Recently a friend reached out to me to build him a custom bike.. his budget was $2000 all in all. I decided to try to build him the most clean, OEM thing i could come up with.
The problem was that neither me nor him had much time. I was working 60 hour weeks, and he was moving to another state in a month.
We kind of haphazardly picked up a Giant NRS1 bike locally at a low price, and i decided i'd figure some way to make the build work anyway.
Ordered up a new 0.27mm MAC, anxious to give the new motor a try.. plus a small triangle pack from em3ev, so that the controller could fit in the frame bag for a nice appearance.
We ordered the cassette version of the motor, and to my dismay, it came with 12mm axles on both sides instead of 14mm. This was a huge disappointment that i did not expect.. and i have got in touch with MAC's people to let them know that this sucks. ( my eZee cassette motor has 14mm axles.. no reason a MAC can't have the same design )
However, with the 52v battery and a 10T wind, i observed a nice ~90W no load wattage which was truly impressive compared to my old 0.5mm lam MAC.. this impressed me enough to continue and not send the motor back in favor of a threaded freewheel edition.
This was my first em3ev battery order and i was very impressed with the fact that a discharge test sheet and all the correct connectors came included. Battery construction looked great.. i didn't cut it apart to see the guts though.
The first thing i did was make a shock protector plate that the triangle bag could rest on. I experimented with several designs and ended up reverting to the one i used on my super hiryuu bike. This design is already proven to be super solid, so no problem. I had to use a longer shock bolt and a series of 1mm thin washers to pull this off though.
The rear dropouts were a nightmare due to the curves on the dropouts, and lack of mounting bolts. In addition to this, we are dealing with a 12mm axle, which is rather weak to begin with.. so the dropout game needs to be extra good. I'm just glad we don't have regen here.
After experimenting with several brands of torque arm, many of which were loose in the slot, i eventually settled on ebikes.ca's front motor design for one side. Being able to rest the arm part directly on the frame made this pretty solid.
The other side had a disc brake hole to mount to, but a ridge in the dropout and disc brake mount that was a pain in the ass. I ended up using a variety of washers of various thickness in order to get a good solid contact with yet another ebikes.ca design. The dropout ridge is also acting to resist torque.. the torque arm is actually resting right on it.
The end result is quite nice and clean. I told the friend to swap out the front chainring. I didn't have time to do it, and sent him off on his way.
We used em3ev's cycle analyst v3 throttle setup. Honestly, i was disappointed with it.. no matter how i tuned the CA, it would induce throttle lag or slight stuttering in the MAC. This is no surprise.. the MAC is a difficult motor to control, and putting another layer of controller between the throttle and controller is a bad recipe. Luckily, said friend didn't mind.
Otherwise, the CA v3 is a pretty nice display and i knew it would make the geek in him happy..
The bike is tuned for about 1300w peak and 30mph continuous. I think the high efficiency super thin stator lams made for good power anyway because so much of that power is hitting the rear wheel. The Samsung 30Q pack only dips about 1.5-2.0v at peak load. It would be very difficult to get any part of this setup to overheat while climbing some major hills and i think it'll last a long time for him, which is exactly what you want when your customer is 700 miles away.
If i could go back and do this build again, i'd do the following things:
1) use a 14mm freewheel thread MAC.
2) paint all the mounting hardware black. ( didn't have enough time )
3) use CA V2 and direct to controller throttle.
4) Bump peak power up to 1600W.
Even with the little flaws, he still got something more powerful, durable, and with longer range than a commercial bike of the same price.. so i think this build was a real success anyway.
The problem was that neither me nor him had much time. I was working 60 hour weeks, and he was moving to another state in a month.
We kind of haphazardly picked up a Giant NRS1 bike locally at a low price, and i decided i'd figure some way to make the build work anyway.
Ordered up a new 0.27mm MAC, anxious to give the new motor a try.. plus a small triangle pack from em3ev, so that the controller could fit in the frame bag for a nice appearance.
We ordered the cassette version of the motor, and to my dismay, it came with 12mm axles on both sides instead of 14mm. This was a huge disappointment that i did not expect.. and i have got in touch with MAC's people to let them know that this sucks. ( my eZee cassette motor has 14mm axles.. no reason a MAC can't have the same design )
However, with the 52v battery and a 10T wind, i observed a nice ~90W no load wattage which was truly impressive compared to my old 0.5mm lam MAC.. this impressed me enough to continue and not send the motor back in favor of a threaded freewheel edition.
This was my first em3ev battery order and i was very impressed with the fact that a discharge test sheet and all the correct connectors came included. Battery construction looked great.. i didn't cut it apart to see the guts though.
The first thing i did was make a shock protector plate that the triangle bag could rest on. I experimented with several designs and ended up reverting to the one i used on my super hiryuu bike. This design is already proven to be super solid, so no problem. I had to use a longer shock bolt and a series of 1mm thin washers to pull this off though.
The rear dropouts were a nightmare due to the curves on the dropouts, and lack of mounting bolts. In addition to this, we are dealing with a 12mm axle, which is rather weak to begin with.. so the dropout game needs to be extra good. I'm just glad we don't have regen here.
After experimenting with several brands of torque arm, many of which were loose in the slot, i eventually settled on ebikes.ca's front motor design for one side. Being able to rest the arm part directly on the frame made this pretty solid.
The other side had a disc brake hole to mount to, but a ridge in the dropout and disc brake mount that was a pain in the ass. I ended up using a variety of washers of various thickness in order to get a good solid contact with yet another ebikes.ca design. The dropout ridge is also acting to resist torque.. the torque arm is actually resting right on it.
The end result is quite nice and clean. I told the friend to swap out the front chainring. I didn't have time to do it, and sent him off on his way.
We used em3ev's cycle analyst v3 throttle setup. Honestly, i was disappointed with it.. no matter how i tuned the CA, it would induce throttle lag or slight stuttering in the MAC. This is no surprise.. the MAC is a difficult motor to control, and putting another layer of controller between the throttle and controller is a bad recipe. Luckily, said friend didn't mind.
Otherwise, the CA v3 is a pretty nice display and i knew it would make the geek in him happy..
The bike is tuned for about 1300w peak and 30mph continuous. I think the high efficiency super thin stator lams made for good power anyway because so much of that power is hitting the rear wheel. The Samsung 30Q pack only dips about 1.5-2.0v at peak load. It would be very difficult to get any part of this setup to overheat while climbing some major hills and i think it'll last a long time for him, which is exactly what you want when your customer is 700 miles away.
If i could go back and do this build again, i'd do the following things:
1) use a 14mm freewheel thread MAC.
2) paint all the mounting hardware black. ( didn't have enough time )
3) use CA V2 and direct to controller throttle.
4) Bump peak power up to 1600W.
Even with the little flaws, he still got something more powerful, durable, and with longer range than a commercial bike of the same price.. so i think this build was a real success anyway.
