[arber333]

Hm...

If the wheel has rating only up to 27mph i expect it is due to tire design or structure. Auto tires have speed rating as well. I think this tire will pop at anything over 40mph. https://www.tirerack.com/tires/tiretech/techpage.jsp?techid=35

The other is the KV of the motor. It will not want to run over that speed.
So in case you want to use gas engine on the rear wheel and hub motors on front at highway speeds i recommend you verify if controller has release function. I have seen this as a switch in wiring. When applied this feeds motor with just enough modulation for magnets to turn with 0A. This would allow you to run the gas motor without hubs in regen. But i cant imagine what inverter would do at highway speeds. My guess is backEMF spikes from motor would fry reverse diodes in transistors.

First of all i recommend you use scooter or motorcycle wheels for that speed and go and compare at what speeds motor is designed to run at. Any PM controller with correct voltage would do if motor has correct KV and battery voltage is sufficient to turn it at such speed.

I suspect if you have $$$ to buy LiTi cells you should have money to buy proper hardware also.

Some items to compare...
https://www.kellycontrollers.eu/hub-motor-48v-3kw-high-speed
https://www.kellycontrollers.eu/hub-motor-72v-45kw-high-speed

https://www.ebay.co.uk/itm/60v-72v-...290571?hash=item26255823cb:g:7F0AAOSwhpBdTFEx
https://kellycontroller.com/shop/kl...tribute_pa_current=350a&attribute_enclosure=N
https://kellycontroller.com/shop/kls-h/

 

[arber333]

Oh... there were tadpoles with proper design made before. Check...
https://rqriley.com/product/xr3-hybrid-plans/

http://www.evalbum.com/559

 

[slomobile]

Well thanks for redesigning the vehicle, but I really only need information about the way the controllers and oversped motors interact. I'm not sure I followed what was being said about the inverter. I've not seen anything described as a release function. Are you talking about open circuiting all motor phases or something else? When is it OK to reconnect? Is stopping and power cycle required? I suspect the KLS controllers will be ill suited due to inability to change firmware. I need regen to kick in automatically(with minimal predictable drag) above electric drive speed (27mph) and kick out before it becomes unsafe. 27mph is probably not unsafe, its just the fastest the electric motor alone will drive the bike. If that is kv limited, I'm ok, if the limit is due to the commutation speed limit imposed by a slow controller, then I could be in trouble when phases get out of sync. I don't even know the right language to use when asking for this spec. Ideas?



I've pedaled 45mph downhill on a mountainbike with nothing worse than brake fade. Land speed record on a bicycle is 280MPH+ so the mechanicals can be tamed. Its the electric pixies that concern me. 1000W on 2 wheels with 12GA spokes is all the strength, power, and weight that I require. The batteries were chosen to meet the specific need to absorb regen current in a short ride, not to throw money around.

 

[arber333]

I am not sure how the wheels will hold up at lateral load in the turns. Consider that in a single track vehicle tire load should be always with the N force.

Well engine speed over the available voltage is possible. We call it field weakening range. With more rpm motor backEMF rises to the available battery DC bus voltage and motor refuses to spin more. In this case we need to weaken the force of the magnets so that voltage does not rise too high. We induce some counter force in stator so rotor can now turn faster. BUT in the case something goes wrong and nverter shuts down magnet force returns and DClink voltage will rise. If the battery is powerfull enough it can absorbe some voltage otherwise this can quickly kill inverter transistors or capacitors. Usually inverter with FW are more expensive and will require some method to detect rotor angular position.

If you would need more RPM from one motor and do not have FW inverter, you could always try to redesign motor winding to higher KV number. This will lower torque, but RPM will be higher.

Hm...or you can simply chose inverter made for higher voltage. That should increase your speed-voltage margin.

 

[brian_]

My plan is to add 2 of the best I can find 20" electric fatbike (bicycle hub motors) in place of the front wheel on a functional gas motorcycle. The purpose is to increase ground patch and achieve all wheel drive for riding on grassy hills without tearing up the grass or spinning out.

It might not matter a lot to the EV aspects, but just for clarity...
Will this become a trike (with two separate wheels in front) which doesn't lean into corners, or a bike with a double front tire (still leaning into corners), or something else?

 

[slomobile]

It is a design similar in many ways to a Piaggio MP3 except both front wheel axles are supported on both sides. There will be 2 front wheels that lean with the bike sharing the load in turns as much as possible. Check the video at 1:49 for a better view.



I cannot build my own hub motor. At least I've never tried. I want larger diameter than typical scooter wheels for negotiating obstacles off road, lighter weight, lower power than car hub motors. 20" ebike hits the sweet spot.



I thought field weakening had been almost universally adopted in modern controllers but unmentioned in marketing because anything with "weakening" sounds bad. Is that not the case? Most of the good hub motors I've seen have redundant hall effect sensors for position sensing and controllers which use them. Are you saying that I could use a 48v battery, 48v hub motors, and a 96v capable controller and just feed it 48 volts? motorkv * max RPM = the voltage of the controller I need?

 

[arber333]

It is a design similar in many ways to a Piaggio MP3 except both front wheel axles are supported on both sides. There will be 2 front wheels that lean with the bike sharing the load in turns as much as possible. Check the video at 1:49 for a better view.



I cannot build my own hub motor. At least I've never tried. I want larger diameter than typical scooter wheels for negotiating obstacles off road, lighter weight, lower power than car hub motors. 20" ebike hits the sweet spot.



I thought field weakening had been almost universally adopted in modern controllers but unmentioned in marketing because anything with "weakening" sounds bad. Is that not the case? Most of the good hub motors I've seen have redundant hall effect sensors for position sensing and controllers which use them. Are you saying that I could use a 48v battery, 48v hub motors, and a 96v capable controller and just feed it 48 volts? motorkv * max RPM = the voltage of the controller I need?

I took my EV motor to a electro technician which was also rewinding various other motors. I asked him to rewind my motor to my configuration. This changed my 28kW motor at 75Vpp into 190Vpp which goes very well with my battery votlage of 360Vdc. Power remained the same, but i feel i can now run at higher power for longer due to lower amps in the system.

Nope! FW has to be implemented in software or not. This is the ability to go over the normal rev specs of the motor and HW design of inverter has to be beter in case something goes wrong. Like if wheel hits a pothole in the road while in FW range. Transistors have to be higher voltage and main capacitance is usually higher. This is to absorb power if DC contactor goes offline and battery is disconnected from inverter in FW region.

In your example if inverter would have FW in code you could rew your motor from 48Vpp to some 72Vpp which is some 66% over the specs.
Well in real world when applying more rpm you also have to feed the winding to vary phase excitation. This creates heat in stator which slowly builds up.
In the real you would use some 25% FW and make use of good stator cooling mechanism.

Also kW is kW and if your motor is rated for some continuous power, you have to abide to this power, even if you increase RPM. That means lower mechanical torque at higher speed, which in electrical terms is lower amps at higher volts. If not life of motor is reduced...

Some motors do not want FW, they become loud and release various schratching noises etc... This is all losses that go into heat.

 

[brian_]

It is a design similar in many ways to a Piaggio MP3 except both front wheel axles are supported on both sides. There will be 2 front wheels that lean with the bike sharing the load in turns as much as possible.

Thanks for the clarification. I was wondering if it might be like the MP3, but didn't want to introduce that complication to the discussion unnecessarily.

In this design, front wheel speeds will be slightly different in a turn, so the two motors will need to be controlled in a way which allows that.

Are you saying that I could use a 48v battery, 48v hub motors, and a 96v capable controller and just feed it 48 volts?

That seems correct to me. The controller's voltage rating is a limit, not a requirement. There will be a minimum voltage in the controller specs, but it won't be the same as the maximum voltage.

motorkv * max RPM = the voltage of the controller I need?

Almost. That (KV * speed) is the voltage requirement to overcome back EMF and spin with no load; you'll need more to push current through winding resistance and produce useful torque.

 

[slomobile]

Thank you brian, that was very helpful. You also anticipated my next question. How can I maximize repeatability/linearity/predictability of 2 motor/controller combos? I imagine headroom from using a high volt/high current controller far below its max ratings(in low speed mode) will go some way toward this.


Also, do I need to dyno all this for safety before a test ride? Tadpole is less flip prone than 3 wheel ATV but there is still danger I think. What is the worst that could happen Edit: if the controllers lose sync with the rear wheel or with each other? That isn't hyperbole, I'm really asking for potential worst case scenarios so I can prepare for them.


I need more than open loop throttle proportional torque control. I need closed loop wheel position control accurate down to (I think) below pi/6; while also minimizing acceleration and jerk in low speed mode. Changing acceleration/jerk limits in high speed mode to not interfere with spirited gas acceleration, provide smooth regen braking, and not cause unwanted steering torque.
Low speed = wet grass hill. High speed = dry flat freeway pavement. I would be interested in other opinions of how accurate wheel speed or wheel position control needs to be for traction control on wet grass hills.



I understand that throttle applied doesn't necessarily translate directly into a specific RPM so this will largely be a factor of the update speed of PID and sensor loops, and consistent step response from the system in various states.


Applying torque inappropriate for a given wheel path can create unintended steering forces. That was yet another reason for choosing motors that barely meet torque required in the low speed situation; to mitigate severity. In High speed mode, the electric system's job is basically to get out of the way, but provide regen braking when it can. Not any contribution to highway driving except for a brief holeshot torque to help save gas and emissions.



There is a single twist throttle connected to the gas engine. The Throttle Position Sensor(TPS) on the intake provides a 0-5v signal to ElectricPowertrainControlUnit(EPCU) and GasPowertrainControlUnit(GPCU). A selector switch for Low/High speed mode and a reverse button are the only controls added to the usual motorcycle controls. A brake fluid pressure sensor or brakelight switch triggers regen depending on which controller is being tested.



When in low speed mode, optical flow sensors vs wheel speed sensors trigger traction control interventions including modulating electric drive and sending "slow down" signals to GPCU which alters ignition timing and/or selected fuel injection map. That allows some control of rear wheel torque without authority over throttle butterfly position or hydraulic brakes.



GPCU outputs a signal whenever gas engine is running. If high, EPCU engages only when transmission is in one of 5 forward drive gears. EPCU monitors rear wheel speed VSS in transmission output and attempts to match front wheel speeds to pace +/- depending on steering angle sensor. The amount of differential is not fixed yet. It depends on the algorithm chosen, preferred feel, and the amount of Ackerman built into the mechanical steering system. Ackerman is still somewhat adjustable right now, so software has to be updated whenever there is a mechanical change.



If low, engine dead, key on, EPCU engages only when transmission is in Neutral. This enables totally electric front wheel drive and is the only mode where the reverse button is active.



At some point I might add a hidden override to allow bump starting, or even control of gas starter motor as hybrid cars, but right now that is off the table.

 

[slomobile]

Almost. That (KV * speed) is the voltage requirement to overcome back EMF and spin with no load; you'll need more to push current through winding resistance and produce useful torque.

It just struck me that your statement says

I still need to produce the higher than battery voltage to equal back EMF

even if it doesn't need to have much current behind it.

I don't need or want it to produce any torque at that speed, forward or back.


So that introduces a new problem, how to design a higher than battery voltage supply that can be integrated into the typical Ebike system and only kicks in when needed without breaking anything else.


Thoughts?

 

[arber333]

It just struck me that your statement says

I still need to produce the higher than battery voltage to equal back EMF

even if it doesn't need to have much current behind it.

I don't need or want it to produce any torque at that speed, forward or back.


So that introduces a new problem, how to design a higher than battery voltage supply that can be integrated into the typical Ebike system and only kicks in when needed without breaking anything else.


Thoughts?

Ah! Now i see clearer around your application. You would want to ride high speed downhill not under power then. Why not have rear drive with sprag clutch like on the bicycle then? Do you need regen that much?

Heh, maybe the easiest solution would be to just use one ACIM motor connected to the rear wheel. ACIM inverters have FW implemented allways, and if you stop PWM there would be no excitation and no torque.
This will also eliminate the need for torque regulator for 2 controllers on the front wheels. Its installation would be heavier though...

Tadpole is more stable in a turn only if you put CG in the 1/3 body length over the front axle. And i think in that case it would be better for traction to have a rear drive. Remember when you apply power to the rear wheel it tends to push that wheel downwards whereas with front wheel drive it tends to raise them.

 

[remy_martian]

Your controller needs to be rated at least 3x the voltage rating of the hub motors at 27MPH.

Being weird means building your own controller.

Riding that thing with the front wheels driving will be the same as hitting a patch of gravel with the front wheel in a turn on a roadbike. That also includes your regen, downhill, mode. Unlike a car, your bike can have NO front wheel slip at all -- your biggest problem here is not the controller, but detecting wheel slip prior to *onset*. Good luck with that one.

Also, if your battery is full and you still want downhill active braking, you'll need to dump the energy in a controlled fashion into a resistor. Again - custom controller.