Worst nightmare came true. Help me find out why…

My biggest fear when going around on an eskate are a sudden electronic failure or something similarly, throwing me of the board out of the blue. That happen tp me today and I can’t figure out what the hell happened.

I was doing my speed runs I just launched for the one where I really would give it all. Slightly downhill and the wind in my back. But I didn’t get very far before the board threw me off. It kinda felt like that the board just stopped / blocked or something. Luckily it was in the beginning of my acceleration so I was not going very fast (20 mph), but it was enough speed to launch me forward. Reflectively I took a long step and landed first on one foot and then down on my stomach and hands sliding. I sled long enough to think “hey now I am sliding. Wonder how my shirt will hold up. Kinda nice feeling sliding when I think of it :slight_smile: I was wearing knee pads, hillbilly crash pants, kvelar shirt, elbow pads, and slide gloves. So I am fine and nothing seriously happened.

The thing is that after the board worked fine. No sign of things that were wrong. I only took the board to 17 mph on my way back. But no issues.

I had the metr app running when I crashed and the only thing that stands out are the motor current that did go all the way to 118 two sec before the crash (I did start the acceleration from standing and slightly up hill). That might be the reason. But would that block the motors and then afterwards work fine again? When I look that other vesc data I have I see that the 118 motor amps are the highest I have seen yet. Normally its around 80a when It’s the most. I haven’t been close to 118 before.

Anyone have any suggestions on what might have happened. I know you don’t have much info to guess from. But Anyway.

This is the metr data: https://metr.at/r/LEFou

2 Likes

Motor amp spike caused your motor controller to shut down. The sudden loss in accelleration feels like the board is braking but it is not. What is your motor max set at and gear ratio/top speed? The taller the gear ratio the more susceptible the system becomes to current spikes when pushing the throttle.

7 Likes

If the Vesc short down wouldn’t we see that the Voltage would drop very low in the metr data? It only drops to 35V. Or would the short down be so short that it wouldn’t be caught in the data? I have tried to short the board down while giving full throttle and the V drops tp 9V And sometimes, but not every time, the metr app gives a warning saying “Under Voltage”. That didn’t happen with my crash.

My motor max is set at 80a and my gear ratio are 16/40 and I hit 37 mph as max speed.

Crash data: image

Under Voltage Warning when testing shirtinget of the board. But that warning only came once, and I turn off the board multible times: image

I am really surprised about that spike in motor amps. I haven’t seen that before. Normally I am around 80 when it is highest. And I didn’t feel I was accelerating that hard. Strange.

Sounds like motor desync due to hard throttle input that can’t be followed and then resulting in cogging at high speed felt like locked motor. Huge amp spikes also indicate that your esc had a super hard time driving the motor and/or the current ripple from the desync overflow the current management loop Do a better detection and try again Also hardware side is extremely important, make crazy strong soldering job and ensure electricity can flow with no bottleneck. Use ton of tin, big cables, and be sure vibrations can’t loosen connection

8 Likes

Would you do the detection with or without wheels on?

The vesc has a hard cutoff at 120 amps and in your app data you hit a peak of 120.7 amps. This only cuts throttle and does not shut the whole system down. It looks like your board is set up with a top speed just under 40 mph, this is high enough to start triggering these spikes when throttling hard at 50% duty cycle.

If you lower your max motor amps down to 50 or 60 amps it will give the soft cut off more time to engage and prevent the hard cutoff.

Motor detection should be with the motor free to spin, no belt or gears should be attached.

12 Likes

What kind of VESC are you using, if you don’t mind me asking?

Ahh that makes sense… Thanks Chaka.

Would lowering the motor amps make the the top speed less?

It sunds like I need to upgrade my setup to handle more strain if I want keep doing these speed runs. What would you suggest I look at?

I don’t mind… two focboxes.

It will not lower the top speed but it will lower the torque you feel during acceleration. Looking again at your app data and it looks like your motors are being driven very hard, you can see the temp rise when the board stopped moving. They could have picked up some resistance from heavy use so It might be worth while to recalibrate to help prevent another hard cut-off.

1 Like

I think it is time for all of us to move on to a more robust MOSFET like the to-247 package in order to get the power we need from a dual motor setup.

6 Likes

“recalibrate” now we are entering unknown territory for me.

You are right. I properly drive my motors to hard. they are 6365 200kv. I run pneumatics 150mm.

I did adjust the belts a little before this run. Made them a little tighter. And one motor are saying a noise. But thats not in the motor when there is belt on. I think its from a loose motor pulley, where the scrub screw is funck and cant be tighten anymore. So the bully sit loose and makes a noise.

But I will check everything again.

What would be a reasonable temp to have. When I see other data I have the temp hit 60 degrees C.

When a say recalibrate I mean run motor detection again. :wink:

I also recommend taking a mental note on how much torque you were feeling when the cutout happened. You most likely hit the maximum power you can pull out of a vesc based on v4.12

6 Likes

Thanks Jeramiah. You helped me find a plausible reason why this happened. That takes away the unknown. I was afraid of the unknown. Now I feel ok comfortable hitting the speed track again tomorrow.

4 Likes

Photos of the shirt plz!

4 Likes

Could also be that the motor saturated and the observer lost track of it

You could a bit with saturation compensation, but only way to test would be induce the problem again, not sure if you are up to it

2 Likes

Looks like the experts are finishing up this case :+1:. I recently had a scary fall. I was not so composed in the aftermath. I put the remote in a vice and crushed it.

I think I’m going to add Metr Pro to my essential riding gear right after helmet… and pants . Heal up bro :call_me_hand:

6 Likes

So I had the exact same fall and I figured out the reason eventually by doing a bench test. Accelerate the board by pulling the full throttle on a bench, does one of the motors stall? In my case the motor was stalling only when accelerating too quickly. The reason was bad motor detection. That fall fucked me up real good. To avoid this do a bench test every time you do any changes to the drive train. If there is cogging, time to re do the motor detection.

3 Likes

I did a lot of bench throttling afterwards including full throttle for 5 min. Nothing really happened. Everything seemed fine. I will try again.