Current state.

In the early days of brushless DC motors, I developed a mild need to build an electrified bicycle, since it is convenient in a city to transport kids and smaller goods. My finalised utility is here. The only maintenance it needed for the last 5000 km was occasional rainwash. This thing has all-wheel drive mode and tank turn capabilities.

Attempt one

Let's step through a complete chain of failures set by the initial goal. Since my pockets were not very deep, I had to think about how to get things done under a nonexistent budget. My first motor was a car generator. I drilled out a regular electromagnet and replaced it with neodymium magnets. Now it can be called a motor. And then I fitted a regular bicycle gear on the shaft.

This setup has three problems. The magnetic field was weak, heavy, and had no freewheeling (directly connected to the pedals). Since BLDC wants switching in electronics, I had to wire hardware and write code to get the motor to spin. Battery technology was not there, and I had to use a second-hand, super-heavy, regular 12V car battery. To cut BS, this thing was not convenient to use. I accepted losses and paused. Procjet state was recorded here

Attempt two

The second attempt was made when I could buy an el-cheapo BLDC controller and a skateboard motor. The idea was to print out the nylon gravity pusher and the drag front wheel. I wanted to get around 50W of power transfer from it. Unfortunately, it started bouncing on the wheel due to irregularities. Then I pulled it towards the front wheel, got my happy 2 km and blew the controller. Luckily, my self-made battery survived because I did not forget to install a proper fuse.

Attempt three

Since pockets had grown deeper and the world around me had evolved, I managed to buy my first "proper" hub motor and controller. It turned out that starting from a standstill (hall sensors) and using an electronic brake were essential, mainly because speed (weight) increased, and brake pad and rim wear were significant.

I upgraded batterys and got around 120 km range out of it.

Then, after a significant number of cycles ( around 17 years). Bikeframe started to show its age. Initially, the front fork fell off. Then I discovered small cracks on the seatpost.

I repaired and replaced everything, but the ageing issues did not go away. And then winter happanes - again.

Final attempt

After another reality check, I started to plan my new two-wheeler. Requirements were:

1. Maintenance free
2. Robust frame, steel fork
3. Comfort riding position
4. All wheel drive, e-brake
5. Bicycle, by its nature

First, I considered recumbent bikes, but I gave up because of their rarity.

Then no one will sell me a decent quality diamond-framed bicycle with a belt drive. High-end bicycles were built for racing. Regular city bikes were built to compete on price. It took half a year and countless rejections for custom-built bicycle manufacturers until I found 99% what I wanted. The only thing I had to do was lace the motor myself. Build a battery, install a front rack for it ( no bending wires) and hope for the best.

I decided to stick with my proven e-bike controller provider. I accidentally purchased a lower-voltage-rated controller during the COVID supply turbulence. I managed to get it to run, but just in case, I contacted the seller since the setup was too hard for a kit. The company owner wrote back to notify me that I have a 50% chance of blowing up MOSFETs — yet another controller swap. Then I discovered that my new battery BMS was weak. This one was luckily fixable in the e-bike software by doing a 0. something seconds softer start. After that, I had to build a PAS sensor inside the gearbox. The initial idea to let the belt drive the PAS looked clumsy. But there was a way to fit everything inside the gearbox's removable cover cap. After numerous failed prints, I got it done.
Now the final touch was to set up reasonable limits for this beast and hit the road.