V12 Cutout tracking

[Paul g]

On 6/12/2022 at 2:07 AM, Silverfish said:

Picked up my V12 from eevees today and took the long way home so I rode about 33k. So, there's a success story to add to the mix.

Did they told you which Mosfets does it comes with, HY5012(TO247) or the HGP039N15M(TO220)?

Edited June 15, 2022 by Paul g
my bad English 😭

[Tawpie]

56 minutes ago, Paul g said:

And @Tawpie, were you worried the tiny- tiny tension in the MOSFETs legs could create problems after installation?

100%, and from well before the V12. The way they did the MOSFET attachment on both the V11 and the V12 bothers me. I personally know one (former) V11 rider who had a MOSFET literally crack (and dump him on his face 'like a sack of potatoes' at 20 mph). Other V11 owners have reported not-burnt-up MOSFET failures as well. Parts don't crack on their own.

If it were me and my skin on the line, I'd either hot air reflow all of the MOSFET solder joints after attachment to the heatsink (has some risk, but if you know how to solder and have the right soldering tools it's not an issue) or try to figure out a way to allow the MOSFETs to 'float' a little at the heatsink attachment point. Not cinched down hard. The problem with not cinching them down is that you may not get good thermal contact to the pad, and goobering on thermal compound is not the way to fix that problem.

The problem with cinching them down hard is the heatsink will expand and contract at a different rate than the printed circuit board—and it's long. Because the heatsink will get longer than the printed circuit board will (or won't get as long) thermal expansion and contraction will introduce stress into the MOSFET leads and eventually work harden the leads and the solder, or could cause the epoxy body of a MOSFET to crack. Will it? Probably not.

As much as I was tempted to buy a V12 as an upgrade to my 16X, the heatsink attachment to the MOSFETs is enough of a worry to me to make me steer clear.

I'm not saying it's a bad design, just saying that it's tricky enough to not allow enough room for error and tolerance stackup for my skin.

Edited June 15, 2022 by Tawpie

[Paul A]

Would be informative if Inmotion were to make public the qualifications/experience of their electrical engineers.

[Tawpie]

6 minutes ago, Paul A said:

Would be informative if Inmotion were to make public the qualifications/experience of their electrical engineers.

This is actually what we called a "packaging problem", and it's one for the gear heads. EEs (wire weenies) can tell you how much heat to expect and how quickly the heat load will change, but the MechEs are the pros at dealing with moving heat around. And the Manufacturing Engineers responsible for assembly should have tricks up their sleeve—but those folks would only work on to the process of creating the board+heatsink assembly. Before I went to the electrical design side, I spent a couple of years as a Manufacturing Engineer in the PCB assembly factory trying to figure out how to assemble the nightmares handed to us by the design side.

Edited June 15, 2022 by Tawpie

[Paul g]

2 minutes ago, Tawpie said:

This is actually what we called a "packaging problem", for the gear heads. EEs can tell you how much heat to expect and how quickly the heat load with change, but the MechEs are the pros at dealing with moving heat around. And the Manufacturing Engineers responsible for assembly should have tricks up their sleeve—but those folks would only work on to the process of creating the board+heatsink assembly. Before I went to the electrical design side, I spent a couple of years as a Manufacturing Engineer in the PCB assembly factory trying to figure out how to assemble the nightmares handed to us by the design side.

Yes, because the manufacturing engineers and the ones doing the design were not working together. Tesla is very aware of these issues in the manufacturing and learned the lesson the hard way. The problem is some companies never ever figure this out.

[Paul g]

22 minutes ago, Tawpie said:

100%, and from well before the V12. The way they did the MOSFET attachment on both the V11 and the V12 bothers me. I personally know one (former) V11 rider who had a MOSFET literally crack (and dump him on his face 'like a sack of potatoes' at 20 mph). Other V11 owners have reported not-burnt-up MOSFET failures as well. Parts don't crack on their own.

If it were me and my skin on the line, I'd either hot air reflow all of the MOSFET solder joints after attachment to the heatsink (has some risk, but if you know how to solder and have the right soldering tools it's not an issue) or try to figure out a way to allow the MOSFETs to 'float' a little at the heatsink attachment point. Not cinched down hard. The problem with not cinching them down is that you may not get good thermal contact to the pad, and goobering on thermal compound is not the way to fix that problem.

The problem with cinching them down hard is the heatsink will expand and contract at a different rate than the printed circuit board—and it's long. Because the heatsink will get longer than the printed circuit board will (or won't get as long) thermal expansion and contraction will introduce stress into the MOSFET leads and eventually work harden the leads and the solder, or could cause the epoxy body of a MOSFET to crack. Will it? Probably not.

As much as I was tempted to buy a V12 as an upgrade to my 16X, the heatsink attachment to the MOSFETs is enough of a worry to me to make me steer clear.

I'm not saying it's a bad design, just saying that it's tricky enough to not allow enough room for error and tolerance stackup for my skin.

How would you solve MOSFETs and heatsink contraction and expansion issue?

[Tawpie]

17 minutes ago, Paul g said:

How would you solve MOSFETs and heatsink contraction and expansion issue?

That needs analysis to know if, and how much of a problem it actually is. They very well may have considered this already!

If you can't change how everything is arranged, then you look to reduce the differential. One way is to try to match the TCE (thermal coefficient of expansion) of the heatsink material to the PCB as best you can. That's a materials choice for the heatsink, or at least for the section that bolts to the parts. Mostly I'd do what I could to avoid 'long' heatsinks bolting a row of parts together. Saw the heatsink into sections?

Keep in mind I'm just eyeballing this from the outside. What I see really only says "you'd better take a good close look at your numbers", it doesn't say that what they've done isn't satisfactory.

Another option is to figure out how to get enough cool air to the MOSFETs so that you can cool them without relying as much on the thermal connection to the heatsink and might be able to get away allowing them to 'float' a little more. At that point you can stand a bit more thermal impedance between the parts and the heatsink because air is taking some of the heat away. Fans?

You could also make sure the heatsink doesn't actually get very hot, that'll reduce its expansion/contraction. If I were to bet, I'd bet "heatsink never gets very warm" is what IM is relying upon.

Edited June 15, 2022 by Tawpie

[Silverfish]

50 minutes ago, Paul g said:

Did they told you which Mosfets does it comes with, HY5012(TO247) or the HGP039N15M(TO220)?

I didn't ask. I only know what a mosfet is in the first place because of this experience, so it would be in one ear and out the other. 

[Freeforester]

It would appear that they do not help the situation by posting them out in inadequate packaging, from what I have seen. Who benefits from such an approach, one wonders.

[Paul g]

18 minutes ago, Tawpie said:

Mostly I'd do what I could to avoid 'long' heatsinks bolting a row of parts together. Saw the heatsink into sections

Was thinking at the same thing. Or making a separate heatsink for each mosfet? Or even better: provide from factory mosfets with integrated heatsink? That is why close integration in manufacturing is important. One screw holds mosfet directly on the moterboard ( can be even glued to it ) has its own heatsink- no expansion issues. 

 

 

[Paul g]

19 minutes ago, Freeforester said:

It would appear that they do not help the situation by posting them out in inadequate packaging, from what I have seen. Who benefits from such an approach, one wonders.

They appear to try to solve this in their last announcement. 

[Paul g]

38 minutes ago, Tawpie said:

 

You could also make sure the heatsink doesn't actually get very hot, that'll reduce its expansion/contraction. If I were to bet, I'd bet "heatsink never gets very warm" is what IM is relying upon.

Still, there can happen to have big discrepancies between internal board temp and heatsink temp. Example: it is winter, had the wheel inside at worm temp. take the wheel outside, very cold. the heatsink contracts, the motherboard inside is still worm-> mosfets soldering cracks!

[Paul g]

39 minutes ago, Silverfish said:

I didn't ask. I only know what a mosfet is in the first place because of this experience, so it would be in one ear and out the other. 

Could you please ask your provider? Would be very important for us to know about what kind of board-mosfets combination your V12 has so we know what we’re evaluating when you report about your experience with it. 

[Richardo]

On 6/10/2022 at 5:56 AM, adrianqbs said:

We also took the extra step of testing with each of the 3 MOSFET feet individually

I'm not an expert, but I believe you want to avoid touching the middle MOSFET foot (the gate) at all! 

the gate is extremely sensitive to current, and the MOSFET can be ruined by a static spark too small to see or feel. The continuity tester has to run a current through the circuit to test it, and I believe that could potentially damage the MOSFET. 

Can an EE chime in to verify my paranoia? Or dismiss it?

[Chriull]

51 minutes ago, Richardo said:

I'm not an expert, but I believe you want to avoid touching the middle MOSFET foot (the gate) at all! 

the gate is extremely sensitive to current, and the MOSFET can be ruined by a static spark too small to see or feel. The continuity tester has to run a current through the circuit to test it, and I believe that could potentially damage the MOSFET. 

Can an EE chime in to verify my paranoia? Or dismiss it?

A continuity tester has (normaly) a low testing voltage. Such Power Mosfets as used in EUCs allow often +/-20V between Gate and Source.

So there is no concern or damaging potential.

Static electricity is very different!

Edit: But there is nothing to be gained measuring from all the mosfet legs - so there is no reason to touch/measure them...

Edited June 16, 2022 by Chriull

[techyiam]

22 hours ago, jimjam.nyc said:

i lowered my footplate position from highest to lowest when taking the wheel apart. It makes a decent difference. 

After the lowering of your pedals, did you noticed any increase in wobble tendency, especially in emergency braking. 

 

22 hours ago, jimjam.nyc said:

Streets here in NYC suck bad.. So i am gaining my trust quickly, it handled the potholes and bumps really well so far

What was your tire pressure? I believe you had changed your tire, and it is now a H666?

[jimjam.nyc]

2 hours ago, techyiam said:

 

2 hours ago, techyiam said:

After the lowering of your pedals, did you noticed any increase in wobble tendency, especially in emergency braking. 

It is hard to say really. I wouldn't say i noticed any increase in wobbles with emergency braking. I noticed more of a gyro effect in general which threw my Chi off a bit for the first few miles. After about half hour or so i adjusted. While I feel like it takes more energy to get the wheel to carve at the lower height, it is not any less nimble or stable. Just feels different. I may put the pedals in middle at some point. I just like them lower now that i have a seat, its just a bit more comfortable going from standing to sitting. The adjustment in ride style will be something that becomes second nature sooner or later. 

 

2 hours ago, techyiam said:

What was your tire pressure? I believe you had changed your tire, and it is now a H666?

I did change to the H666 pretty close to when i got the wheel. My tire pressure is about 23-24psi. For context i am a small guy. about 5'3 and 135lbs. So i keep my pressure sort of low. I bounce between 28psi and 23psi. These days and on the h666 i am liking the lower pressure. The h666 tire can just handle more than the stock. Plus at higher pressures and my low weight, i tend to bounce like a beach ball going over bumps at high speed :-).

I believe my 16x has the same tire on it as stock v12. I keep that at around 26psi. I recently rode the 16x. When i take a train to go out to my folks house, i usually grab the 16x. Its just a bit easier to travel with on small commutes.

[techyiam]

2 hours ago, jimjam.nyc said:

It is hard to say really. I wouldn't say i noticed any increase in wobbles with emergency braking.

Thanks. The reason I asked is because my V12 had a death wobble issue before. I just wanted to confirm whether others have them. Back then, if I lower the pedals to the lowest height, the wheel would go into a death wobble during braking over 50 km/h+. But if I put the pedals right back to the highest height, the death wobble would go away instantly. I don't have wobble problems anymore, though.

2 hours ago, jimjam.nyc said:

These days and on the h666 i am liking the lower pressure. The h666 tire can just handle more than the stock. Plus at higher pressures and my low weight, i tend to bounce like a beach ball going over bumps at high speed :-).

You must have good knees. We have bumps on our side streets but there are not enough real pot holes that are worth mentioning. But even then, and at 20 psi with stock tires, my knees complain. And this for only below 40 km/h riding. But now, I am wanting to try some 50 km/h sections. I think I have managed to overcome the wobble issues, and now I focusing on overcoming the issues arising from bumps.

[jimjam.nyc]

1 hour ago, techyiam said:

You must have good knees. We have bumps on our side streets but there are not enough real pot holes that are worth mentioning. But even then, and at 20 psi with stock tires, my knees complain. And this for only below 40 km/h riding. But now, I am wanting to try some 50 km/h sections. I think I have managed to overcome the wobble issues, and now I focusing on overcoming the issues arising from bumps.

If you are a new or newish rider you will hear alot of people say keep your knees bent. Which is true, but i find its only half the story. After many miles i realized my knees were bent but I was still keeping everything stiff.. I think once you are on autopilot with bent knees (not straight legged), start getting used to keeping your lower body in particular loose. When you hit the bumps, your legs and knees will move with the wheel taking some of the punch out of the bumps. Obviously there is a limit to what a body can take, but this has helped me a lot on the crap streets over here. Saved my knees and more than a few falls. 

It has taken me quite a while to get comfy riding and learn to keep loose!

Since this is way off topic here i will just add.. Still going strong on the new v12 board. So far no issues. Did close to 60 miles so far. The board seems to run quite a bit hotter than the old one. Not sure if this is a known thing or i botched heat sync stuff.. It has not overheated yet, but its def something i am noticing.

[Steve Evans]

1 hour ago, jimjam.nyc said:

The board seems to run quite a bit hotter than the old one. Not sure if this is a known thing or i botched heat sync stuff.. It has not overheated yet, but its def something i am noticing.

the new board might  be just APPEARING to run hotter. the new board has repositioned the temp sensor from the middle of the board to right between 2 of the mosfet legs. so it might be that the new board is better at temp measurement.

btw, i have noticed the same. seems to be about 10/15 degrees C hotter.

Edited June 16, 2022 by evans036

[techyiam]

1 hour ago, jimjam.nyc said:

After many miles i realized my knees were bent but I was still keeping everything stiff.. I think once you are on autopilot with bent knees (not straight legged), start getting used to keeping your lower body in particular loose. When you hit the bumps, your legs and knees will move with the wheel taking some of the punch out of the bumps.

Thanks for the tip. I think you are quite right. If you can get your body to move up with the wheel when impacted by a jolt, that is going to help a lot. A good example would be the world class slalom skiers. Their legs work like shock absorbers for the upper body.

 

1 hour ago, jimjam.nyc said:

Still going strong on the new v12 board. So far no issues. Did close to 60 miles so far. The board seems to run quite a bit hotter than the old one. Not sure if this is a known thing or i botched heat sync stuff.

I wonder if it would help it to run cooler if the old thermal paste was removed first, and then put on the new thermal pad without the thermal paste?

Edited June 16, 2022 by techyiam

[Tawpie]

1 hour ago, jimjam.nyc said:

When you hit the bumps, your legs and knees will move with the wheel taking some of the punch out of the bumps

So true!

When you watch the hardcore single track riders, watch their head. It (basically) doesn't move up and down. Then watch their knees... it's amazing. I've discovered that on pavement I tend to get locked up whereas when I'm off-road, the constant bounding about keeps things looser. I continue to work toward that fluid knee action, for me it's taking much time and practice practice practice.

[Richardo]

My body just started understanding this when hitting speed bumps at speed.
 

There is some magic to keeping your legs “loose” and boom they can just flex with the movement, leaving your torso alone. 

I think i was slow to learn this because i went from a mTen3 that i dont take any bumps with to a v11 that took the bumps for me. Its only after riding the v12 for a while that i started to adapt. 

[rolis]

I would just like to share on what to look out for when installing the new board besides shorts, and checking for blown mosfets.

 

There are two very important steps when to look out for mistakes:

1. When putting together the case of the two boards - Be sure, that the motor phase wires are not in the way of anything. There is a heatsink on the blue board that might squish one of the wires. And be careful that the other wires are not pushing on any other components. The phase wires are very tough to bend.

 

2. When installing the top part with the display and the handle - There are a lot of wires that can get stuck in between parts and get damaged. Check before and during the installation of the part if there are any wires getting squished.

[Jeff Earl]

So I've finally accumulated 100mi on the new controller and l'm beginning to feel pretty confident on the V12. Not shorts and flip-flops confident, mind you. More like commuting to work without full body-armor confident. I'm not an aggressive rider by any means. I've made no attempts to stress test or overpower my wheel, but I have experienced the real-world majesty and variety of NYC terrain it's finest.

I've traversed the frighteningly uneven expansion joints and scarred assents oofthe interboro bridges. I've bounced my way over unmarked speed bumps and over-filled potholes patched by inebriated road workers. I've slalomed  my way through the war-zone caliber minefields that pretend to pass for city streets. I've dodged oblivious Uber drivers and zoned-out tourists simultaneously jaywalking and gawking at skyscrapers; mouth agape.  I've performed emergency maneuvers to avoid the bloody entrails of urban road-kill, and snaked my way around steaming mountains of unidentified mammalian feces. And remarkably, I've somehow even survived the brutal and appallingly unimaginative verbal assaults of Spandex-clad cyclists on a mission to save the world from the lazy.

All in all, it's been a dream. The wheel has performed flawlessly for me. Aside from some footplate angle weirdness (probably attributable to operator error), I have no real complaints to report. I have noticed that the V12 seems to favor a slightly lower tire pressure than my 16X running the same CST tire. For some reason, the 38psi that I was accustomed to sporting on the Kingsong results in tooth rattling vibrations on the V12.  I seem to be settling on a pressure in the neighborhood of 33-34psi for my "muscular" 200lb riding weight.  I'm also still experimenting with the various split-riding modes, but seem to be gravitating to a profile that suits me well enough to stop tweaking it with every ride.

It's nice to finally be riding this wheel after months of parking it in my apartment's corner of shame. Hope others are finally experiencing some well-deserved fun with their V12s as well.