**********As of August 18, 2017***********
After getting my ass kicked in college for a few years and finally finishing up my physics degree, I am about to start an actual engineering education in parallel with stepping down on the metaphorical gas pedal of this project.
First and foremost, I am trying to build a small community around this project. There are many areas where an extra hand can help tremendously, therefore I am publishing all of my progress on my github page: https://github.com/lolomolo/LongboardMarkII
P.S. I'v posted some pictures of what I have made so far over the past year or two.
After gaining some confidence from a working first prototype (see Mark I) I am making some major upgrades. Inspired by the erganomic design, but daunted by the price of the Boosted boards, I have decided to explore the full potential of the electric longboard.
- This is not a complete guide as the board is still in progress.
- This project has (and will have) a lot of parts to it and will take an estimated 4 months to complete (mainly because I dont have too much free time due to classes).
- The vocabulary and terminology I will use wont make much sense to you unless you have done considerable digging and research on your own.
- This project is expensive (although hopefully still less than a boosted).
- Do everything at your own risk, this is only a general guide, not meant to be replicated.
- Basic electronics know-how
- Circuit Theory
- THIS BOOK!: Practical Electronics for Inventors by Paul Scherz (its a fun read and got me started from absolute 0)
- An Oscilliscope (you can find 20Mhz+ scopes on ebay for ~$50, this is more than enough)
- A soldering iron (Look for one that you can set the temperature on)
- PCB CAD software (eagle is good enough)
- ~$500 to work with
- Basic Cpp programming (for arduino or other micro-controllers)
All of this took me about 8 months to learn. Before last spring (~march 2015) I had close to no knowledge of electronics, I barely knew what a capacitor was. (I will make a list of all reading material that I went through)
Again, I anticipate for this to be used as a general guide and not step by step instructions, I will try to brush up on most of the mistakes that I made, and mess-ups that you can avoid.
Step 1: Breaking It Up
Since this is such a large project, the most practical and efficient method to tackle it is breaking up into parts... and those parts into more parts.
From the start I split the project into # parts:
- 3 phase half bridge (since the motors I plan on using are 3 phase BLDC)
- Gate drivers
- PWM control
- I opted for LiFeO4 batteries (lots of power per volume/weight, and they maintain capacity over many more cycles than standard LiPo batteries)
- Lithium batteries need a battery management system (BMS) that will make sure that each cell charges evenly and none of them are over-discharged.
- We need a charger that can output significant power (looking at max pull from a 10A wall outlet which is 1200W)
Receiver/Transmitter (the remote)
- RF communication
- For now I will use the chassis from the Mark I board.
- Different chips need different voltages so we need a converter that take the ~40v from the batteries and output constant 3.3/5/12v
- Buck Converter
- Head lights for night riding
- USB charging port, cause you never know when my your phone gets low
- Underlighting, cause why not.
Step 2: Bare Bone Motor Control/Power Subsystem
The first thing I did was to make the bare minimum to get a single motor running.
I designed a Power Bridge module that contained 3 buck converter ICs. I designed it to output 3.3, 5, and 12 Volts but after designing and ordering the PCB I change some of the chips I planned to use and no longer needed a 3.3v output. I just didn't solder on the third buck converter.
The 3 phase half bridge used the most powerful and (reasonable priced) mosfets that I could get. Those turned out to be the IRFS7530. I added another stock diode across each body diode (the intrinsic diode in a mosfet) in order to handle the emf that might result from sudden stops of the motor. Also I added pull-down resistors at each gate, however this was a mistake. The gate driver has a built in pull down resistor for each output, and the external pull-down resistors caused interference in other gates and made them turn on during high pulses of current from the gate drivers.
I made the PCB design in Altium designer but have since switched to Eagle because I didn't need the vast amount of tools and complexity that altium provided (its overkill). Eagle is free and I found that SparkFun offers the most comprehensive tutorials.
I had the PCBs manufactured by a website called pcbway. They had the most friendly user interface out of the manufacturers that I came too and turned out to be the cheapest and one of the quickest. They printed 5 copies of both the power bridge and buck converter for 26$ plus 25$ shipping, and delivered in 5 days. The quality of the boards was also on par with many of the local manufacturers.
Step 3: Testing - Gate Drivers
After testing the gate drivers (NCP5181PG) using some random fets that were laying around and shorting/burning and then repairing the buck converter multiple times, I made a circuit to test the full 3 phase bridge. I used the arduino nano to send a 20khz pwm signal (using the PWM1 library) to an inverter (effectively shifting the phase 180 degrees). I used the inverted and none-inverted signal to control the gate driver. However the problem I ran into was that I was getting about 4ns of overlap and since this gate driver didnt have any shoot through protection I was opening both Mosfets at the same time for that 4ns. I figure that wasn't very healthy for them... Although im sure they are fine because they are rated for 1500amp pulses at <10ns. Nevertheless this is extra pressure on the system and a lose of unnecessary power so I ordered a faster inverter and gate drivers with shoot-through protection.
to be continued---
(I am currently learning about ARV programming because I want to move away from arduino and have more control over my electronics).
In the next update I will probably have replaced the arduino nano (which I actually killed a few days ago) with at atmega32.