riding before the wheel died on them.

Not sure what I think of that, it's gonna be in the back of the mind all the time while riding.

[RagingGrandpa]

52 minutes ago, Paul g said:

simply not the proper mosfets for the job

Yup, and the community rang the alarm bell 6 months ago...

On 7/1/2021 at 5:39 PM, RagingGrandpa said:

Production model: https://ecodrift-ru.translate.goog/2021/07/01/inmotion-v12-pervaya-partiya/?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=ajax,elem

FETs rated for only 100V ?!

Some guinea pig needs to do a cold-weather full-battery braking test before one of these would get my money...

 

45 minutes ago, supercurio said:

Are the MOSFETs supposed to degrade with use if stressed with voltage and currents beyond their specs

Yes. "Damage" is the reason engineers refer to rated maximum specifications. Exceed the ratings, damage likely.

Bear in mind, there is always component variation...
Semiconductors are made by dissolving chemicals in other chemicals, to create conductive pathways. This chemical process cannot be perfectly repeatable, and so part-to-part variation of some amount is expected. The 'average' part likely exceeds the rated limits by a notable amount; but a few parts may end up right at these limits. Which would help explain a 10% failure rate for controllers of identical construction.

For the cutout issue: we should be honest about we know and don't know. The failures are at a low (but significant) occurrence rate; and there has not been a measurable cause of failure identified yet. The cause could be some additional thing, unrelated to the FET spec. But this out-of-spec part is concerning, in any case.

And now Inmo seems to intend to find and reject any low-end FETs after EUC assembly. I don't like that answer, because the worst-case riding condition is not well understood and likely difficult to recreate in a factory setting. Is a revised board so much to ask for??

  

6 minutes ago, The Brahan Seer said:

ISO9001 accreditation

That's good too. But choosing an IC and then not respecting its ratings is a design issue, not a process quality issue...

 

Edited January 4, 2022 by RagingGrandpa

[.....]

Let me cry a little more here...

WHY hasnt this rudimentary no load testing been done before shipping? WHY must there be injuries and accidents, before such simplistic and inadequate testing is performed? I guess we should all be overjoyed that a minimal ammount of testing is decided to be done, AFTER units are sold and shipped..

Yes, an iso9001 type standard would be awesome. Unfortunately, the game plan is to push new wheels out one after another, who has time to adhere to a standard or test a model that will only be the 'newest model' for about 2 months?

The more info we get about this, the less I trust ANY of it. 100v mosfets on a 100.8v wheel? Am i reading that correctly? Sounds good to me. Whip it up, ship it out, let the chips fall where they may. I hope they figure it out, and I hope people excited for having or getting this wheel, don't run into any of these issues. Even more so, I hope to see a company err on the side of parts designed to take WAY MORE LOAD than needed during extreme use. I think they call it 'adequate overhead'. Hopefully each company is watching each other and at LEAST paying attention to where they can and can't use the cheapest solution. Personally, blaming a parts supplier is a bullshit excuse. Proper QC and testing of the sum of the parts, is what we expect. I won't quit crying until I see it become a reality.

Edited January 4, 2022 by ShanesPlanet

[The Brahan Seer]

59 minutes ago, Paul g said:

??? Makes no sense what I read here

What I am saying is there is a way to mitigate the risk of faulty components and wheels and that is by using a company that is accredited with ISO9001. What this means is that all the components have to be tested, the procedures, the installation etc etc with Quality Assurance. The drawback to this is to buy products with ISO9001 accreditation costs more. 

Example Fluke Multimeters cost a lot more than other brands. But they have ISO9001 accreditation. Which should mean they have very robust systems in place to ensure their products are good. 

[.....]

14 minutes ago, The Brahan Seer said:

What I am saying is there is a way to mitigate the risk of faulty components and wheels and that is by using a company that is accredited with ISO9001. What this means is that all the components have to be tested, the procedures, the installation etc etc with Quality Assurance. The drawback to this is to buy products with ISO9001 accreditation costs more. 

Example Fluke Multimeters cost a lot more than other brands. But they have ISO9001 accreditation. Which should mean they have very robust systems in place to ensure their products are good. 

yup. And my Fluke meter is now 23 yrs old, has been run over by countless cars and is disintegrating to dust, finally. Guess what? It still works and is reliable. I experience the same thing with a lot of my premium music gear, cars and motorcycles. They were made propery DECADES ago and still operate as they should. Quality control is a REAL thing and has REAL consequences. Of coure it can also be really expensive. My cheap made in China gear is crap from day one and barely makes a decent doorstop after. I've learned to quit buying doorstops finally.  Extensive QC checks wont make issues go away entirely, but numbers prove that it isnt a wasted exercise.

So the question is... how much is your front teeth and ability to eat with a fork, worth to you? Now that we are beginning to know better, is it safe to say we will be more willing to pay more? I'd bet those who have spent days in a hospital room or have ruined expensive gear, would likely pay a 20% premium to see drastically reduced failure rated and quality parts. Im not just picking on Inmotion here, this is something ALL vehicle manufactures should be considering. If we're gna be product testers, shouldnt the fist batch be free and come with a waiver of liability?

Edited January 4, 2022 by ShanesPlanet

[wheel-life]

6 minutes ago, ShanesPlanet said:

yup. And my Fluke meter is now 25 yrs old, has been run over by countelss cars and is disintegrating to dust. Guess what? It still works and is reliable. I experience the same thing with a lot of my premium music gear. It was made propery DECADES ago and still operates as it should. Quality control is a REAL thing and has REAL consequences. Of coure it can also be really expensive. So the question is... how much is your front teeth and ability to eat with a fork, worth to you? Now that we are beginning to know better, is it safe to say we will be more willing to pay more? I'd bet those who have spent days in the hospital room or have ruined expensive gear, would likely pay a 20% premium to see drastically reduced failure rated and quality parts.

Agreed, I would pay A LOT of money to avoid a drastic quality of life changing accident that was caused by cheap components. Obviously plenty else can go wrong while riding but at least give us good quality components that don't randomly die.

[Waulnut]

Would this mean Inmotion does not do independent test on controllers to make sure they function over certain limits? 

Besides having guinea pig testers, a manufacturer should have some equipment to test parts for failures. 

I'd only feel safer with a revamped board with bigger MOSFETs if that's the case. 

[.....]

Just now, Waulnut said:

Would this mean Inmotion does not do independent test on controllers to make sure they function over certain limits?

Thanks that was the comic relief I needed.

[The Brahan Seer]

11 minutes ago, Waulnut said:

Would this mean Inmotion does not do independent test on controllers to make sure they function over certain limits? 

Besides having guinea pig testers, a manufacturer should have some equipment to test parts for failures. 

I'd only feel safer with a revamped board with bigger MOSFETs if that's the case. 

At the moment we have to trust that they do. But each time any manufacturer releases a wheel with a fault it doesn't fill me with confidence. Because this is still a new industry with lots to learn from trial and error and experience means we need to take on some of the 'risk' too. As long as we are aware of this then its OK. But as more and more people get into EUC's and it becomes more mainstream perhaps many aren't aware at all. In which case this is a problem.

Edited January 4, 2022 by The Brahan Seer

[RagingGrandpa]

34 minutes ago, Waulnut said:

independent test on controllers to make sure they function over certain limits

They do test the controllers, of that you can be sure. Assembling a PCB can go wrong in a myriad of ways, and does. End-of-assembly-line testing, usually through connecting the controller to a electrical test machine, is the norm in industry now and really the only way to produce electronics in large quantities, without losing all your money on returns and warranty. We have business competition to thank for this.

But it's misleading to ask for more testing here... you can't "test" your way into making an out-of-spec part become in-spec. It simply isn't, by design.

34 minutes ago, Waulnut said:

a manufacturer should have some equipment to test parts for failures

Assuming for a moment that the "FET spec" is root-cause: in order to have discovered it through testing, they would have needed to validate more than 20 controllers at or above the worst-case condition, which is some unknown combination of voltage, temperature, speed, load, and time. Clearly, that wasn't achieved.
But it's not the practical way. Instead, leave margin in the design!

 

Edited January 4, 2022 by RagingGrandpa

[Chriull]

9 minutes ago, The Brahan Seer said:

At the moment we have to trust that they do. But each time any manufacturer releases a wheel with a fault it doesn't fill me with confidence. Because this is still a new industry with lots to learn from trial and error and experience means we need to take on some of the 'risk' too. As long as we are aware of this then its OK. But as more and more people get into EUC's and it becomes more mainstream perhaps many aren't aware at all. In which case this is a problem.

Like last year. And the year before. And the one before...

Seems to be the story with EUC. Actually Inmotion drew a short straw - let's see who's next ;(