Inmotion V12 HT board: "More durable MOS"

[Tawpie]

1 hour ago, supercurio said:

if components has been indeed conservatively selected, they would not pop on 22% of wheel across all production batches.

Assuming they were conservatively selected (benefit of the doubt), some of the other whisperings about how long batch 2 took to ship and changing suppliers and parts shortages and improper storage makes me think that somebody in the supply chain bought parts on what I used to call the grey market. They didn't come from the 'normal' supplier, but the army of buyers that was scouring the planet for parts managed to find some somewhere and they entered the manufacturing process. The issue with the grey market is that parts you find there generally don't have as good a pedigree—they might have been part of a legitimate over-buy and are fine, but they might be parts that do not actually meet all of their original specs (seconds). The might have been rejects, or salvaged from other assemblies. You just don't know for sure. The grey market is a risky route to take—sometime it works out but it's called 'grey' for a reason. If you do allow use of grey market parts, your design margin has to be able to absorb the extra tolerance consumed by including parts that may not meet the original design specifications.

I'm not saying these are gray market parts or that they do or don't meet Infineon's OEM specifications, I'm just pointing out that it is easy for a final assembly house to get burned by their supply chain when parts are in short supply.

 

A war story about supply chain management:

I once designed in assemblies supplied by the professional products arm of a huge consumer electronics firm and some of the crystals (filter components) they delivered to me didn't meet specification. It was kind of a big deal, I didn't discover the problem until we were well into qualification testing. When the huge consumer electronics firm discovered that their crystal manufacturer had delivered 'bad' parts, their corrective action was literally "they are no longer a supplier to our corporation". I jokingly asked "To all of xxx, even your consumer products? What did you do? Cancel 2 boat loads of crystals?" The answer was simple: "Correct, they no longer supply to any division and yes, it was several container ship loads of lost business to the supplier". To me, this was a bit of a heavy handed response (you can fit a lot of crystals in a container ship) but perhaps the supplier had a history of this behavior I don't know. Nevertheless making such a decision had to cause huge production problems for the consumer side of the business and was enormously costly to xxx as a whole. I on the other hand, gained more faith in xxx corp's willingness to aggressively manage their supply chain, and consequently some trust was restored that I could expect to receive 'good' assemblies from them in the future. It all turned out fine, except for me having to explain the schedule slide in my project.

 

Whatever the root cause, the corrective action, where the blame falls—I hope and do believe that IM will emerge from this able to deliver safer products overall, and that the industry as a whole will take notice and pay attention to even the small details. This incident (and the battery fires) should make us all safer in the end. It better... I'd hate to have to give up riding, it's too much fun.

[UniVehje]

3 minutes ago, supercurio said:

We still don't know, but a trusted distributor told me 30% of his summer batch V10F failed also. And aren't they supposed to use the same Infineon IPP023N10N5?

It was not discussed anywhere but Inmotion has this information, although they left the distributor in the dark without answer.
This probably contributes to their conclusion that V12 failures are due to defective parts in the in the FET supply instead of a design issue.

V10 series has been out for a long time and I used to own one and now my spouse has another unit. The later unit is definitely better than my old first batch unit. It used to be very reliable (after first batch) and one of the most sold wheels. But I’ve lately noticed a lot more issues with them. Just by following the discussion here. It is certainly strange that an old design would start to fail more lately. Something strange is going on. 30 % failure rate is too much.

[Paul A]

Great explanations on what might have happened and sourcing from grey markets.

____________

 

Heavy handed response maybe deliberate.

To make an example, to send a message to other suppliers, to establish a reputation.

Costly in the immediate term.

Might be an investment in discouraging future cheats and save on future costs.

[supercurio]

Thanks @Tawpie for the insightful story.

 

40 minutes ago, UniVehje said:

V10 series has been out for a long time and I used to own one and now my spouse has another unit. The later unit is definitely better than my old first batch unit. It used to be very reliable (after first batch) and one of the most sold wheels. But I’ve lately noticed a lot more issues with them. Just by following the discussion here. It is certainly strange that an old design would start to fail more lately. Something strange is going on. 30 % failure rate is too much.

Indeed an odd coincidence that new V10F units start to fail the same way on a proven design (burned MOSFET)
The trusted distributor mentioned that he unofficially learned that Inmotion was looking for a new mainboards supplier last spring. Reasons unknown.

Both supply chain quality issues and unsuitable MOSFET selection can be compounding problems.

Let's see if we can find what the V12 HT "More durable MOS" are.

Edited January 7, 2022 by supercurio

[Paul g]

12 hours ago, mrelwood said:

 But I have a habit of trusting working professionals in any decision making or especially assessing the validity of any circuit design on any product. I have repaired guitar amps and other musical gear as a side job for over a decade. Yet I will never even give an assumption of being able to assess whether a mosfet choice in an EUC or even in a guitar amp is the right one or not. I don’t quite see how you can.

     Sorry, @mrelwood, would you say the same thing about the several Boeing 737 that crashed and killed hundreds of people until now? Were the working Boeing professionals, who had many years of experience, unquestionable ? 
 

     And if all reduces to blind trust in “professionals” than the issue that plaques the V12 reduces to two possibilities:

- 1) The Infineon “professionals” made faulty mosfets and sold them to clients

- 2) The InMotion “professionals” designed faulty boards and sold them in V12s

     In both cases “ professionals” were to blame, that is why this blind trust in “professionals” makes no sense to me.

     Second point I want to make, is that most of us have usually good trust in engineers. It is the “bead counters” decisions we fear, and these ones lead to disastrous outcomes.

      Third, we did not come with this idea of underrated mosfets from thin air. Though I’m not an electrical engineer, on specialty forums and sites, it is repeated over and over again that when choosing a mosfet for a motor you have to have the Drain-Source Breakdown Voltage -Vds higher then the maximum working Voltage available. 

 

 

[RagingGrandpa]

Putting aside breakdown voltage just for a moment, I wanted to consider:
Do these common Inmo and Gotway FETs meet our needs for output current?

... and if their datasheets are to be believed, the difference is huge.

(My expectation of 100A @ 40us is based on my observations of stall currents just above 200A in Gotway EUC's, with that 200A supplied by two FETs in parallel. This is overlean at zero speed, truly the highest output current that can occur during normal riding.)

It seems the TO220 FET simply should not be used for such pulses (30A max), whereas the TO247 has plenty of headroom here (600A max).

Inmo could set the current limits (using firmware) at whatever value they please, so perhaps they simply don't permit such high motor current. But I don't think they can limit it to only 1/3rd of the Gotway value, because it would mean very weak torque for a same-size same-speed motor.
My personal reaction to riding V11 (which uses the same TO220 FETs as V12) was that it had low-speed torque very similar to the 84V MSX; definitely not "much weaker." V11 is an 84V system, so the limit does become higher than the 100V numbers shown below, nearing our 100A target. And V11 controllers have been more reliable- so this all fits.

From my short look at other similar FETs (www.mouser.com), it seems like a real challenge to find a substantially higher-current alternative in the TO220 size. And upsizing the package (e.g.: TO247, or SMT parts) means an enclosure/heatsink redesign.
Back to the drawing board, gents.

Edited January 8, 2022 by RagingGrandpa
grammar

[supercurio]

Okay I'll try:

 

Enhance!

 

AI upscaling couldn't restore the text on V12 HT board MOSFETs 😂

[supercurio]

36 minutes ago, RagingGrandpa said:

It seems the TO220 FET simply should not be used for such pulses (30A max), whereas the TO247 has plenty of headroom here (600A max).

Edit: got it.

100V drain source voltage, 40 micro seconds ends up here. However isn't it for continuous operation instead of pulse?

Edited January 7, 2022 by supercurio
Understood by taking a better look at your annotated graph thanks!

[mhpr262]

2 hours ago, UniVehje said:

V10 series has been out for a long time and I used to own one and now my spouse has another unit. The later unit is definitely better than my old first batch unit. It used to be very reliable (after first batch) and one of the most sold wheels. But I’ve lately noticed a lot more issues with them. Just by following the discussion here. It is certainly strange that an old design would start to fail more lately. Something strange is going on. 30 % failure rate is too much.

Its not just the electronics... my third V10F arrived with sagging pedals, straight from the factory, apparently never ridden. Left pedal was exactly horizontal, right pedal sagged downwards ever so slightly. It was barely noticeable, but holy crap how unpleasant it was to ride. I was able to more or less fix it with shims, but they have to be rechecked and adjusted every couple 100kms because those damn shims just wont stay put. My new V11 made me remember how great it is to have pedals that are angled correctly and that dont need any maintenance, LOL

[enaon]

 

23 hours ago, Paul g said:

They even blame Infineon for faulty or unreliable products. That is even more ridiculous that said handling.

 

11 hours ago, Planemo said:

Nothing wrong with Infinion FET's (assuming genuine and used within their specs). They are reputable, reliable and I have used 1000's of them with repeatable results.

this is what I didn't like too, blaming Infineon is funny. I do not understand how they are supposed to check the fets they are going to send to the resellers. My bet is that they had a feeling they were using clones. 
 

[RagingGrandpa]

5 hours ago, supercurio said:

isn't it for continuous operation instead of pulse?

Ah. Talking about stall (zero wheelspeed), the motor will hold a relatively continuous current, but the controller achieves it by pulsing on and off the battery using the FET array. FETs turn on and increase the current through the inductive load of the motor, and then turn off and allow the current to recirculate naturally for a moment (through body diodes or other FETs), then on again, then off again, with the on/off time ratio (aka "duty cycle %") determining the level of current that is achieved.

We cannot go to 100% "continuous" duty with the motor stalled- the coil resistance is about 0.1 Ohm and so 100V would produce 1000A. Everything would melt

 

Edited January 8, 2022 by RagingGrandpa

[mrelwood]

 

7 hours ago, Paul g said:

     And if all reduces to blind trust in “professionals” than the issue that plaques the V12 reduces to two possibilities:

I never even hinted towards “blind trust”. Just like durability (), my trust for professionals as well is a scale. What I said is that I tend to trust professionals more than unknown forum members who are agitated, angry and some even furious. Just like I’m taller than my brother, but I’m still not the tallest person in the world.

 

7 hours ago, Paul g said:

     In both cases “ professionals” were to blame, that is why this blind trust in “professionals” makes no sense to me.

I agree, and it would make no sense to me either!

 

7 hours ago, RagingGrandpa said:

It seems the TO220 FET simply should not be used for such pulses (30A max), whereas the TO247 has plenty of headroom here (600A max).

I understand that you are taking only about the pulse current, and that there are several different mosfets specs related to current. When I was deciding on whether to buy a new board for my MSX a few years ago, I also checked if I can source mosfets I could use as a possible replacement instead. Trying to find a TO247 mosfet locally with comparable specs to the HY5012 turned out very difficult, mostly because the continuous current rating was hard to match.

Edited January 8, 2022 by mrelwood

[RagingGrandpa]

42 minutes ago, mrelwood said:

Trying to find a TO247 mosfet locally with comparable specs to the HY5012 turned out very difficult

Luckily new parts keep being designed... this one looks excellent to me, should have lower dissipation than HY5012 and 150V tolerant.

44 minutes ago, mrelwood said:

the continuous current rating was hard to match

Ah, but that's what I'm getting at: in an EUC, we never have current above about 100A per FET, and the most important behaviors are pulsed. So the dreamy >300A specsheet numbers aren't necessary.

p.s. HY5012 is package limited to 90A continuous... sneaky authors hid it on page 4, hoping you'd stop reading at page 1

[evX_Mick]

On 1/7/2022 at 2:18 AM, mrelwood said:

First, nobody can disagree that the situation is really bad, and would harm Inmotion’s reputation no matter what the reason for the failures were, and no matter how they would handle the situation. Thus it’s no surprise that it will create emotions, in some people more than in others. And I feel terribly bad for every single one who have hurt themselves because of this, and even for those who are forced to refrain from riding because of this.

 But I have a habit of trusting working professionals in any decision making or especially assessing the validity of any circuit design on any product. I have repaired guitar amps and other musical gear as a side job for over a decade. Yet I will never even give an assumption of being able to assess whether a mosfet choice in an EUC or even in a guitar amp is the right one or not. I don’t quite see how you can.

You being so sure about a large safety oriented EUC/PLEV manufacturer making a mistake so obvious that random end users and discussion forum members are able to reliably point it out doesn’t sit right with me. We’re talking about a manufacturer who has designed and sold possibly over 100000 vehicles with the same basic motor controller principle. That’s a hundred times more than I have posted on this forum. If only they had asked the right forum member to do the final check on the circuits, right?

That’s a bit thick isn’t it? You don’t have any of the countless crucial pieces of information required to know anything about the manufacturer’s perspective. “Some” reading between the lines is clearly an understatement.

 

I know you’re much smarter than this so it’s probably a language barrier. Basic lesson in logic: Me being taller than my brother doesn’t mean or imply in any way that my brother weren’t very tall as well. Being durable or not is not a black and white distinction. Some things can always be more durable than others, and the secondmost durable thing is usually still pretty darn durable.

 

That is indeed quite fitting…

I have noticed that you have been one of the most vocal participant in the recent V12 discussions. It’s clear that you take the issue strongly. What can happen though is that seeing the forest from the trees becomes increasingly more difficult as you keep strengthening your beliefs. That still doesn’t make it a fact though.

 

If I’d ask you how you know that, you would probably point me to evidence of wheels that have failed. But if I’d ask the same from an electronics engineer specializing in BLDC motor controllers, he’d be able to explain in detail how the system works and why exactly the design is faulty, included with calculations. Do you understand the difference?

 

 Could you be right in that the company we have learned to trust the most all of a sudden no longer understands the meaning of “voltage”, decide to create a product designed to fail, and they just pull out testing instructions from their hats, hoping that the western fools would yet again be fooled? I guess you could. But it all just sounds so extremely improbable. I’d rather ask someone who knows about the matter. And I don’t mean someone who has read about the cutouts. I mean someone who actually knows the circuit diagram, inside company policies, Inmotion engineer hiring requirements, and so on.

Great rebutle here from mrelwood. I think the original poster is making a lot of assumptions to be passed off as analysis of fact, and leaning heavily on malevolence.  We are once again in ANOTHER Inmotion blunder. But I do not believe it’s due to bad engineering by some team lacking intelligence. Nor is it a conniving marketing team to cover it up.  (That is what this post is getting at.) Guys this isn’t some western pharma company lol. These companies still see themselves are glorified factories. 

 

 

[Paul g]

12 hours ago, mrelwood said:

 

I never even hinted towards “blind trust”. Just like durability (), my trust for professionals as well is a scale. What I said is that I tend to trust professionals more than unknown forum members who are agitated, angry and some even furious. Just like I’m taller than my brother, but I’m still not the tallest person in the world.

 

I agree, and it would make no sense to me either!

 

I understand that you are taking only about the pulse current, and that there are several different mosfets specs related to current. When I was deciding on whether to buy a new board for my MSX a few years ago, I also checked if I can source mosfets I could use as a possible replacement instead. Trying to find a TO247 mosfet locally with comparable specs to the HY5012 turned out very difficult, mostly because the continuous current rating was hard to match.

Makes sense to me. I would take your example if I were one of the people that had an injury from a V12 burned mosfet and would trust InMotion more than be “agitated, angry” and “even furious”. Also the same would be advisable to all family members of Boeing crashes. Why be angry and agitated, when you should trust more the Boeing company, which is a renowned and trusted company, with so many years of experience, and so many undeniably successful and safe products throughout the years.

  To be noted here, it is only the pressure put on governments by protests of “angry”, “furious” and “ agitated” surviving family members of those deceased in Boeing case given as example that led to the investigations which proved Boeing was putting peoples lives in danger for more profit. Until very late in the investigations they were blaming pilots, but when the investigation brought to light undeniable proves they finally conceded to some of the wrongdoings they were blamed of, even then not in totality, they were still trying to hide some of them.

   Now it is true, I’m no expert, but that does not mean I can’t investigate my self the issue, and in a more or less perfect way, with the help of people that know what they’re saying, find possible reasons for V12 problems. 
 

    In any case, it is InMotion that produces the V12 and therefore it is their, and only their, responsibility when it comes to the dangerous failures. If the failures happens because of bad quality mosfets, it is their failure. If it happens because of underrated mosfets, it is again their failure. And the ones that pay the price it is us.

  

[supercurio]

5 hours ago, evX_Mick said:

Great rebutle here from mrelwood. I think the original poster is making a lot of assumptions to be passed off as analysis of fact, and leaning heavily on malevolence.  We are once again in ANOTHER Inmotion blunder. But I do not believe it’s due to bad engineering by some team lacking intelligence. Nor is it a conniving marketing team to cover it up.  (That is what this post is getting at.) Guys this isn’t some western pharma company lol. These companies still see themselves are glorified factories. 

Hey @evX_Mick, can you please use your access to ask what are the board specs & MOSFET used in the V12 HT?
And if there is any other difference.

I would feel much more confident with these "Increased max power by 23%" "More durable MOS" installed in my incoming V12 vs the current ones:

• maybe passing a stress test just fine
• maybe surviving the stress test by only 0.5% and dying on the next ride
• or.. popping on the stress test and I should have just canceled my order instead

Thanks.

 

By the way, at no point I am saying or implying that Inmotion attempted to cover up anything, and I don't believe so.
Only that the V12 HT material might candidly / inadvertently reveal the internal knowledge that the V12 has reliability issues. And there is no argument they have been alerted of these reliability issues by distributors and customers, before it became a thing in the forums.

I do believe it is bad engineering however instead of bad luck. Lack of headroom in the parts used, no effective stress testing before shipping to customer. Hopefully that'll become the new standard for EUCs like it already is for other small EV manufacturers (like NIU)

Edited January 8, 2022 by supercurio

[mrelwood]

In order to keep the discussion relevant to the topic, I’ll skip most of @Paul g’s reply.

On 1/8/2022 at 6:35 PM, Paul g said:

Also the same would be advisable to all family members of Boeing crashes. Why be angry and agitated, when you should trust more the Boeing company, which is a renowned and trusted company, with so many years of experience, and so many undeniably successful and safe products throughout the years.

So, under how many aspects or topics has Boeing lied or been deceitful? And in how many aspects they have told the truth? The ratio is what I’m after. Then the same ratio from the average discussion forum whistle blowers.

 Nothing anyone can ever say will help in any meaningful amount with the sorrow that an event such as the Boeing one lays on countless people’s lives. And there’s of course no question that if it had been created intentionally, the people who have decided to do so or have let it slide, should be charged with manslaughter or similar. 

But even when Boeing does something like that, does that automatically make everything they have ever said a lie? Of course not. Does it make all manufacturers evil and deceitful? Of course not. Does it mean that 911 was an inside job? Of course not. And finally circling back to the topic, does it make manufacturers less trustworthy than angry discussion forum members? Of course not.

 

On 1/8/2022 at 6:35 PM, Paul g said:

  To be noted here, it is only the pressure put on governments by protests of “angry”, “furious” and “ agitated” surviving family members of those deceased in Boeing case given as example that led to the investigations

And investigating is exactly what should be done and is being done with the V12 as well. But it has nothing to do with blaming Inmotion being a deceitful liar who tossed out a decade of hard earned trust and reputation simply to make a quick buck.

If you believe that governments should investigate the matter, don’t rant it out at a web forum. Be active and reach out to politicians with your suspicion. They’re not reading the forum for clues on what they should investigate.

 If you want the manufacturer to do something, spend a minute to think what kind of an approach would make them do what you want. You really think that blaming them for being deceitful and only being after an easy short term sale will even make them read your message all the way through the first time?

 If you want the distributors to do something about it, reach out to them. And again, think what kind of a message would make you want to do something about it.

 But if you don’t want to be taken seriously, don’t want anything to actually happen, and just want to rant it out a bit, go to Reddit or any other discussion forum, and make a post with accusations and present your theories as if they were well known truths. It will serve your intent perfectly.

I’m pretty sure that even the immense amount of “911 was an inside job” posts, tweets, videos and talk shows made any government or other official party do diddly squat to lift a finger towards opening up an actual investigation. Because it’s the wrong approach.

 

 (I’ve learned that amongst Americans, disclaimers are often required: All of the above were examples to make a point, and none of it was directed towards any single forum member. For example @supercurio is actually doing mostly great job in educating himself and others about the matter, is being proactive, and despite a few misses seems being after the facts instead of slander.)

[supercurio]

We know today that the stress tests are not sufficient to identify wheels which will later fail during real-world use.

So far I gathered hints indicating that replacement boards:

• Will have a number of differences
• Should have more durable MOS
• Is likely the V12 HT board

My guess is that future V12 batches will also be equipped with the V12 HT board - as predicted in the original post.

No official quote or statement, please consider this as rumor for now.

Edited January 12, 2022 by supercurio

[Bikemike2777]

4 hours ago, supercurio said:

We know today that the stress tests are not sufficient to identify wheels which will later fail during real-world use.

So far I gathered hints indicating that replacement boards:

• Will have a number of differences
• Should have more durable MOS
• Is likely the V12 HT board

My guess is that future V12 batches will also be equipped with the V12 HT board - as predicted in the original post.

No official quote or statement, please consider this as rumor for now.

Yeah I would love to have the standard v12 not the ht. But with the ht board    it would be nice if inmotion would identify what fets they are running for the new ht so some of us can order a dozen fets and install them ourselves to act as testers and hopefully have a safe first and 2nd gen wheel.   

[supercurio]

12 minutes ago, Bikemike2777 said:

Yeah I would love to have the standard v12 not the ht. But with the ht board    it would be nice if inmotion would identify what fets they are running for the new ht so some of us can order a dozen fets and install them ourselves to act as testers and hopefully have a safe first and 2nd gen wheel.   

It would be interesting out of curiosity, no doubt.

However it appears that the HT board would have more differences than the fets only (TBD)
The firmware would likely require a different tuning as well to take into account these components in the tuning of the motor controller.

Still worth investigating, how much are you willing to crash and get hurt as a tester tho!

[Bikemike2777]

5 minutes ago, supercurio said:

It would be interesting out of curiosity, no doubt.

However it appears that the HT board would have more differences than the fets only (TBD)
The firmware would likely require a different tuning as well to take into account these components in the tuning of the motor controller.

Still worth investigating, how much are you willing to crash and get hurt as a tester tho!

There is other options for testing. Climbing harness attached to rafters of a garage and a gutted out treadmill. Maybe some shock absorbing lanyards to the wheel too   ain't no speed limit on a treadmill if there is no motor hooked up lol if the body and wheel are stationary that should have better end results if the failure occurred.

[Paradox]

2 minutes ago, Bikemike2777 said:

There is other options for testing. Climbing harness attached to rafters of a garage and a gutted out treadmill. Maybe some shock absorbing lanyards to the wheel too   ain't no speed limit on a treadmill if there is no motor hooked up lol if the body and wheel are stationary that should have better end results if the failure occurred.

This would be an interesting video to watch.

Edited January 12, 2022 by Paradox

[supercurio]

6 minutes ago, Bikemike2777 said:

There is other options for testing. Climbing harness attached to rafters of a garage and a gutted out treadmill. Maybe some shock absorbing lanyards to the wheel too   ain't no speed limit on a treadmill if there is no motor hooked up lol if the body and wheel are stationary that should have better end results if the failure occurred.

Beefy treadmill which can take 70 km/h or handle 10kW peak 😋 - actually providing some resistance, or you're thinking by using gravity maybe.

Cool project nonetheless. 

[fbhb]

1 hour ago, RagingGrandpa said:

Kingsong has been on the TO220 size for all of their modern EUC's. Sorry I'm not aware of which model chip it is though.

Even the much maligned S18 actually uses 12 TO-247 MOSFETS despite it's much lower performance threshold, as can be found on King Song's website:

KS-S18-18 inch series-King Song

[Paradox]

7 minutes ago, fbhb said:

Even the much maligned S18 actually uses 12 TO-247 MOSFETS

The S20 seems to be using the same.

 

https://www.kingsongs20.com/spec