Ok so yesterday I pretty much completed my Brompton build in one day. The first thing I found is that the Crystalyte G209 motor may well have a 74mm dropout spacing but it WILL NOT fit straight into Brompton forks! The only way the motor will clear the forks is by using the supplied washers which makes it effectively 10-15mm wider!
So I had to stretch the forks somewhat using threaded rod so that the wheel would go in. Incidentally there is no indication on the motor which way it will rotate either so for reference if anybody else uses this motor the wire exits on the left of the bike (which I found after getting it backwards).
I modified the original Brompton retaining washers by filing them out to fit over the Crystalyte axle. There is absolutely no room for torque arms or thicker anti-rotation washers. As you can see in the pictures there is barely room to fit the supplied washers and the mudguard clip onto the axle.
Having fitted the wheel I then had some trouble with the fold due to the slight width change. This was solved by moving the brake levers around a bit to stop them fouling the wheel when folding.
I then wired everything up to my Juicy Bike battery using temporary block connectors and got the wheel running nice and smoothly; everything worked well first time although the rpm seemed quite low. I put all the wiring and controller in the front bag and took it for a little ride; it worked well but was quite slow (14mph on the flat) but seemed to go up hills nicely. I then had a tinker with my programmable controller and managed to up the speed to an unloaded 18.5mph which seemed a little more hopeful.
Then the fun bit...all the cables were routed into my aluminium box (decided not to paint it for now) through rubber grommets. I wired everything up by soldering and double shrink wrapping every connection...this took hours!! I wired in my DC-DC converter and also connected up two splash proof toggle switches to the left of the box. These operate the Magicshine light and ignition. I soon realised that I would never have managed to fit all this into the original smaller box that I considered. For now I have an umbilical power cord which just plugs into my Juicy Bike battery; when I get my proper battery I will have it connecting through the luggage block like the Brompton Nano...no loose wires and nothing to plug in....pop the bag on and go. I mounted the Cycle Analyst directly on top of the controller box and dropped the wire straight down to the controller; I couldn’t mount it directly aligned as it interfered with the fold (of course ).
And then the first test ride – what a perfect setup this is for what I need....speed on the flat unassisted is between 15-17mph; perfectly acceptable and comparable to my Juicy Bike. The Brompton also flies up any hill although my battery doesn’t much like supplying high current so I didn’t push it too hard. I’m really looking forward to getting the big Lifepo4 hooked up!
So for now I think I have just about achieved my aim....a Brompton which still folds small enough to pop on the train but seems to power up hills at a good 15mph. I have also found that the motor is not that restrictive when going downhill; I can easily pedal over 20mph so my average speed should be just fine.
I also have regenerative braking activated on this controller. I’m not too bothered about the extra little bit of range this will create but I am super impressed at the actual braking sensation. As it kicks in I get a nice gradual reduction in speed without having to cause any wear to my brake blocks or rims...if I need to stop quicker I just squeeze harder to activate the actual brakes. Brilliant!
The fly in the ointment?.....it’s heavy. Without the battery it’s 16.3kg so with my lifepo4 this will be 22.3kg. But this is what I expected and contrary to some opinions in this thread I should hopefully have produced a reliable long range commuter Brompton (I guess time will tell).
I am struggling to get the Cycle Analyst calibrated properly...it’s reading up to 70A at the moment which is totally wrong (controller is programmed to 20A max). Some more fiddling with settings needed (I suspect an in correct Rshunt value has been given to me by Ed Lyen). I therefore cannot comment at all on the efficiency of the motor until I get this sorted out.