Newbie planning a DIY and wanted some feedback

[stormbeta]

Long-time bike rider, but first time building an ebike, looking to convert an old Diamondback Ascent Ex (26" MTB rigid steel/chromoly) and want to make sure I'm on the right track before I order any expensive parts.

Questions / advice wanted:
• The Diamondback only has 130mm dropouts - will this be an issue, especially for mounting the torque arm(s)? I've heard cold-setting mentioned but don't know if that would weaken the frame.
• I want the motor to be as quiet and smooth as possible - is it true that the Phaserunner would help with that?
• How does proportional regen/braking signaling work mechanically? The brake lever hookups look like on/off switches, and I can't tell if the lever throttle from Grin goes both directions or not.
• Do I need two torque arms given that I plan to make use of regen?
• Is it worth waiting to try and get the Crystalyte H3540 instead of the MX3006? They always seem to be out of stock... but the simulator indicates it can handle even higher loads before overheating.

Background/Details

I don't own a car, and this bike will be a bit of a workhorse for going longer distances or up steeper hills (10-11% grade max, most riding will be much less than that). I don't plan to go much faster than 25-28mph max. Base weight (me + bike + electronics + minimal gear) I'm guessing around 240lbs, and I'll want to be able to carry at least another 50-80lbs if needed. Budget is $1600-1700. Not sure exactly how much range I'll need but more is always better. I've test ridden a few ebikes and by far preferred direct drive systems - even with drag from cogging the quiet and smooth operation feels more natural to me than whirring geared motors.

Grin/ebikes.ca:
• MX3006 direct drive rear + statorade
• CAv3
• At least 25A controller (?)
• Cadence sensor to start with - would like torque eventually but BB replacement is a bit more complex than I want to mess with up front
• Torque arm and other misc parts

Battery:
Both EM3ev and Luna Cycles offer a 52v, 24-25Ah triangle battery; EM3ev is $700 (w/shipping) but takes a lot longer to ship and is 3lbs heavier than the Luna Cycle battery ($830 w/shipping).

 

[Jebe]

Anyone ? Interested in answers to this as well

 

[pwd]

The Diamondback only has 130mm dropouts - will this be an issue, especially for mounting the torque arm(s)? I've heard cold-setting mentioned but don't know if that would weaken the frame.

The 130mm dropouts shouldn't be a problem on a steel frame. They can usually be spread a bit without any ill effects.

I want the motor to be as quiet and smooth as possible - is it true that the Phaserunner would help with that?

A Phaserunner would help with that. Any other FOC or even a "Sinewave" controller will be more quite than a traditional trapezoidal controller.

How does proportional regen/braking signaling work mechanically? The brake lever hookups look like on/off switches, and I can't tell if the lever throttle from Grin goes both directions or not.

I have never used proportional regen, but I've read that you use the throttle to modulate the regen level when braking. Your best bet is to watch a video.

Do I need two torque arms given that I plan to make use of regen?

Two is always better than one, but it really depends of how much regen force is being applied to the axel.

Is it worth waiting to try and get the Crystalyte H3540 instead of the MX3006? They always seem to be out of stock... but the simulator indicates it can handle even higher loads before overheating.

That will come down to what power level you intend to use the motor at. At 52V x 25 amps, I think either motor would be fine. Play some more with the simulator to figure that out.

 

[stormbeta]

Thanks for the feedback!

The Diamondback only has 130mm dropouts - will this be an issue, especially for mounting the torque arm(s)? I've heard cold-setting mentioned but don't know if that would weaken the frame.

The 130mm dropouts shouldn't be a problem on a steel frame. They can usually be spread a bit without any ill effects.

I want the motor to be as quiet and smooth as possible - is it true that the Phaserunner would help with that?

A Phaserunner would help with that. Any other FOC or even a "Sinewave" controller will be more quite than a traditional trapezoidal controller.

Decided to go with the Phaserunner - I did more reading up on it and it has a lot of features I'd want:

* Torque-based throttle: fits neatly with torque sensor, and constant torque for acceleration is slower but more efficient based on the motor simulator. Not quite as efficient as a geared drive but close.

* Tight control over regen phase amps, which is important as I don't want to exceed the cells' rated rapid charge rates.

* Thermal rollback for controller

Plus from past experience I usually prefer things that I can tweak and configure over black boxes all else being equal.

Do I need two torque arms given that I plan to make use of regen?

Two is always better than one, but it really depends of how much regen force is being applied to the axel.

I ordered a second one just in case.

Is it worth waiting to try and get the Crystalyte H3540 instead of the MX3006? They always seem to be out of stock... but the simulator indicates it can handle even higher loads before overheating.

That will come down to what power level you intend to use the motor at. At 52V x 25 amps, I think either motor would be fine. Play some more with the simulator to figure that out.

I discovered that the trip simulator can plug in routes from Google Maps, which helped a ton - my main concern is that the steepest routes near me, while averaging 6-9%, have short segments (no more than a 1-2 hundred meters) that are 15-20%. It looks like I should be fine as long as I try to gain some momentum ahead of time and pedal hard for those sections.

I went with the MX3006 because I didn't want to wait for the Crystalytes, and it caps out at 28-29mph instead of 24 like the H3525 (@52v). Plus it uses a cassette natively, which I already have spare parts for.