V12 Cutout tracking

[Paul g]

1 hour ago, vBlitz said:

The new plates in my arm seems to say otherwise

unless if it's a faulty installation. I don't know yet. I'm still at the hospital.

 

Thanks for the kind words @Paul g

Your welcome! 

We must find out what causes these crashes. I don’t really know what kind of data InMotion has from its right-before-cutout logs, but I have some doubts about the amount of info these logs provides to them. These wheels that had cutouts must be opened and then thoroughly checked by a proper EE which should be provided by the seller. @vBlitz I hope you will manage to specifically ask that from your provider.

The seller must not use a guy who can change the tyres for installing driver boards, otherwise it must bear the responsibility for the end results. This must be done by an EE. 

If one chooses to change the board by oneself, then one is also aware of the risks.

Broken bones from cutouts must have consequences for the producer if the board is faulty.

@James M, it is difficult to appreciate at what speed it becomes dangerous to ride a wheel prone to cutouts. But generally it is advised to not go over 15kmph(10mph), as this speed may allow you to bail out and remain on your feet, or if not, at least not get too badly wounded. Unfortunately there is no guarantee you won’t get hurt at this speed. There are instances of people falling at very low speed (~5kmph-3mph) and still ending up with broken bones.

[Cress]

@vBlitz - I am very sorry for your crash + injuries.  My injuries have all been by bicycle, fractures of my clavicles (collar bones).  You will do more work in rehab compared to me. Your injuries are difficult BUT . . . I agree with @Paul g   you will do a good thing by reporting the incident to your re-seller and start a technical review.  Status of the V12 Hunteck motor-control board is cloudy unless each failure is reviewed and understood.  You have the EUC community waiting for your recovery while anxiously waiting to find out what broke on the V12.  {Please take that last statement in the best way  }

Edited July 13, 2022 by Cress

[vBlitz]

I will do a full report when I can, but I can say that for now:

Swap was made by an experienced guy (over 40 replacements), appointment provided by my reseller.

Reseller assured me it was 130v mosfets board.

The board did not came wit the heat dissipator already mounted.

The wheel does not work anymore and the motor oppose resistance (MOSFET related).

Cutoff happened on a very small bump on the road.

Fell backwards on my elbow, breaking  my humerus (twisted fracture).

I was wearing the same gear as always: motorcycle kevlar pants, kevlar jacket (back, elbow, shoulder pads), full face helmet, wrist guard, least dual axis .

Edited July 13, 2022 by vBlitz

[RArtem]

4 hours ago, vBlitz said:

I will do a full report when I can, but I can say that for now:

Swap was made by an experienced guy (over 40 replacements), appointment provided by my reseller.

Reseller assured me it was 130v mosfets board.

The board did not came wit the heat dissipator already mounted.

The wheel does not work anymore and the motor oppose resistance (MOSFET related).

Cutoff happened on a very small bump on the road.

Fell backwards on my elbow, breaking  my humerus (twisted fracture).

I was wearing the same gear as always: motorcycle kevlar pants, kevlar jacket (back, elbow, shoulder pads), full face helmet, wrist guard, least dual axis .

How many kilometers on new board? Was it first ride?

[vBlitz]

1 hour ago, RArtem said:

How many kilometers on new board? Was it first ride?

First ride, maybe 5km at most.

[RagingGrandpa]

On 5/10/2022 at 1:15 PM, mrelwood said:

It appears that the first customer has gotten their new upgraded V12 (og) board. They asked for a confirmation on IM Telegram chat for this Aliexpress bought board, which shipped on the first of May.

 The photos clearly show mosfets with a marking GP039N15M, which points to a 150V mosfet by Hunteck in the TO-220 package size:

http://www.hunteck.com/Datasheets/HGP(B)039N15M_V1.0_090320.pdf

Uh-oh, Ecodrift's batch of replacement boards showed up looking pretty sloppy... 

 

[Paul g]

2 hours ago, RagingGrandpa said:

Uh-oh, Ecodrift's batch of replacement boards showed up looking pretty sloppy... 

 

Thanks, @RagingGrandpa! I found a video with wave soldering process. 

Ecodrift.ru translated with google translate:

“Such a transistor will not be able to fit snugly to the radiator. And it is completely impossible to install it through a ceramic gasket.

But, unfortunately, it's even worse than it looks. The fact is that these transistors are overheated with poor-quality installation. That is why the traces are only on the first 4 pieces. The board with transistors is fed into a special conveyor, which passes it through a bath of solder (even before the installation of all planar elements). This process is called wave soldering. And the operator needs to make sure that the board goes smoothly, without stopping. If jamming occurs when the board is fed, then the first transistors simply overheat (in our case, these are the 2nd and 3rd). Then the board is pushed and it goes through the conveyor smoothly, so the rest of the elements do not have time to overheat. Overheated transistors can lose their properties already due to overheating itself, since they are critical to temperature, in addition, a very uneven surface is obtained, with which it is no longer possible to provide normal heat exchange.

In this story, once again, Inmotion shows that they cannot overcome quality control issues, as well as their subsequent attitude towards the problem, which they again blame users for.”

Edited July 13, 2022 by Paul g
add info

[Chriull]

1 hour ago, RagingGrandpa said:

Uh-oh, Ecodrift's batch of replacement boards showed up looking pretty sloppy... 

"Sloppy" is a very very nice word used for this ... 

 

[techyiam]

I just want add another possible source of causation for new board failures.

What I discovered when it was time to mount the new board to the heat sink, was that the holes of the mosfets don't match up that well to the holes in the heatsink. The holes in the old board actually matched better. The second thing I found was that almost all the original spacers were cracked, with pieces missing from in the shoulder area of the plastic mosfet screw insulators. At the time I discovered this fact, I hadn't tightened the screws yet, nor were the plastic pieces that cracked off can be found.

Theoretically speaking, even with the cracked insulators, no short would occur should the holes be aligned because the diameter of the screws is sufficiently smaller than the holes of the mosfet tabs. There is simply no contact. The issue arises when the holes are not centered, and by chance, one or more screws are forced to one side with the cracked off part of the insulator sitting next to the side of the hole of the mosfet tab. For those who have done the the installation, you would know not to tighten the screws too tightly. So, let say by chance, no contact is made initially and passes the continuity test. Everything gets put back together and off we go. Over time, depending on circumstances, the vibration and shakes may just be enough for one of the mosfet to move just enough for the screw to make contact. And that mosfet gets fried.

For my own installation, I decided to do something outside the box so I don't have to disturb the soldering of the mosfets, which would be required in order to align the holes. Also, I avoided having to get new insulators, and be severely restricted in tightening torque. 

Because the holes are somewhat offset, there is not enough room to fit in new insulators that are not broken. I instead apply super glue to just below the heads of the screws. I made two applications. This is the part of the screw that can make physical contact with the mosfet tab. I also use those red motherboard washers from the PC days to insulate the heads of the screws from the tabs. I use nail clippers to trim them to the right size. With this set up, I could tighten the screws as firmly as I like.

I also remove all thermal compound from the new thermal pad with alcohol. One interesting thing is when the thermal paste-free pad is aligned and pressed onto the heatsink, it sticks to the heatsink much better. It is not unlike applying tint film to the windows of cars. One needs to " squeeze out the bubble".

The first thing I did was to check the temperature. Everything looked good. Now after about a hundred or so kilometers in the hot weather that we are currently having, I found the temperatures to be not much higher than before.

Note that I am happy with the quality of my replacement board. There were no observable defects found on my new board. Even the tabs were all upright, basically perfectly aligned to each other, forming a straight line.

.

Edited July 15, 2022 by techyiam

[Paul g]

1 hour ago, techyiam said:

I just want add another possible source of causation for new board failures.

What I discovered when it was time to mount the new board to the heat sink, was that the holes of the mosfets don't match up that well to the holes in the heatsink. Actually, the holes in the old board actually matched better. The second thing I found was that almost all the original spacers were cracked, with pieces missing from in the shoulder area of the plastic mosfet screw insulators. When I discovered this fact, I hadn't tightened the screws yet, nor were the plastic pieces that cracked off can be found.

Theoretically speaking, even with the cracked insulators, no short would occur should the holes be aligned because the diameter of the screws is sufficiently smaller than the holes of the mosfet tabs. There is simply no contact. The issue arises when the holes are not centered, and by chance, one or more screws are forced to one side with the cracked off part of the insulator sitting next to the side of the hole of the mosfet tab. For those who have done the the installation, you would know not to tighten the screws too tightly. So, let say by chance, no contact is made initially and passes the continuity test. Everything gets put back together and off we go. Over time, depending on circumstances, the vibration and shakes may just be enough for one of the mosfet to move just enough for the screw to make contact. And that mosfet gets fried.

For my own installation, I decided to do something outside the box so I don't have to disturb the soldering of the mosfets, which would be required in order to align the holes. Also, I avoided having to get new insulators, and be very limited in tightening torque. 

Because the holes are somewhat offset, there is not enough room to fit in new insulators that are not broken. I instead apply super glue to just below the heads of the screws. I made two applications. This is the part of the screw that can make physical contact with the mosfet tab. I also use those red motherboard washers from the PC days to insulate the heads of the screws from the tabs. I use nail clippers to trim them to the right size. With this set up, I could tighten the screws as firmly as I like.

I also remove all thermal compound from the new thermal pad with alcohol. One interesting thing is when the thermal paste-free pad is aligned and pressed onto the heatsink, it sticks to the heatsink much better. It is not unlike applying tint film to the windows of cars. One needs to " squeeze out the bubble".

The first thing I did was to check the temperature. Everything looked good. Now after about a hundred or so kilometers in the current hot weather we have, I found the temperatures to be not much higher than before.

Note that I am happy with the quality of my replacement board. I observed no abnormalities, and the tabs were all upright, basically perfectly aligned to each other.

Nice! You might have some issues though with the superglue, because it is stiff when it hardens it cracks easily and falls off. Superglue also forms a very thin hard film which doesn’t add much distance between the screw and FET. I would probably try to find a glue like silicone, that keeps being elastic. It has to be a heat resistant one as well. Can you imagine what a shitty plastic they use for those insulators? 

[EUC Phenomenon]

1 hour ago, techyiam said:

For my own installation, I decided to do something outside the box so I don't have to disturb the soldering of the mosfets, which would be required in order to align the holes. Also, I avoided having to get new insulators, and be very limited in tightening torque. 

Because the holes are somewhat offset, there is not enough room to fit in new insulators that are not broken. I instead apply super glue to just below the heads of the screws. I made two applications. This is the part of the screw that can make physical contact with the mosfet tab. I also use those red motherboard washers from the PC days to insulate the heads of the screws from the tabs. I use nail clippers to trim them to the right size. With this set up, I could tighten the screws as firmly as I like.

I also remove all thermal compound from the new thermal pad with alcohol. One interesting thing is when the thermal paste-free pad is aligned and pressed onto the heatsink, it sticks to the heatsink much better. It is not unlike applying tint film to the windows of cars. One needs to " squeeze out the bubble".

A couple of questions. Did you still use the old insulators? and did you glue the washers on to the screws?

[techyiam]

4 hours ago, Paul g said:

Nice! You might have some issues though with the superglue, because it is stiff when it hardens it cracks easily and falls off. Superglue also forms a very thin hard film which doesn’t add much distance between the screw and FET. I would probably try to find a glue like silicone, that keeps being elastic. It has to be a heat resistant one as well. Can you imagine what a shitty plastic they use for those insulators? 

Thanks for the insight. However, the particular cyanoacrylate adhesive that I have been using for over ten years doesn't do that. In any case, without the insulators, the space between the screw and the hole in the tab is a good margin. Even if the glue flakes off, I still don't think they will short. I have tightened the screws tight enough. The mosfets won't be able to move.

I may inspect them again to see how they are fairing after more heat cycles. Currently, we are going through our hottest months.

 

Edited July 14, 2022 by techyiam

[techyiam]

36 minutes ago, EUC Phenomenon said:

A couple of questions. Did you still use the old insulators? and did you glue the washers on to the screws?

Nope. No insulators. Extremely high risk in my case to use the old ones, and the new ones won't fit. Nope. The washers are free to rotate.

[Cerbera]

On 12/31/2021 at 1:36 PM, Paul g said:

But we would end up looking like some star wars soldiers.

*nods enthusiastically

CBR

[Cress]

For-what-it's-worth, I used the approach @techyiam outlined, above, to remove corrosion in a marine power supply but marine polyurethane varnish instead of superglue.  The boat still works.  I'll trust @techyiam's  method to work on an EUC, to make the Huntech TO 220 MOSFET more reliable compared to original mounting hardware. 

I'd like to meet the guys who designed the shoulder-washers. 

[techyiam]

30 minutes ago, Cress said:

polyurethane varnish instead of superglue.

Interesting idea.

 

30 minutes ago, Cress said:

I'll trust @techyiam's  method to work on an EUC, to make the Huntech TO 220 MOSFET more reliable compared to original mounting hardware. 

Thank you for your kind words. However, I am still experimenting myself. But it makes sense to me. I hope they work as well for you as they have for me.

 

30 minutes ago, Cress said:

I'd like to meet the guys who designed the shoulder-washers.

I suspect the issue has to do with Inmotion choosing to use the conventional mounting method for mosfets, but soldered in the mosfets first, instead of soldering the mosfet after they are mounted on the heatsink. Since the alignment can't be perfect, there just isn't enough room for the insulators to fit. So when the screws are tightened, the shoulder of the insulators get crushed. They could have used a non-conducting bar to clamped the mosfets onto the heatsink, but the screws don't go through the holes in the mosfet tabs.

[jimjam.nyc]

10 hours ago, techyiam said:

I suspect the issue has to do with Inmotion choosing to use the conventional mounting method for mosfets, but soldered in the mosfets first, instead of soldering the mosfet after they are mounted on the heatsink. Since the alignment can't be perfect, there just isn't enough room for the insulators to fit. So when the screws are tightened, the shoulder of the insulators get crushed. They could have used a non-conducting bar to clamped the mosfets onto the heatsink, but the screws don't go through the holes in the mosfet tabs.

I had to slightly adjust each mosfet while screwing them in. I broke so many of those little plastic insulators installing my board. I also had to test each screw as I screwed them in. I can definately see why some of the installations are failing, even from experienced techs. There is a little bit of anxiety involved with how fragile those plastic insulators seem. I do plan on opening this thing up every 500 or 1k miles to check on it. Next time i do I may try this method to feel a little safer. Unless a better hack or official method comes to light before then. 

The new board seems solid so far. I have close to 200 miles so far on it, and i feel a little more confident every ride. 

[Funky]

Bigger plastic/rubber washers. Plastic screws?

Edited July 17, 2022 by Funky

[OldFartRides]

“ plastic screws ?”

Dunno, sounds kinda funky to me….

[techyiam]

14 minutes ago, Funky said:

Bigger plastic/rubber washers. Plastic screws?

If you have the right plastic screws, the washers would be optional. I think those mosfets can get pretty hot.

Edited July 17, 2022 by techyiam

[Funky]

1 hour ago, techyiam said:

If you have the right plastic screws, the washers would be optional. I think those mosfets can get pretty hot.

Ohh.. Yeah.. Forgot they become hot. Dohh, brain fart.  

Edited July 17, 2022 by Funky

[daniel1234]

If somebody in EU have busted v12 board, send it to me for repair and analysis. I can replace those joke "hunteck"  fets for real ones, like infineon or IXYS IXTP150N15X4A . Seems like you have guys a lot of bad experience - i think HGP039N15M is root cause - i would replace all of them. Except those cutout issues it seems like nice wheel.

[Paul g]

3 hours ago, daniel1234 said:

If somebody in EU have busted v12 board, send it to me for repair and analysis. I can replace those joke "hunteck"  fets for real ones, like infineon or IXYS IXTP150N15X4A . Seems like you have guys a lot of bad experience - i think HGP039N15M is root cause - i would replace all of them. Except those cutout issues it seems like nice wheel.

Are you a EE, and you know the IXYS IXTP150N15X4A will work with that board?

[daniel1234]

15 hours ago, Paul g said:

Are you a EE, and you know the IXYS IXTP150N15X4A will work with that board?

There are just 3 things you need to check replacement mosfet (all written in datasheet - don't believe people in internet :D):

- need to be approximately as fast in switching ns range and same purpose.

- similar load in A and voltage.

-  same pinout

Not sure what EE is . They seem similar - there is a lot of option depending on your region. Since there is semiconductor shortage you need to pick one available close to you in quantity you need . I am confused why people even wait for immotion fix. They are already working on v13 ... not sure who will want to drive 100 km/h with board quality like they showed at v12. I think they should redesign v12 to have some lightweight suspension and mainly fix board issues first.

Then there is price good fet cost 4+ dolars/euro in minimum. You need 12 of them with same characteristic ideally from same batch - cheep fets usually have high failure rate so one fet is different from other and that creates imbalances in circuit. They work - they are great for cheap stuff like UPS / computers e.t.c. 

[spitfire1337]

6 hours ago, daniel1234 said:

There are just 3 things you need to check replacement mosfet (all written in datasheet - don't believe people in internet :D):

- need to be approximately as fast in switching ns range and same purpose.

- similar load in A and voltage.

-  same pinout

Not sure what EE is . They seem similar - there is a lot of option depending on your region. Since there is semiconductor shortage you need to pick one available close to you in quantity you need . I am confused why people even wait for immotion fix. They are already working on v13 ... not sure who will want to drive 100 km/h with board quality like they showed at v12. I think they should redesign v12 to have some lightweight suspension and mainly fix board issues first.

Then there is price good fet cost 4+ dolars/euro in minimum. You need 12 of them with same characteristic ideally from same batch - cheep fets usually have high failure rate so one fet is different from other and that creates imbalances in circuit. They work - they are great for cheap stuff like UPS / computers e.t.c. 

If you don't know what EE is then your surely not qualified to make those kind of decisions... just my 2cents