Sunday, 6 July 2014

Working on bike.1, part 1

Bike.1 has been suffering from an occasional (very occasional, it's only happened twice) problem with the motor. After being shoved through vegetation (which happens more often than I'd like), the motor runs so roughly that it's not usable; wiggling the wires entering the motor has fixed it both times. But I'm concerned that it might not next time, and I'll be left with a motorless bike for the rest of the circuit. Not a disaster, but not what I want.

My guess is that this is a loose connection on the hall sensor wires where they enter the wheel. Unfortunately, to check this, I'd have to take the motor apart, and to do that, I'll have to remove all the spokes and get the motor separated from the wheel. Which is a big job, but not as big as the job of replacing it, and truing and balancing the wheel.

So I've bought an Infineon controller that will run the motor without the halls; my plan is to take it with me (it's not large or heavy) and if I have an unfixable problem in the middle of a circuit, I can swap to the sensorless controller. But for that to be viable, it has to be a matter of unplugging the old one, and plugging in the new.

I have an sensored Infineon controller. The easiest solution will be to install that. Then, to make the swapover, I just unplug a few wires, and plug them into the sensorless controller. I use EC5 connectors for the power (male at the controller, female on the battery side) and 4mm connectors for the phase wires (male on the controller, female on the motor side), so it's easy for me to plug any controller into any battery, and and controller into any motor. But the throttle connectors are non-standard, but that's not a big problem, it just means that I have to solder up a little adaptor.

Today, I started to work towards that, and I ran into an immediate problem. When I put on the sensored Infineon controller and powered up the bike, the wheel didn't turn.

There's a way around this. There are three phase wires from the controller, and three to the motor; yellow, green and blue. But these colours aren't in any way standardised. Everyone uses yellow, green and blue, but the colours don't always mean the same thing.

But there's only six possible ways you can connect three wires to three wires. I tried them all. On four combinations, the wheel didn't turn. On one, it ran backwards, roughly and very fast. And on one it ran forwards, smoothly, with a top speed (with the wheel held in the air) of 19 mph. That top speed worries me a bit; if it's 19mph under no load, what will it be when I'm sitting on the bike and riding along? There's only one way to find out, and I'll do that another day, it's too late now for testing.

I also tried the sensorless controller. That gave me an immediate good result, but again, a top speed of 19 mph.

I also found a couple of problems. The yellow Hall wire was almost broken off, so I had to replace the connector. And the black hall wire is almost broken, that's going to need soldering. Worse, though - the front wheel was loose! I have no idea how it got like that, but a loose wheel is *very* dangerous. You can imagine what would happen if the front wheel came out of the forks when I'm going along at speed! Ouch. Ow ow ow. So I sorted that out.

I'm not finished yet. I have to do a road test with each of the two Infineon controllers, then a road test with the original controller, then I can decide whether to use the original controller or the Infineon, and if I use the original, I have to do a bit of soldering to make it all plug-compatible (meaning, adaptors for the throttle), so I can do a controller swap in the middle of a muddy field.

And I need to sort out a good connection for the handlebar voltmeter - I think I know how to do that on the old controller. There's a four-way cable that goes to the original throttle/blobmeter. Three of those will be for the throttle; 5v, hall and earth. And the fourth one, I think, will be the battery voltage, so all I need to do is use that. With the Infineon controller, I've already worked out a suitable adaptor, taking the voltage reading from the six-way cable.

Then I can neaten up all the cables using some curly-wurly, use cable ties to secure the cables to the bike, more cable ties to secure the controller to the back rack (reusable cable ties, so I can do a swap in the field), put on lengths of inner tube to waterproof the cabling, and I'll be ready to roll.

But tomorrow, I'll be using bike.3. I've bolted the side-stand on more firmly, because I don't want a repeat of part of it falling off as I go across a field. And I've bolted a piece of plastic to the rear carrier, to keep the pannier well clear of the back wheel. A pannier rubbing against the back wheel will soon wear through.

No comments:

Post a Comment