V11 Just... Died. AGAIN!!

[MacGyverCanada]

After completing my previous V11 repairs I did a couple of short test drives and commuted to work a few times.  It was a real nice day today so I decided to head out on an adventure ride.  There's a long rails-to-trails system nearby so I drove to the trailhead and got rolling.  I had a vague destination in mind, but really I was just cruising in the sunshine.  I have the speed on my V11 limited to 45 km/h and I wasn't pushing it to the point of beeping, but was cruising at the upper end, 40'ish or so.  I was on a flat section of packed crusher dust.  No ruts or potholes, wasn't picking up any gravel, everything was smooth sailing. 

And then it cut out.

My head was catapulted into the ground and took most of the impact.  I don't think I'm concussed, but we'll see how I'm feeling in the morning.  My pads took care of me pretty well, but my leg is gonna have a nasty bruise and my shoulders aren't working quite right. 

The wheel is dead.  Again.  The front headlight assembly of the V11 was busted up pretty bad; no response from the power button.  As I pushed the wheel along to get to the nearest roadway I found that the wheel was cogging with some significant force, which I'm pretty sure means blown MOSFETs; this would be on the new driver board which has less than 100km of usage on it.  I'm not so sure that I even want to repair this wheel again, as I can't say that I have much confidence in it now.  Le sigh...

[alcatraz]

That's a terrible experience. Can you return it somehow or exchange it?

Yeah something's definitely off. 

I had something like that happen on my msx. I had ridden it for thousands of km at speeds up to 45km/h and then it cut out at 15-20km/h without even accelerating. I still to this day don't know why it blew a mosfet at low speeds but it did. I got it repaired and now it's working again.

I have suspicions of what it was. I think one of the mosfets wasn't bolted down into the heatsink very well. The control board can report 40-45degC temperatures all day and that mosfet would still be running hot for hundreds/thousands of km, until it gave up one hot summer day. At low speed of all times. (thank goodness) I think it must have been fatigued or something. Imagine it reaching like 100degC and then cooling down a few thousand times. (I ride carefully. I'm lightweight. I don't lean on the wheel squealing up a hill for long periods. So it's been getting hot and cooling countless times.)

So what I did was tripling the fan capacity by adding two additional 40x40mm fans next to the existing one, and I ensured that all mosfets are nice and tightly torqued into the heatsink.

I can't say for sure this was the problem but I'm hoping it was. I never ride the wheel under a 20% safety margin in EUC world so I don't accept that I overleaned it. This is on a wheel with 9000km on the clock too, so it's not some dud product in constant repair either.

Edited December 5, 2022 by alcatraz

[alcatraz]

Another thing I considered doing was to build my own crude little MOSFET thermal monitoring system.

There are these tiny temperature controlled switches you can buy for pennies. Imagine gluing one to each mosfet, or between two, and then connect them to a buzzer or flashling light or something. Pick a switch that closes at like 50degC or something conservative. Then it seems nearly impossible to blow a mosfet. Without heat I don't think you can blow them.

[Tawpie]

A local rider had his V11 dump him... careful inspection of the driver board revealed a cracked MOSFET. No magic smoke escaped, and no visual indications of excess heat. I think it was due to mechanical stress. In the past I've whined about how difficult it would be to build the V11 (and V12) driver boards because if you don't assemble it to the heatsink just so you can create mechanical stress in the MOSFET body. I believe this is why replacement V12 boards were eventually sent out with factory attached heatsinks... it's not something you should be doing if you don't have the tooling and processes to get everything aligned. The V12 inherited the mechanical design of the MOSFET-to-heatsink interface from the V11.

I feel bad for you getting dumped (especially dumped again), and do hope you're recovering well.

[mrelwood]

Having a cut out at those speeds is definitely no joy. I can easily see how confidence is tough to get back.

I agree with the above comments suspecting the mosfet to heat block interface. I modified mine to use mica discs, and resoldered the legs after reinstalling the board. That way I tried my best to make sure that all mosfets sit flush, have a great thermal connection, and have little or no tension in the legs.

I did the same mod on a local V10F since the stock system just didn’t cut it by my standards.

I guess there didn’t use to be anywhere near the requirements for impeccable cooling mechanisms that there are now, and Inmotion hasn’t been ahead of the game as much as they could’ve.

[MacGyverCanada]

Been back and forth with Andrew @ Speedy Feet who's been great; I ordered the replacement driver board from them and they've submitted a warranty claim to InMotion (pending).   Andrew was thinking that this is prob an issue with the signal from the Hall sensor and suggested that I strip and re-insulate the cables leading to the motor before replacing the driver board again:
 

Quote

From what ive experienced in other inmotion wheels hall effect failure or any interruption to the signal feed often results in mosfet board failure. A wheel failure nearly always includes a board failure. And when ive replaced a board without checking the hub motor, i often soon after get another board failure. This is probably because of how the board interprets the (often) absence of a signal. On the other hand twisted wires could inductively effect the signal going up the hall effect wires(5x thin wires) as its 5v and the motor wires have in excess of 80v(depending on the wheel) it would seem fairly easy to think of a pulse being induced from movement/water/corrosion ect. This would confuse the board thinking its fully on or fully off and damage electronics at the same time. 
 

If it was me, id remove the motor, strip back the wires to where it is worn, and individually heatshrink each one with high temperature heatshrink. Then reassemble with glass fiber insulation between the 3 motor wires and secure everything well. Using automotive silicon(RTV or similar) to secure at entry/exit points(such as out of the axel). Think of it surviving under salt water permanently and you get the level needed.

 

Afterwards you should check the hall sensors function. directions for this can be found online and is the same for ebike hub motors. 

I haven't actually cracked open the V11 since the crash; been busy as heck and low on workspace due to home renovations.  But I will check for MOSFET damage when I get around to it.  Any concerns with leaving the batteries connected in the wheel's current state (pun somewhat intended)?  I've also been shopping around for replacement wheels, reading reviews on off-road wheels and... gotta say, it seems like repairing this V11 is the way to go, if it can be done with some assurance of safety.  The V11 suits my needs quite well.

Edited December 7, 2022 by macgyvercanada

[Tawpie]

8 hours ago, macgyvercanada said:

Any concerns with leaving the batteries connected in the wheel's current state (pun somewhat intended)? 

I would disconnect them... if something is wrong with the motherboard, no sense keeping it hooked to the batteries!

[alcatraz]

Bad hall sensor does sound like a culprit. I hope it works well after a switch.

I hope they have a way to measure the hall sensor output in like a testbench before they assume it's all ok and let you be the testdummy.

About the cooling thing. I'd love to have some way to see mosfet temperatures while riding. I wonder if I should simply hotglue a little temp sensor to one and put a display on the outside. What about the other mosfets? 

[mrelwood]

On 12/9/2022 at 5:27 AM, alcatraz said:

I wonder if I should simply hotglue a little temp sensor to one and put a display on the outside. What about the other mosfets? 

The mosfet casing can get pretty hot, so hot glue would probably melt.

 As long as all mosfets are all installed correctly, their temperature should be practically the same.

[alcatraz]

Right. I'd love to be able to see their max temperatures after a ride. Like a FLIR recording or whatever. Obviously that's not possible. 

Haha, maybe just put a dab of hotglue on each and see later which melted off first.  Hmm, I'll keep on brainstorming. 

[MacGyverCanada]

Huh well I do have a FLIR camera.... I would just need to ride with the hood open, so to speak.

[MacGyverCanada]

Wrapping this up, I received my warranty parts and had my local e-scooter repairman help with the repairs.  

...and everything was going great until:
https://forum.electricunicycle.org/topic/34022-v11-please-repair-below-100-charge/

[MacGyverCanada]

I thought y'all might enjoy this...  big arcflash scar on the dead driver board.  Methinks this caused my December 2022 cutout crash...

[mrelwood]

2 hours ago, macgyvercanada said:

I thought y'all might enjoy this...  big arcflash scar on the dead driver board.  Methinks this caused my December 2022 cutout crash...

The small bridges between the large surface track areas are 30A fuses (each). They really have gone out with a serious bang.

[alcatraz]

If you want to avoid this type of failure (and perhaps introduce other failures) try Begode. They rarely have fuses. Today they might though.

It's a tricky subject. Self-destruct vs self conservation (+rider injury).

[alcatraz]

What kind of "minimum safety margin" are you seeing in Eucworld on rides like these? Are you going past 0% or what?

[mrelwood]

 

4 hours ago, alcatraz said:

It's a tricky subject. Self-destruct vs self conservation (+rider injury).

Generally the fuses are rated high enough that currents that high would indeed do damage elsewhere already. Way past the wheel’s intended performance. It’s beneficial not to short the batteries in case of a jammed tire etc, as the controller wouldn’t be able to ride it out anyway.

 If one destroys V11 boards like underwear, one really should think about getting a more powerful wheel. It sounds like them might be expecting aspects from the wheel that it just doesn’t have. There are many more powerful wheels available, from all manufacturers. Though all much heavier of course.

[alcatraz]

It's like flavors of ice cream. 

[MacGyverCanada]

Quote

 If one destroys V11 boards like underwear, one really should think about getting a more powerful wheel. It sounds like them might be expecting aspects from the wheel that it just doesn’t have. 

Well I have had one board failure while walking my wheel with the trolley handle and another when cruising on flat ground at 40 km/h.   If that's pushing it too hard...

[alcatraz]

Begode can have board failures too but fuses blowing (self preservation) isn't one of them. (unless new designs have fuses, I ride older wheels.)

[Punxatawneyjoe]

2 hours ago, alcatraz said:

Begode can have board failures too but fuses blowing (self preservation) isn't one of them

that's great.Good point and probably true. The good news is , a couple of blobs of solder and Mac is back in action....problem solved..

 

On a serious note: glad you were not seriously injured. Let's face it, we are all just a little bit crazy for riding these contraptions.

[mrelwood]

18 hours ago, macgyvercanada said:

Well I have had one board failure while walking my wheel with the trolley handle and another when cruising on flat ground at 40 km/h.   If that's pushing it too hard...

Wow. Definitely not pushing too hard. It beats me why some users have continued issues with their boards, while all the rest of the thousands of units are fine.