julesos said:
One project I want to do is showcase long range for a fixed amount of battery (to showcase efficiency of electric motors and the aerodynamics of a velomobile form factor) - so hypermiling around a track as long as possible until the battery runs out. This would prioritise overall vehicle efficiency.
Another would be undertaking a longer journey (say Perth to Sydney) to again showcase a unique vehicle design and measure for energy use.
These would be like 100% electric, nil pedals - and involve data collection throughout.
The most efficient motor you can get would be good for those, but it does matter that you pick a motor winding that comes as close as possible to matching the speeds at which you wish to use it. If that is highly variable, so will the efficiency be variable.
AFAICR, the Leaf hubmotor in this thread
https://endless-sphere.com/forums/viewtopic.php?f=30&t=66489
is one of the most efficient commonly available ones, but generally the thinner the laminations, the better they'll be, especially at higher rotor speeds (meaning, it's more important in a geared hubmotor than a DD hubmotor). There's other factors, too, some discussed in that thread.
Outside these distinct projects, it would be good to use as an ebike (reinstall pedals) for use around town, again as part of outreach.
For this, if you want to demonstrate it's ability to be pedalled without power, a geared hub with internal freewheel/clutch that reduces resistance to moving iwhtout motor power will help. LIke the regular MAC, vs the GMAC (which has no freewheel/clutch).
Being a velomobile form factor (with some equipment) the weight will be high (compared to bikes) - say with rider up to ~ 160kg. Low aero.
Speed would be similar velomobile / ebike speeds ~ 30 - 40km/hr and able to do more as needed.
Then one question that remains is where you want the peak motor efficiency to be. The system should be geared (motor winding) for that to occur at the speed you want to use most often. But that may leave you with the inability to reach a high enough speed, if the speed you need efficiency at is too low. (teh simulator should help you see how that works).
Highest priority is to achieve the above. (So this alone, I think, takes out mid drive motors and some other factors).
Medium priority would be regen braking, reversing, sizing for vehicle, ability for hill climbing.
Lower priority would be cost, high maximum speed, high acceleration.
The GMAC is the only easily available higher-powered geared hub I know of that can do reversing and regen. All the others use a freewheel/clutch inside so they can't apply power in reverse (which means they cant' regen brake either). It's possible to "weld" that on at least some (perhaps all) geared hubs to defeat it, and make it possible to regen/reverse.
All DD hubs I'm aware of can regen/reverse.
Theoretically, as long as both motors are equally picked for the usage the system will see, the geared hubmotor should end up closer to it's higher effiency range more of the time vs a DD hubmotor, because the actual motor inside is able to stay spinning faster in a geared hub.
However, a DD hub can shed the waste heat much better, as it only has one thermal barrier set, where ageared hub has two.
Probably, a geared hub will be better for your purposes, and likely the GMAC is the only one that would do all the things you need it to.
You may have to consider what voltage the system needs to run at to give you the speed you need, though, if you need something faster than the typical ebike packs would give in the wheel size you want to use. (something else the simulator can help you figure out).
I’ll have to research controllers more about the capabilities and functions that are both available and what I need. (One feature that would be good would be to configure between ideal output for off-road EV use, and then configuring back when pedals are added to make road legal).
What I would do is simply make a list of the things you need the system to do, in as complete a detail as possible. Then we can help you figure out which (if any) controllers have those features, so you don't have to waste time doing it the reverse way.
For example, on my SB Cruiser trike, I need (not prioritized!):
--torque-only PAS startup, but
--cadence-only PAS control, with
--option for torque-pAS control, plus
--throttle-override control at any point,
--all of which controls motor current (vs speed), with
--adjustable throttle curves.
--FOC sensorless motor control, with
--silent operation, with
--perfectly smooth acceleration, with
--perfectly smooth electric braking, which
--doesnt' have to be regenerative, but does have to be able to go down to zero speed and actually stop the trike, and
--preferably to hold it in place against a slope
--with at least 8000w of power and option for more
--twin controllers that talk to each other to help adjust torque during turns (less inboard more outboard), and
--overheat protection that ramps down power rather than cutting it off
--overcurrent protection that protects against controller destruction
--ability to change any of the settings in the controller on the fly from a "dashboard" while riding
--abilty to see all the readings from internal sensors/etc on the dashboard (phase and battery currents, voltage(s), temperature(s), RPM, status readouts, error messages, etc
--easy programming that doesnt' require rocket science to setup
--autotuning for new motors,
--other stuff I forgot right now.
In my case, only about half of those things actually exist in any one system, at most. Combining three systems (Cycle Analyst v3, some electronics I have to design and build, and the HI-Lebowski controller project) gets me something more like 2/3 to 3/5 or more at once. The rest would require custom software and hardware for an android tablet (to make the dashboard) along with rewriting the Lebowski code to support it...and that's never going to happen (unless someone volunteers to do it for me :lol: ).
So...what's your list of "must have" features?
