I finished installing the new controller in bike.3. It uses IRFB4110 Mosfets, those are semiconductor devices for switching large currents. It has a very nice feature; there's a three-way switch mounted on the handlebars that let's you change the amount of power you use. I programmed it to have 50%, 75% and 100% of power. I took the bike out for a test run today, and it's great!
At 50% power, it's fairly ordinary, the top speed on the flat (no pedalling) is 17 kph (that's about 11 mph). At 75% power, I feel the boost, and it's giving me 22 or 24 kph. And when I switch it up to 100%, it really zips along; on the flat I got about 30 kph. This power setting is for off-road use, because when I'm caching, speed isn't important, because I'm stopping every few hundred yards anyway. But power is; riding on a poor track takes a lot more effort.
While I was out, I was surprised to notice that the left hand brake was the front brake, and the right hand was the rear. Oops! I must have got the cables mixed up. So I sorted that out.
When I looked at my wattmeter after the test ride, I found that the maximum power draw was 48 amps! The controller is limited to 25 amps, so I'm thinking that this must have been a momentary draw.
I also found that the bike doesn't like starting from stationary on a steep hill. Well, who does? But the problem here, is that this is a sensorless motor (no Hall effect sensors). That means that the controller doesn't get the benefit of knowing where the motor is until it's actually started to move. So you can see that starting up on a steep hill is going to be bad news. Fortunately, I don't often need to do that; it happened on my test ride because I stopped half way up the steep hill at the bottom of my road, just to see how starting up would be.
Another interesting thing - it doesn't like being given wide open throttle at low speed. It's OK once the speed is up, but WOT at low speed leads to many complaining sounds from the motor.
So, overall, I'm very pleased with the result.