How hard would it be to build your own hoverboard?

File: a jolly wanker2.jpg (15KB, 400x400px) Image search: [Google]

15KB, 400x400px

pretty hard

actually i was thinking about this, maybe buy a broken one and use the controler off of it and just put in a proper motor and batteries.

i wanna see someone put one of these in there http://www.hobbyking.com/hobbyking/store/__25413__Turnigy_RotoMax_150cc_Size_Brushless_Outrunner_Motor.html

>>920852
depends what you mean by "hoverboard".
if you're thinking about something on wheels, then it would be pretty easy.
building something that would bend the laws of physics on the other hand would be difficult

>>920877
talking about a self-balancing scooter.
was poking around and found a parts list here:

http://bestelectrichoverboard.com/hoverboard-faq/how-do-self-balancing-scooters-work/

Step 1: turn in your Man-Card
Step 2: become massive faggot

From there on, I'm not sure, I imagine a bunch of electrical wizardry.

>>920852
Lots of tuning of loops.
Built a self balancing robot in college about 7 years ago, was pretty easy with a group of 4.

>>920865
>150kv
>50 volts
>5" wheels
>100+ miles per hour

No thanks.

Hardest part would probably be the gyroscope.

File: motorized_skateboard_wheel_fj328rj.jpg (107KB, 1221x575px) Image search: [Google]

107KB, 1221x575px

There's places on aliexpress that sell all the parts you'd need.
The gyro could just be a arduino + 9DOF board.
I know there are guides online for making balancing arduino "segway" style robots (with only two lateral wheels)

Pic related: found these motorized skateboard wheels while browsing once. I don't know if these are the same motors used for the hoverboards or not.

File: Change_with_Ki.png (8KB, 465x367px) Image search: [Google]

8KB, 465x367px

>>920852
Try learning PID controller first

then we are talking...

File: Darth Vader v Hoverboard.webm (2MB, 406x720px) Image search: [Google]

2MB, 406x720px

>>920852
Not very difficult if you know anything about electronics. Most of the stuff you need is like putting together legos.

>>921078

There wouldn't be any saving over buying a chink import then.

>>921128
For as cheap as these things are you ain't going to make one cheaper in your basement.

>>921075
The chink versions don't even use a gyro. It's a clever short circuit that achieves the same result.

>>921149

Yeah exactly. And then once you Factor in time and quality.

Just get a chink import OP and get bored of it before you waste your time and money making your own.

File: 12321323.webm (166KB, 371x209px) Image search: [Google]

166KB, 371x209px

>>921154
But if the OP learns how and makes his own he'll have tons of knowledge to make other cool things.

>>921158
So make something that the chinks haven't already done cheaper?

>>921167
it's kind of just a hobby off to the side to keep me busy between classes.
im an EE student in ieee so the electrical stuff isn't too bad for me. was just wondering if anybody actually has done it before/knows what it would cost.

>>921152
>It's a clever short circuit that achieves the same result.
What do you mean?

not as hard as inventing a sideways hoverboard

>>921531
The original kickstarter version used a gyro.
But, the gooks figured out a feedback loop that could replace the gyro, making a cheaper product that (seemingly) works the same.

>>921167
>You shouldn't make anything some else does better.
Are you sure you didn't mistook this board for /g/?

>>920852
>hoverboard
>posts skateboard

so much of >>921084

>>921813
A skateboard is better than that piece of crap. You can't clear a curb on that thing.

>>920852
I dont understand why anyone would want a segway sans handlebars when you can get/build a motorized skateboard for far cheaper, which goes faster, and doesnt make you look like a tool.

>>921859
>electric skate board
Five years ago my neighbor wanted to make a motorized skate board.

I helped him with the circuitry and everything. He of course wanted it to go insanely fast.

Three weeks after he finished it he was out riding it around flying down the street and hit a rock. Wasn't wearing a helmet and slammed the back of his head into a curb.

Now he has an electric wheelchair and a drool rag. I kept scolding him to wear his fucking helmet but never did. I had even told his wife to hound him about, she told me he'd be fine. Now that she divorced him, his mother moved in to change his diapers and bedpans.

I get so infuriated when I see people not wearing their helmets.

I want to build something similar, does using 3 phase BLDC motors complicate things? Most of the balancing robot tutorials I can find use stepper or servo motors.

Is it just a matter of writing a balance program for a MCU + gyro and having it send commands to the BLDC controller? Or does it need complex current and hall sensor information from the controller?

All the chinese ones share a MCU for the balance control and motor driving so I can't tell how tightly integrated the two functions are.

>>921935
>I helped a man cripple himself and it makes me SO! ANGRY! AT! STRANGERS!
cool story bro

>>922123
You can get BLDC motor controller boards with PWM input.

>>921813
>>921839
Still, this, why the fuck are they called hoverboards? They literally don't do what they advertise.

>>922165
The BLDC control is easy, what I don't know is if the balancing algorithm only uses the gyro or if it needs to be integrated into the BLDC driver.

>>922174
Call them what the chinese do, shilly cars or electric wheelbarrows.

File: shock.gif (1MB, 330x271px) Image search: [Google]

1MB, 330x271px

>>921120
>legos

>>921935
They make it difficult to shower effectively though.

>>920852
You'll need to drive 2 BLDC motors independently with Hall sensors for proper commutation sequence. You need an MCU that can drive 2 BLDC interface chips which in turn drives 2 three phase inverters. The original used an ST F1xx series ARM to interface to the BLDC drivers and read the LSM330 IMU. You'll also want to look into LiFePO batteries and really understand their characteristics. We upgraded the microcontroller to an ST Cortex-M4 but I don't know what made it into production.

>>922248
Can you tell me more about the software?
Does the balancing part just work using the gyro? Or does it need data from the motor (hall, current, RPM) to balance?

>>920852
>How hard would it be to build your own hoverboard?
Quite hard. Infact near impossible

>>922248
What did China do to make them explode at such alarming rates? They're still really not that popular to justify all this. Everyone's drone isn't catching fire.

>>923131
pulling too much current from batteries that cant handle it, or draining them passed the point of no return.

>>923122
Honestly, the simple answer is: the entire project is built around a reference design and software provide for free by ST Microelectronics on their website.

It's a couple of PID loops monitoring the lean angle of both footpads and the rate of change in lean angle. The commutation pattern of the BLDC is handled by a special driver chip and the MCU. You do need Hall effect sensors to detect when to switch phases on the BLDC however. RPM isn't really a relevant quantity since the output of the final PID loop generates a term that translates to a torque vector. I'm other words, the controller doesn't care how fast the motor is turning and doesn't seek to directly control RPM. Rather, it modulates the field strength from coil to coil as the motor turns. The application and outcome is deceptively simple for how complex the math is.

Look up BLDC drivers on ST's site and be prepared for really interesting math.

>>923131
The knock offs and, unfortunately, the original all have a couple main issues with the battery. First, in the original, it's a lithium iron phosphate battery which is unlike lithium metal hydride or lithium polymer. LiFePO can handle 100s of amps of pulse current discharge but it must be kept within strict operating temperature limits for charge and discharge. So one issue is that the pack has no charge or discharge limit by design because the motors can draw huge amounts of current. What happens is the pack generates gas (not hydrogen) and blows its safety vent. This is not harmful in itself but without temperature monitoring the external charger will eventually overcharge the batteries and the loss of electrolyte causes cell failure. Second, the manufacturers cheap out on battery packs and aren't using matched cells or a charger for LiFePO and the required cell balancing method. If I remember it is a LiMH charger which has a very different charge curve than LiFPO. So the combination of cheap cells and POS charger cause premature cell aging and excess heat generation. I can't confirm this but I suspect that the Chinese copies are not using LiFePO cells but the cheaper LiMH or LiPo cells which absolutely can't handle the 50 or 75 amps of surge current under heavy acceleration. They also aren't suited for regenerative braking.

>>923141
Yes, to put it simply. Lithium chemistry batteries are tough if charge and discharge are handled properly. LiFePO is still fairly new so there aren't a lot of charger topologies to support the various implementations like LiMH or LiPo.

Is there a teardown out there done by a sufficiently tech-minded individual?

>>923159
I haven't seen one yet, pictures would be helpful though. I don't have the photos from the early development stages anymore.

>>923159
There's a few teardowns online. I'm not sure how detailed you want it.

>>923151
Just watched a teardown of one.
The batteries are charged UNBALANCED. No wonder these motherfuckers are exploding.

>>923159
They are fairly simple.
The primary board consists of a STM32F1 MCU, 2 BLDC motor control circuits. Secondary boards consist of a cheap MCU, MP6050 6 axis gyro and a pressure sensor for each side. The two control boards send gyro and pressure sensor data to the main STM32F1 board.

>>920852
dont build your own they are pices of shit

>>923159
Hopefully AvE will get his hands on one of them soon. Got my dick ready and a bag of ice in the freezer waiting for it.

>>920852
It will be extremely painful

>>920865
10 fucking kilowatts in something 10CM in diameter.
What in the fucking shit would use something like this? A fucking smartcar?
Seriously though, you could build a fucking go kart out of that shit. A go kart with 10% the go of a fucking car

>>925755
>Abusing SI prefixes and unit symbols
Please re-read the SI Brochure.

File: Smart_electric_car.jpg (347KB, 2592x1944px) Image search: [Google]

347KB, 2592x1944px

>>925824
Kilowatts and cm, happy?
Anyway, smart electric drive cars are only 20 Kilowatts, two of those motors could build something as powerful.

>>925855
kW and cm bro

>>925856
Fugg me I'm dum
Thanks cleverAnon, I shall remember this forever.

File: untitled_key2.jpg (13KB, 271x296px) Image search: [Google]

13KB, 271x296px

>>920865
A hoverboards' rate of acceleration is determined by the rider's lean angle:

tan <lean angle> = <acceleration> / <gravity>

The rider would have to lean to a ridiculous angle to make use of a 10kW motor's power! Practically, you would be limited by rider's skill in their ability to lean in a controlled manner. Most hoverboards use 150W motors which are adequate for most riders' skill. How much power could a pro rider use? 500W? No where near 10kW.

>>925855
>'Kilowatts'
>In the middle of a sentence
You didn't read the SI Brochure ;_;

>>925880
>No space between the quantity and the unit symbol