i gotta see how it’s done
cause i can set to bldc but there’s no way to run parameters… unless imm dumb and blind
and they removed it from the wizard
which isn’t bad but it’s sad
i must find a way
well there’s the non wizard detection pages in the app.
1 Like
i’ll mess around on friday, hope i din’t blow something up
2 Likes
You have to do a manual motor detection to use BLDC, don’t use a wizard
yup, if figured
i want it
1 Like
been thinking about it being rpm and not erpm since skyart wrote this.
kept thinking it’s got to be ERPM because that’s related to the number of times the magnetic field changes
And this seems to concur.
Its a lot more complicated when comparing between different motors.
But if we’re just talking about the same motor (regardless of what motor), then iron losses increase linearly with rpm or erpm. Erpm just being rpm * pole pairs which is a fixed value for any given motor.
1 Like
I thought eddy currents increase exponentially with RPM and hysteresis is linear?
2 Likes
Yes you are right.
Copper loss also increases with the square of current, so making lots of power from a small package gets pretty difficult.
7 Likes
oh all this time i thought eddy current loss = hysterisis loss = iron loss.
and I thought they were all the energy lost reversing the magnetic field. which i now think is just the hysteris loss. that one i understand more than the eddie current loss. more stuff to look into.
I now think it’s eddy current loss + hysterisis loss = iron loss.
2 Likes
Yep and a lot of the losses come from higher harmonics which is why simulation and good design is very important to minimize these harmonic losses. The switching frequency and shape of the sine wave from the controller also contribute to harmonics and iron loss in the motor. Even the switching characteristics of the particular mosfets used in the controller has an effect on the losses in the motor. You can go pretty deep with this stuff!
7 Likes
I cannot recommend this channel enough. Most of the things I’ve learned about the systems we use in esk8 come from RCexplained. He explains everything and walks you through it with real world calculations.
Brushless Motor Copper Losses and Iron Losses Explanation and Calculation - YouTube
3 Likes
Wed 2023-10-11
another secret sesh.
course was too flowy I again didn’t get up to punching hard out of corners. maybe tomorrow at Intro2Speed.
214 max motor amps. 97 max battery. (/2 motors). but the board feels great at these paces.
runs were short so temps remained under control. < 70c
@Ruxx and @tuckjohn also rode the board. (didn’t get Tucker’s time, but we did get a time on his board)
| v5#1 (vesc) fess | AnkleWreacher @tuckjohn (vesc) | v5#3 (castle) fess | v5#1 (vesc) @ruxx |
|---|---|---|---|
| 33.23 | 33.54 | 33.68 | 32.91 |
vesc tool telemetry
i haven’t used this much before. in favor of metr/robogotchi.
about 4kwatts a side max. so as i said another day of not pushign it hard enough. i feel i’m failing in this testing. but… the board is fun. lots of fun.
I’m also confused why these two have different distance traveled and effeciencies. soemthing seems off there.
and on a broader note @HAIRYMANJACK posted to instagram story (since deleted) a 14kw power max screen shot. and we were discussing wether that’s dual or single. I thnik vesc_tool only shows per vesc numbers there so i was confused.
my trust that i understand what i’m getting from vesc_tool rt summary is low.
robogotchi telemetry
and the amps here (dual) on robogotchi, don’t match I Max in vesc tool real time at all. not sure what that diff is either.
throttle response
balance
I noticed the difference in fidelity of current control mode vs duty cycle because of those moments where you are using the throttle to balance you and lift you up out of a turn.
switching
Switching between the two is still hard, but i get better at it. Some of the Fidelity stuff still might be managing the throttle differently between current control vs duty_cycle.
range
I notice depending on how you’re using the throttle. current control mode gives you more range,
a really simplified model:
as you set duty_cycle you send watts. (whatever duty_cycle is) watts turns into rpm + torque
So, with duty_cycle control mode. if you’re at 50% rpm, 50% throttle. you have 50% more throttle to apply the remaining power to accelerate from there. however with. current control if you are at 50% speed and maintain you’re not asking for more current, so the throttle is further back and you have more range across the throttle to apply the remaining power to accelerate from there.
I think this might be an advantage you’d miss.
and if the fidelity can all be attributed to PID tuning. that might be better in the long rrun
Ruxx opinion
@Ruxx gave it a try for the first time. he’s used to the g2tb on castle xlx. so he could make a reasonable comparison. with he g2tb on vesc. He also described it as smooth, ( partially as a joke referencing a certain obnoxious person)
after that he said it was kinda like stick vs manual. with the current control being smooth like a torque converter taking away some of the fidelity and power. I can see this viewpoint.
again I think this might be eliminated with PID tuning which I haven’t gotten to.
updates for next time.
-
@Titoxd1000 mod for g2tb. done.
-
adjusted dead band down to 7% from 15%. @tuckjohn called out the wider deadband and I had been meaning to adjust it. .
Todo still.
- get some more power down. try to test those limits. (fear of faults setting in. @zero_ads not helping me here.
)
- drop fault stop time from 75ms to 50ms
- PID tuning
- learn it better.
- test it
- do no load thermals test.
- compare my erpm vs @tuckjohn’s think about iron losses more.
- watch more of the vid from @Shadowfax
3 Likes
welcome to the club 
what’s the pid tuning for?
1 Like
duty_cycle throttle (castle) seems to have better fidelity than vesc current control. (trying to rule out familiarity bias) strongly suspect this is a delay of the PID controller and can be tuned out. keep saying that haven’t done the experiments yet.
1 Like
I recall you faulting once on the setup I don’t recall if you figured out why?
Seriously lol, we will figure it out…. I hope. I don’t think I’m pushing as hard d as y’all are though. And it’s sad.
we’re gonna play every card we can 
1 Like
the inital one was unfair. no slow abs, abs max set too low. moved from 225 to 350/side.
after i had flutters for which i didn’t catch a fault, but I suspect i failed to check both sides properly.
also max motor amps was to 200 for both of those, (after hitting limits of 150/side and pushing for more) i set them back down to 150/side and simultaneously haven’t pushed the limits of 150/side since.
yeah you’re getting the skippy tires, which is different. shock loads. but to skip tires you must be pushing hard.
Ah I see, damn 350 per side niiiice. Wish I could set mine that high lol. Yeah you can probably start creeping it up again 
My skippy tires are just on launch though. Which honestly trigger control could probably help a lot to get rid of it.
@poastoast and I did talk about manually trying to shock the system. Lift board, spin tires, then drop them. Would that sock and induce current spikes? I don’t condone this test and not 100% sure what it would prove.
i’m unsure what it would prove either but it we could at least observe what’s goin on in a “controlled manner” since raising the nose and dropping at say 10 mph is much easier and safer than fuccin roun n finding out at say 30 gunning it