Ground Up E-Freeboard | Custom Control Algorithm: Dual VESC | Dual, HUB 2200W 150KV Motors | Work In Progress |

Hey Kids,

Intent: This marks the beginning of a multi-month ground up build. I will be attempting to create a electric freeboard almost entirely from scratch. I will periodically post my progress here, as well as links to relevant articles, my CAD files, and my motor control Code.

Background: I am a Senior in Electrical Engineering at Purdue with a heavy background in parts fabrication and controls. This summer, I am interning in Detroit, and my company (omitted), also encourages personal projects that double as learning experiences. Thus, inspired by the LEIF board, an idea was born. What if I could completely design, fabricate, and program an electric freeboard from scratch for under $1000?

Thus far, I have CADed a proof of concept, begun my second design iteration, have started performing calculations for desired Motor Spec, and have created a (pseudo)-working control algorithm for a 3-phase brushless outrunner I had lying around.

Lets start with my proof of concept.

Before you say anything, I know that it looks pretty jank, expensive, and overly complicated. The intent was to demonstrate a working design to get funding before moving on to the next couple iterations.

Here are a few more pics:

This design featured off the shelf parts to enable the wheel pods to rotate, mostly a 200lb lazy susan, and a 3 lead, 30 amp slip ring which would allow the passage of current to the pod, without having to worry about leads getting twisted. Additionally, it made use of two, 18" Carbon fiber rods and a series of belly pans to create a superstructure allowing battery mounting and mounting of the control electronics.

This design has several problems. FIRST: It’s expensive, It’s heavy, and overly complicated. The carbon fiber alone would cost about $170, which is too much for (at best) a medium amount of flash. The Aluminum wouldn’t be too expensive itself, but as i no longer own a waterjet, I would need to pay someone to waterjet it for me, which would not only cost more money, but would also dent my pride :wink:

Furthermore, 30 amps is not nearly enough to drive these motors, which usually are happy around 60-70 amps each. I looked around online, and couldn’t find a slip ring that would fit my design for less than $750 each. So again, no go.

So, scrap that. She was pretty, but not feasible.

This brings us to my second design iteration.

This design features a custom-made, combination slip-ring lazy susan and a wheel mounted directly to the exterior of an outrunner. As seen in the first picture, the lazy susan comprises three stages, each electrically isolated from the others, which doubles as a slip ring, allowing individual currents to pass through each of the three “phases” to the corresponding phase of the motor.

This design has not been flushed out yet, but that’s kind of the point of an ongoing blog right?

On to controls.

I plan to utilize an arduino uno to pass PWM signals to the gates of a standart MOSFET array, allowing current to flow into each phase of the motor in time. A great synopsis of this, as well as the control theory behind brushless motors, can be found here: http://electronoobs.com/eng_circuitos_tut4.php

I actually plan to follow his process quite closely, finally diverging in the speed control stage. Rather than use a delay between phases (which will result in somewhat choppy feel) i will use PWM control to modify the duty cycle of the motor, allowing speed to vary that way.

Thus far, I have built my array, and flashed my arduino.

However, When i execute my code, i have no way to determine the position of the rotor, resulting in a high level of shoot-through and reverse driving. I’m overworking the FETs i have, and am not able to produce the smooth drive action i want.

Thus, as any electrical engineer does, I need more sensors.

I plan to buy two motors with hall effect sensors that will allow me to monitor the precise position of the rotor relative to the stator, and create a much smoother motor drive.

However, before i can do that, I need to buy motors, and before i can buy motors, i need to pick them.

I am thinking about running a 4 cell set up. The batteries i have are 61 amp hours, so they’re massive, and can push any amperage i could ever want. But they’re big. Thus, i will use a higher Kv motor, at a lower voltage to achieve similar results to those i would get with a lower Kv motor, and higher cell count.

I have run the math, and it looks like i should be able to hit about 25 mph with 4 cells and two 2000 watt 270kv motors. It’s also noteworthy that I will be using 78mm casters.

I can’t obtain any torque curves for these motors, but i think that they should work. Any feedback from people with hands-on experience would be great.

That’s all for now!

32 Likes

subbed! I guess you already know leiftech’s electric freeboard?

1 Like

Yeah, But the leif board is $1600, I’m shooting for a faster one for under $1000

4 Likes

This is great, definitely watching this one!

Yes there is LEIFTech eSnowboard and they have patents over this kind of stuff so be aware of that. But I doubt they will do anything to stop your build unless you start selling the boards :slight_smile:

1 Like

Thanks for the info!

I was thinking about picking up trucks from Freebord, anyone have any experience?

1 Like

and you should look at this thread: https://www.electric-skateboard.builders/t/nowinds-first-commercial-board-leiftech-e-snowboard/23571

2 Likes

I ride a freebord (not powered) and those bindings are great! they lock your feet in when you need it, but you can easily bail out if needed

3 Likes

Great project! It’s about time someone did this. I still want to see an all terrain version; if that would even work. :wink: Just throwing that out there.

3 Likes

I guess you can do At but the question is how it would ride. Maybe I’m wring but I guess the pneumatic wheels will like stop the drift or act weird. But a very interesting idea indeed! :slight_smile:

1 Like

Few weeks back I was watching a nice video from Mellow Boards explaining how they drive their motors using FOC and TI Piccolo micro controller with patented algorithm to detect position of the rotor without using any sensors which are in their words just another point of failure :slight_smile: Maybe of interest to you and your build.

I know I learned few things from that video :slight_smile:

3 Likes

Hey! Thanks for the link. The way they do that is via the back EMF on the unused coil. When a coil of wire moves through a magnetic field, it generates electromotive Force (EMF) in accordance with faradays law. So like he said, as you power two leads, the magnet turns, moving the field and changing the flux in the third coil. You can actually read the voltage on that coil to determine the relative positions of the rotor. However, because I am using an arduino, I’m limited to 4 milliamps of current into my analog pins. That’s 0.004 amps, whereas that back EMF generates about 1,000 times that. This means that i would need to use a current division circuit to divert most of the current to ground. However, even then I run the risk of picking the wrong resistor values and burning out my board, or a resistor coming loose, and burning out my board.

On the other hand, I could spend the extra $6 for a sensored motor, and get a more precise, more reliable reading of where the rotor is. Also, I’m hoping to go into controls as a career, so the more learning I can force, the better.

Just my thoughts :slight_smile:

2 Likes

@Mobutusan, I think the best way to do AT would be to replace the casters with pneumatic, which would increase your ride height significantly. Not sure if that tradeoff is worth it. But hey, follow your dreams!

2 Likes

This is awesome! I have dreams of doing this myself as I currently ride a non-motorized freebord myself (on that walk back up the hill, the mind starts to dream…). One tip, learn how to ride a non-motorized freebord if you haven’t already to get the feel for it before you use your electric one. I am tracking this thread and it should be awesome! Good luck!

1 Like

Update 2

Today I created a google drive where i will post all my math, code, and CAD.

https://drive.google.com/open?id=0B0zQsjKfE5R7YVlOYko2S1ZCNmc

Knock yourself out.

(I found Code from Electronoobs (Link found in original post), and found the Caliber 50 trucks on Grabcad.com.

I ran some more motor calculations, my target being the ability to drive a 300lb person 30 mph up a 5% grade (yay production criteria!), and it looks like I’ll need to run two 2200 watt motors to get the torque. I haven’t settled on a cell count, so I haven’t picked a Kv either. As of now, It’s looking like I’ll probably make my own 6S4p pack and a motor around 120 Kv.

I posted my Math in that folder, so you all can run your own calculations from the excel file. Just Change the green criteria, and everything else will adapt. Weight is in Kg, Diameter is in mm, Grade is in degrees, Volts are in Volts (go figure), and Kv is in whatever the heck Kv is in.

Does anybody have any experience with Alien Power Systems? I was looking at one of their 5065 motors, and was wondering how long they take to arrive in the U.S. Can’t really afford to wait 10 weeks with my schedule.

I will probably use the VTC6 in my packs, build my own protection circuit, and charging inverter, pending approval from my boss.

Going forward, I will be creating a target production schedule along with design criteria, part spec, and deliverable milestones on certain dates.

After this summer, i will likely build a second iteration on my own, and it’s gonna be shiny af (firefly references ftw).

Also looking into building my own deck, likely with clear grip and racing stripes. I have a lazer cutter, so i can etch a dope design into the bottom, likely something Pink Floyd related. More than open to suggestions!

Later Today, I will continue to flush out design specs, and will (paycheck pending) order some test parts. I’ll also try to make another dent in the second Iteration of CAD.

Here’s a Mock-up of everything i have so far.

Truly Yours,

VillainousJ

Afterthoughts: Since I’m targeting 30mph, I should invest in some cool helmets and body armor :wink:

7 Likes

Wow, awesome. Looks like you really thought it through. I especially like the slip ring design, ingenious! And looking at the calculations, low kV and low cell count, it seems like you trust it will push quite a current. Did you do any testing yet?

No experience with Alien here but DIY Electric Skateboards has some, higher kV though. Or maybe search HobbyKing for some SK3 motors.

@SilentException I’ve looked at the SK3 motors, and they don’t really have the KV I’m looking for. Because the motors will be direct drive, there will be no gearing, so the Kv will directly affect the top speed.

As for the current, The cells are rated to push that continuously, and even a bit more burst, so everything should be fine there. My biggest concern is the testing of my PWM algorithm, and other high level motor controls. I’ll probably run the algorithm on a low amperage motor first, then fine tune it on that scale, before upgrading my MOSFETs and jumping to a bigger motor. Alien is shut down for a few weeks, so i have some dev-time.

You might want to look into rewinding the motor to change the kV. Check this out: https://endless-sphere.com/forums/viewtopic.php?t=36767&f=30

Yeah, the batteries are solid. I was actually asking about the slip rings and their current transfer capabilities. I really love the design and looking at it, it should work but real world testing is many times a different story :slight_smile: For example will it handle the rough and the bumps and still provide nice and clean signal for motor. But man, this will be a sweet build, fun times ahead :slight_smile:

Awesome build. I thought arduinos were not well suited to be used as Brushless motor controllers. Can they handle 60k erpm like the VESC?

@pantologist The arduino clocks around 12M Hz, should be more than enough to do the PWM to the FETs. That being said, I haven’t done this before! I picked an arduino cause i had one lying around. I checked online and noticed that other people have done it before. Plus i only need to run my motors around 6k.

@Silent Exception, Yeah! I’m excited. I’m from a family of 5 engineers, so i have a couple people to bounce ideas off of. I also get a ton of great input from the guys at work. It’s a good time to be alive. As for your comment about the slip rings, If you’ll refer to the picture of my CAD, those little black pieces are T profile delrin runners with a bolt running through them to the bottom plate, I’m thinking you can tighten them down to keep the top and bottom of the Lazy Susan together.

@NickTheDude Doesn’t Soldering wires together to cut the Kv create a ton of leakage? The fact that one lead will hog most of it also creates a ton of imballance. Beyond that, Don’t maxwell’s equations play in to create a bunch of eddy currents? I think eddy currents would be one of the last things you’d want in a PM motor. I only skimmed the article, so maybe I’m missing something