INMOTION V12

[supercurio]

First, apologies for raining on everyone's parade.
It's no doubt great to have a functioning test which allows to remove clearly defective wheels from the pool.

However

And that's a big caveat.

2 out of 15 cutouts reported (compiled) so far left the wheel unharmed, with MOSFETs intact.
It means that: if the MOSFETs die like they do in the spin/yank/reverse test, a cutout happens - no doubt
But it also means that cut-out can happen due to one or several separate root causes.

No test covers that yet, and this/these separate problem(s) has/have not been acknowledged or addressed  by Inmotion yet.

My guess here:

• yes a V12 that passes a perfectly executed stress test has significantly lower risk of cutting out due to the blown MOSFET root cause on the short run
• one or several issues, which lead to 13% of these cutouts still remain

Do you think this sounds correct?

[Paradox]

from @Afeez Kay

[spitfire1337]

28 minutes ago, supercurio said:

First, apologies for raining on everyone's parade.
It's no doubt great to have a functioning test which allows to remove clearly defective wheels from the pool.

However

And that's a big caveat.

2 out of 15 cutouts reported (compiled) so far left the wheel unharmed, with MOSFETs intact.
It means that: if the MOSFETs die like they do in the spin/yank/reverse test, a cutout happens - no doubt
But it also means that cut-out can happen due to one or several separate root causes.

No test covers that yet, and this/these separate problem(s) has/have not been acknowledged or addressed  by Inmotion yet.

My guess here:

• yes a V12 that passes a perfectly executed stress test has significantly lower risk of cutting out due to the blown MOSFET root cause on the short run
• one or several issues, which lead to 13% of these cutouts still remain

Do you think this sounds correct?

One of those tests where the mosfets where intact ended up failing on the spin test

[ESB]

Wheel doesn't want to free spin anymore. Gives this error. It can still self balance and seems to ride okay (just tested going ~5mph indoors despite it erroring at < 5mph when free spinning)

 

Anyone else??

 

[supercurio]

1 hour ago, spitfire1337 said:

One of those tests where the mosfets where intact ended up failing on the spin test

Very good point.

So hopefully 100% of the wheel which cutoff also fail the test.

But what explains the cutoff for these then, if it's not the MOSFET failure?

Inability to drive the motor correctly by using MOSFET beyond its spec, the same/same which will blow on the stress test?

Ping @RagingGrandpa again 😉

[ESB]

14 minutes ago, ESB said:

Wheel doesn't want to free spin anymore. Gives this error. It can still self balance and seems to ride okay (just tested going ~5mph indoors despite it erroring at < 5mph when free spinning)

 

Anyone else??

 

This was 'fixed' or worked-around by disabling "load detection" in the app. But I was able to get it to free spin earlier today without turning that off (and I don't think instructions mentioned that setting). So still a bit confused but /shrug i guess.

[Steve Evans]

4 minutes ago, ESB said:

This was 'fixed' or worked-around by disabling "load detection" in the app. But I was able to get it to free spin earlier today without turning that off (and I don't think instructions mentioned that setting). So still a bit confused but /shrug i guess.

i would expect you need 'load detection' option disabled. cant imagine how it worked with that enabled with your earlier test.

hope this helps,

steve

 

[yoos]

Piece of information from ecodrift news channel: the timeline for replacement controllers [just for the confirmed cases under warranty] is unknown but they hope it will be before chinese new year. They also mention they have been telling inmotion for 1.5 years that their mosfets are bad judging by V11 experience. I guess those fets are significantly less likely to burn on a V11 due to 84V<100.8V, but they still do burn sometimes (and this is unacceptable). In my opinion, 100V rated fets is cutting it too close even on a 84V EUC. The more I learn, the less confident I feel about "upgrading" from my 16S.

[Rawnei]

Curious, what are the ratings of mosfet used in 100V wheels from other manufactures? What does Begode or LeaperKim use?

[Wolverine]

Unfortunately, replacement of defective 100V power MOSFETs with completely new equivalents or brand-new board with these same power MOSFETs does not solve the problem just as I have written before. This is a huge design oversight from Inmotion. Even if nothing happens according to the tests performed by the drivers, the damage to poorly chosen power MOSFETs will occur over time and at some point, it will fail. A 100.8V system requires a significantly higher voltage power MOSFETs than they are using now (100V).

I totally agree with @yoos, 100V rated power MOSFETs are cutting too close even on an 84V euc (Inmotion V11). But V11 is left with a margin that V12 doesn't have. If you have to choose between buying V11 or V12, then obvious choice is V11.

@Rawnei If I’m not wrong, the power MOSFETs used in the Veteran Sherman are rated for 125V.

[yoos]

2 minutes ago, Wolverine said:

Unfortunately, replacement of defective 100V power MOSFETs with completely new equivalents or brand-new board with these same power MOSFETs does not solve the problem just as I have written before

I hope the MOSFETs in the V12HT are higher rated ("more durable MOS" was one of their marketing points for the V12HT even before this mosfet scandal) and will eventually replace the old 100V mosfets even in new batches of older models, first of all the V11.

[Paradox]

So now we have to wonder if a wheel is designed with proper safety margins.   Kingsong and Begode have 116v wheels coming out this year.  Do we need to worry that they remembered to up there MOSFETs and other components?  Early adopters suffer again,  roll the dice.  Batch one buyers= extreme risk takers.

[tombo]

So my v12 is on a truck right now from the distributor scheduled for delivery early next week. Would it be prudent to refuse delivery and get a refund or are we looking at a very small percentage of affected units?

[Paradox]

Just now, tombo said:

So my v12 is on a truck right now from the distributor scheduled for delivery early next week. Would it be prudent to refuse delivery and get a refund or are we looking at a very small percentage of affected units?

Call your reseller.  Ask him if you can refuse shipment.  It may or may not create more problems just refusing the delivery.

[Paul A]

Are these the 100V and 125V Mosfets used in EUC's?

Price difference of about $0.50 between 100V and 125V.

 

https://www.alibaba.com/product-detail/IRFB4410ZPBF-TO-220-n-P-Channel_1600381554965.html?spm=a2700.galleryofferlist.normal_offer.d_title.7a152f14B8jJ4H&s=p

IRFB4410ZPBF TO-220 n / P Channel MOSFET semiconductor transistor integrated circuit 100V 97A IRFB4410ZPBF

1 - 99 Pieces  $0.29          100 - 999 Pieces  $0.25         >=1000 Pieces $0.21

 

https://www.alibaba.com/product-detail/HY3912W-HY3912-TO-247-190A-125V_1600294106614.html?spm=a2700.galleryofferlist.normal_offer.d_title.7a152f14B8jJ4H

HY3912W HY3912 TO-247 190A 125V N-Channel Enhancement Mode MOSFET

$0.67 - $0.70/ Piece

 

[skautas2003]

28 minutes ago, Wolverine said:

Unfortunately, replacement of defective 100V power MOSFETs with completely new equivalents or brand-new board with these same power MOSFETs does not solve the problem just as I have written before. This is a huge design oversight from Inmotion. Even if nothing happens according to the tests performed by the drivers, the damage to poorly chosen power MOSFETs will occur over time and at some point, it will fail. A 100.8V system requires a significantly higher voltage power MOSFETs than they are using now (100V).

I totally agree with @yoos, 100V rated power MOSFETs are cutting too close even on an 84V euc (Inmotion V11). But V11 is left with a margin that V12 doesn't have. If you have to choose between buying V11 or V12, then obvious choice is V11.

@Rawnei If I’m not wrong, the power MOSFETs used in the Veteran Sherman are rated for 125V.

INMO and if you read inmotion statement not voltage of mosfets problem all of them has some safety margin ir reality . Problem is bad quality mosfets. And during test not all mosfets burn but just defective mosfets

[Chriull]

7 minutes ago, tombo said:

 at a very small percentage of affected units?

It was rumored/forwarded from a distributors statistic that about ?10%/30%? (Don't remember which;) were affected. But it's a new wheel and failure rate could rise once the riders gets accustomized to the wheel and burden it higher?

The wheel lift test statistics posted here somewhere show a result somewhere in this range (definitely lower than 30%), but sample size is still very low till now (some 15-20 tests?)

4 minutes ago, Paradox said:

Call your reseller.  Ask him if you can refuse shipment.  It may or may not create more problems just refusing the delivery.

If refusal/return is allowed the important point is who bears the cost (and risk) of the return transport!

[Chriull]

35 minutes ago, skautas2003 said:

INMO and if you read inmotion statement not voltage of mosfets problem all of them has some safety margin ir reality .

Could be true fir breakdown voltage, but still the Safe Operation Area specified fir this mosfet is "borderline" https://forum.electricunicycle.org/topic/26197-v12-cutout-tracking/?do=findComment&comment=399793

 

35 minutes ago, skautas2003 said:

Problem is bad quality mosfets. And during test not all mosfets burn but just defective mosfets

Bad quality infineon mosfets should be a problem of buying from dubious sources, vad manufacturing or bad design choices.

If it comes to (dis)trust specified mosfet characteristics i'd choose infineon vs inmotion...

Edited January 6, 2022 by Chriull

[jimjam.nyc]

Most of the parts in these wheels have a shelf life and will degrade over time.. So pointing out the fets will degrade over time is a bit over reaching. Its always a game of what will fail first.. Battery, fets, control board.. fuses.. etc..

Weather or not they will degrade much quicker than expected is something to think about. 

Personally i have 1000 miles on my v12 1st batch. At this point i trust my wheel.. I have gone through the exercise of gaining trust in every single wheel i have owned.. Doesnt matter who manufactured it.. These wheels have NO redundancy and parts sometimes fail. The truth is, every manufacturer of these wheels faces the same stuff.. Quality of parts can be an issue, and a percentage of that will not be known until the wheels are in the hands of customers. 

I guess my point is, you shouldnt just trust any of these wheels no matter how much said company does for safety.. Things can go wrong, and I feel like we all know that going in to this hobby.. especially on this forum. That is not a pass for wheels cutting out for no reason, Inmotion needs to figure it out and solve it.. But it really is just a reality of getting into this hobby. Especially if you are buying a new product run.

Hopefully inmotion can get this sorted and going fwd source better more reliable parts. 

If you want the latest and greatest... Gear UP! Maybe Fireproof your house!

All kidding aside, i hope anyone who got hurt heals up.. and hopefully with this testing at least we can avoid injuries going fwd until the problem is fully solved. 

[Rawnei]

1 hour ago, Paradox said:

So now we have to wonder if a wheel is designed with proper safety margins.   Kingsong and Begode have 116v wheels coming out this year.  Do we need to worry that they remembered to up there MOSFETs and other components?  Early adopters suffer again,  roll the dice.  Batch one buyers= extreme risk takers.

The way they are testing the S20 with all that off-road I'm sure they would find any weakpoints.

[nosamplesplease]

3 hours ago, jimjam.nyc said:

Most of the parts in these wheels have a shelf life and will degrade over time.. So pointing out the fets will degrade over time is a bit over reaching. Its always a game of what will fail first.. Battery, fets, control board.. fuses.. etc..

...

It's really great you have confidence. That does not mean the engineering was done correctly.

Every electronic component from a coil to a MOSFET has a tolerance. Depending on what part and what manufacturer, often this variance is 10% , (IIRC, my EE classes were a few years back) better quality ones have lower.

Hypothetically (I don't know this model's tolerance) with these 100v MOSFETS, assuming 10%, that's either a real world functionality range of 90v - 110v. There's are myriad other factors for choosing BUT putting 100v MOSFETS in a system where AT LEAST 100.8v is present by design sounds like Inmotion was gambling on a high percentage of them exceeding their voltage rating. Again IN THIS CIRCUIT with their calculated usage may never experience these conditions. But with regenerative braking, these might be difficult to predict over a a broad range of rider weights, braking duration, and temperature.

The test that Inmotion is having V12 owners do would certainly sort out which MOSFETs are outside of tolerances. It does not tell me if the engineering is flawed. Just that the high percentage of failure IMO, lends towards choices made by management. Which choices, which department, whose management? Who knows?

It does not suggest to me a supplier problem.

This "could be" due to the "worldwide chip shortage" with competition for >100v in the same package size driving the prices up when Inmotion was buying. I do not know their design process but choosing a lower rated components doesn't sound safe. Again their design, expected operating range, and programming may account for this.

Without an independent source doing a root cause analysis on a large sampling of failed wheels you HAVE to trust Inmotion. To me, that's concerning. 3rd party analysis would go miles to make some of us feel confident. 

IME in the service side of the oil and gas industry, when an audit is done internally, fudging to cover up a mistake is common.

Inmotion may be more honest than the multi-billion dollar American companies that I worked for.

There are limitless reasons any part is chosen. We only know what they choose to tell us.

For me as soon as the V12 High Torque comes available for preorder I still probably will put my money down. Only because they specified better MOSFETs. You can be damn sure I'll be watching this forum to see how this problem is resolved over that time. I'm would really like to see pictures of the boards Inmotion decides to use on their next revision. 

Time will tell 

 

Edited January 6, 2022 by nosamplesplease

[ESB]

What I'm having a hard time reconciling (no EE background!) is that if 100v mosfets are inherently so risky on a 100.8v wheel (out of spec ones aside) then shouldn't chooch have hit issues riding his v12 as hard as he does? He's not a heavy rider but he accelerates and brakes super aggressively and goes off road with it some. We all know chooch I'm sure - and chooch style of riding on the v12 for several months I think would have revealed a problem with the design. No?

Edited January 6, 2022 by ESB

[jimjam.nyc]

41 minutes ago, nosamplesplease said:

It's really great you have confidence. That does not mean the engineering was done correctly.

 

 

I wouldnt say i have confidence.. I just can't remember any wheel that i have purchased that i didnt read through pages on this forum of cutouts, broken gear, fires, and everyone chiming in on how the engineers got it wrong, etc. 

I think what i was trying to say, is i have as much confidence as i had in any other wheel i purchased.. Which is not based on my trust in the company itself, but trust in the single product i have in my hand and the parameters i use it in. 

Every single wheel i have purchased, it took some time for me to feel confident in its ability not to throw me off. Especially after coming here and reading through this forum.

As far as waiting for the eventual failure of a part of this or any wheel.. I am sure it will come at some point. Its inevitable unless we start replacing wheels every so many miles, etc.. So far i don't see too many people suggesting guidance on how long these should last before they fall apart. Anything with moving parts requires inevitable maintenance. Good Chance is I will be riding it when and if it fails.. That is just a known risk we all take.

Edited January 6, 2022 by jimjam.nyc

[RagingGrandpa]

4 hours ago, Rawnei said:

what are the ratings of mosfet used in 100V wheels from other manufactures?

Gotway/Begode/Leaperkim is 125V...
 

3 hours ago, Paradox said:

Kingsong and Begode have 116v wheels coming out this year.  Do we need to worry?

S20 is 126V!
New hardware is certainly needed, without it there would be no functional prototypes.
We'll hopefully learn about its specifics in the usual way, from an internet teardown report of one of the first S20's.
 

1 hour ago, Rawnei said:

Early adopters suffer again,  roll the dice.  Batch one buyers= extreme risk takers.

Lots of 2nd-batch V12's failing... but I agree with your idea. Safer to wait until an EUC has been in real-world consumption for 6 months to demonstrate reliability, before buying; less fun tho
 

3 hours ago, Paul A said:

Are these the 100V and 125V Mosfets used in EUC's?

IRFB4410ZPBF (100V) 1000 Pieces $0.21

HY3912W (125V) $0.67 - $0.70/ Piece

Not IR.

V12 100V FET is Infineon 23N10N5. About $4 each!

And the Gotway 125V FET is Hooyi 5012. $0.90 - $1.90/ Piece.

  

16 minutes ago, ESB said:

shouldn't chooch have hit issues riding his v12 as hard as he does?

If you get lucky with component variation, it's perfectly likely that some boards will survive...
(we still don't know if the failure rate is due to variation alone; or if truly defective parts that don't meet the minimum spec made it into production)

 

Edited January 6, 2022 by RagingGrandpa

[nosamplesplease]

2 hours ago, jimjam.nyc said:

I wouldnt say i have confidence.. I just can't remember any wheel that i have purchased that i didnt read through pages on this forum of cutouts, broken gear, fires, and everyone chiming in on how the engineers got it wrong, etc. 

I think what i was trying to say, is i have as much confidence as i had in any other wheel i purchased.. Which is not based on my trust in the company itself, but trust in the single product i have in my hand and the parameters i use it in. 

Every single wheel i have purchased, it took some time for me to feel confident in its ability not to throw me off. Especially after coming here and reading through this forum

This actually expresses very well what I am taking about. A large number of individual units very likely have MOSFETS that do not fail. It's a matter of playing the odds, regardless of good engineering or bad.

This is often referred to as infant mortality in some industries. It is expected a small percentage of completed electronic devices will fail in the first hours of actual usage. This often why warrantees and store return policies exist in the US. It's not the store's or final assembler's fault. 

You know what else suffers a failure in the first few hours of usage by customers? Questionable engineering followed up by inadequate testing.

I believe this industry has a reputation.

The data seems to suggest it could be an infant mortality failure. Again only Inmotion knows.

Your "trust period" is THE SMARTEST thing any consumer can do to protect themselves. This seems to be a priority for Kingsong who locks the speed for the first few miles. 

 

2 hours ago, ESB said:

What I'm having a hard time reconciling (no EE background!) is that if 100v mosfets are inherently so risky on a 100.8v wheel (out of spec ones aside) then shouldn't chooch have hit issues riding his v12 as hard as he does? He's not a heavy rider but he accelerates and brakes super aggressively and goes off road with it some. We all know chooch I'm sure - and chooch style of riding on the v12 for several months I think would have revealed a problem with the design. No?

Yes, and no. Most of the "bigger" YouTubers you've seen were/are using units made before the initial large volume production batch. Less units to QC, more care and feedback from users, less expected perfection and most importantly: this is when the most attention for product happens, prior to release.

But you have to consider production engineering changes just due to refinement. This is normal. Beta products have features on that get removed, MANY second revisions are actually "worse" than initial designs. Revisions are most often made for cost reasons. Sometimes for safety and product improvement. Product improvement might not be from the customers point of view. Improved is generally a marketing term.

There's is a long history of products that were pre-ordered that did not arrive to end users as promised.

You have read articles about Kickstarter?

It is in ANY company's best interest to pay close attention to QC when the world is watching.

Inmotion seems to communicate these things better than most companies in this niche, they seem to have made good choices in the past regarding improved reliability.

Seems.

Just sayin' find trust where YOU can.

 

Stay safe.

Edited January 6, 2022 by nosamplesplease
Damn autocorrect. Sigh....