Vesc Dirt bike....tuning help needed

[Alpacaslapper]

Ok so Im losing my mind, I got a mototec kids dirt bike off facebook, it was trashed, my son and I cleaned it up and I ordered a motor from electro and co. 72v motor.

ETM™ Brushless Motor

Electro and co isn’t really wanting to help me with this motor because they only want me to buy their 400$ controller.

I ordered a votol em100gtr from AliExpress. And I used a 12s8p 40t battery I had. Now the votol was a pain to tune but I did get the bike rideable on 48v, but the controller was supposed to be for 72v so I built additional packs to put in series. Wrapped the packs in cardboard and shrink wrap.

And I set the votol for 72v. Well my son rode around the yard twice and parked the bike, the next day it wouldn’t run. I think its a shorted mosfet phase in the controller.

Now I ordered a makerbase 75200 v2. Got it set up in the bike and got the power button sorted out so that I can turn it on and off, etc.

Now the problem, I can’t get this dang thing to stop cogging no matter what I do, I’ve tried with the firmware that came on the vesc, with phase filters on and off. I’ve tried bldc mode, I’ve tried multiple times redoing the motor detection wizard. It’s currently on the 75200 v2 no limits firmware that removes the current noise from the 75200 update thread.

On the stand the motor goes like a bat out of hell, but when we try to ride it, we can get rolling and then after about 5mph it cogs, if we try to give it any throttle it cogs. And now after trying to work out this cogging the power switch seems to not be working anymore.

I’m tired of throwing money at this project so I’ll just use a loopkey or something. I can’t keep buying esc’s.

Here’s my screenshots of my settings. Please take a look and make suggestions

[Alpacaslapper]

[Alpacaslapper]

[Fearless562]

Try changing this from 4000 to 8000

[Fearless562]

Also your motor inductance looks to be abnormally high

[Alpacaslapper]

@Fearless562

Also your motor inductance looks to be abnormally high

What is this an indicator of?

[Jley]

One possible scenario is that it would be indicating a catastrophic lack of a skateboard in the equation.

[Alpacaslapper]

Correct, do larger motors have higher inductance naturally?

Also I just got home and noticed the esc was on, so I hit the power button and it shut off. I hit it again and it came back on, but there is a delay after I turn it on I didn’t notice before, maybe ,5 seconds, then a delay a few more seconds before the throttle spins the motor. At least it’s working I suppose.

[TimR]

How sure are you that the motor worked properly with the first ESC? And, then the shorted FET didn’t cook the motor which is why it now doesn’t work?

Also, I doubt it’s just bad luck a FET failed. There’s probably a reason, so will happen again.

I’m under the impression a lot of the entry level chinese market motors have rotors that don’t produce the best sineosoidal back EMF pattern for FOC control and my bets that the controllers they ship with are programmed especially to compensate for deficiencies, and use basic BLDC / Trapezoidal control at best.

My mates had a shit time with a Kelly controller. Maybe VESC is better though, if you can get it dialled in right.

[Alpacaslapper]

Well it’s done now lol, I got it running, I raised the sensorless erpm and raised the hall sampling, and turned off the motor temperature sensor because it doesn’t have one, then I rode around the yard, the throttle felt sluggish so I played with the throttle settings, I also turned up the motor amps, rode around felt good, tried to hit a wheelie like on the old votol…poof

Smoked it, actually it was on fire, lucky it wasn’t too close to the battery.

Sigh, there goes another 100$

[Alpacaslapper]

Hello my name is brad(hi brad) I’m an addict, I can’t stop blowing up esc’s

l

[Evwan]

it’s possible your motor’s cooked, but I’ve had trouble getting vesc to play nice with big motors at times. It’s just pretty hit or miss. I’m trying a different hub motor on my ebike to see if it wants to behave this time

[Alpacaslapper]

The motor seems ok, no issues from one phase to the next or and shorts that I can tell, how can I further test it?

[TimR]

I’d look inside the motor, check for winding discolouration and weakened magnets.

The only way we got a Chinese inrunner to work for my mate on a Kelly running FOC was to hack some of the steel laminating out from between the magnets in the rotor.

If we ran a hall sensor over the rotor and looked at it on oscilloscope, we would see the result of fringing, causing the control to get messed up. Grinding some metal out helped, but it’s still not perfect.

In eskate we’re very fortunate to have the outrunners with nice sineosoidal back emfs.

I would still say you might have better luck running a trapezoidal control than FOC on said motor.

My crappy drawing is rotor construction and the hall sensor outputs we were seeing.

[Alpacaslapper]

Well… It’s back together again, with the newer 84200 unit, let me tell you, soldering 6awg wire into the qs8 was a bitch! The bullets I put on the phase wires were ok because I used a torch but you can’t do that on the qs8 due to the housing.

Here it is back together. Now to tune it without blowing it up again.

[mulletboardermobbin]

@Alpacaslapper

soldering 6awg wire into the qs8 was a bitch!

Need one of them Big Boys

[Alpacaslapper]

Lol that thing is crazy. I used two 40w irons, one on the wire one on the cup, fed extra solder in with my teeth lol it was quite the experience haha

[Alpacaslapper]

So, I set the battery current to 100a, motor current to 100a. Both should be well below the threshold of the controller, I’m not sure if it’s rated for battery or motor amps on the manufacturer specs.

I’m uncertain the specifics of the votol programming that I was using prior, it was poorly translated and not user friendly, but I was running 12s8p at 80 battery amps it ran great. Supposedly the votol was rated for 72v but when I added the additional battery it blew fets. All the votol tuning information I could find from the manufacturer and user tutorials said to not touch the preset phase current and I believe it was set to max value on that controller, the value was 9000 so not “amps” I assume. But I assumed vesc 100 battery amps at 20s should be more powerful. I could easily pop wheelies at votol’s 12s8p 80a, I can not pop a wheelie with the vesc at 20s8p 100amp. The acceleration just kinda builds up, there is no instant torque like the votol. I am adjusting the throttle mapping and reducing the ramp rate all the way down and will test again today. But I’m leaning towards the votol phase amps being wicked high whatever the “9000” meant, which is also what I am attributing to the blown fets at 72v. I have read other people have blown fets on this same controller the same way, so I think this model in the line up is not made with quality fets as their higher rated units.

Now back to the current issue

It cogs what I believe is at the switching erpm from sensor to sensorless as that seems to clear it up some when adjusted higher, what would cause this problem? Why is it not smooth switching at low rpm? Also this motor only spins up on the stand in foc, if I try any other mode, it only stutters.

On the stand with the sensorless erpm set to 12000, if I give it fast aggressive throttle it kinda still cogs for a split second then runs like a beast all the way up to the moon, I think unweighted the mph went over 100 last night at full throttle, lol. I did notice a little too much slack in the chain, maybe after I adjust the slack a bit that will smooth out some though.

Currently I’m playing around with the sensorless erpm to keep it low but high enough to not stutter, more ride testing needed but I see noticeable difference on the stand as what appears to be power delivery depending on what the sensorless erpm is at, lower seems to have more power but cogging, higher is smooth but less powerful, again that’s to the eye, unweighted, ride testing to come.

Side note, my throttle is reading 2.6v when closed, 3.4v max, so that’s odd and I will replace it.

So today’s testing:
Chain tension
Throttle ramp and curve adjustments
Sensorless erpm adjustment

If anyone knows if the makerbase specs for amp capabilities are battery or motor (I think motor)
Let me know I don’t want to blow anything, the battery is build with .2x30mm strips and 2x 8awg silicone wire for series connectors and 6 awg terminal leads, what I would comfortably rate upward of 250 amps. The motor came back from the vesc readings at 100a

[Soflo]

Try manually raising your motor inductance R up. Don’t go over 100.
Id go 10 at a time

Also try manually lowering your inductance L. Don’t go below 190

Certain hub motors benefit from this, some don’t. Worth a try

[Alpacaslapper]

So the bike topped out at 25mph while riding, something is wrong here. the foc sensorless erpm needed to be ridiculously high in order not to cog, like 15000. I’ll play with the inductance and see what happens

What I don’t understand is, why won’t this motor run bldc or sensorless at all