Regen

ynot

100 W
Joined
Mar 5, 2022
Messages
105
With the current enthusiasm for regen brake assist, why not switch the industry back to brushed type motors, which make regen very simple. If a brushed motor is spinning and not drawing current, it is generating current, about as simple as it can be.
Just how much of an advantage do brushless motors give vs brushed motors?
 
Brushless share that same trait....

"motor is spinning and not drawing current, it is generating current"

Further, if it's generating current, it's also generating drag, a trait many would just as soon be without.
 
ynot said:
Just how much of an advantage do brushless motors give vs brushed motors?
There's been a few discussions about the differences, some of them are listed here, if it's helpful. https://endless-sphere.com/forums/search.php?st=0&sk=t&sd=d&sr=topics&keywords=brushed+brushless&sf=titleonly


Aside from any other efficiency or functionality, covered in the above list:

For hubmotors, or any motor not automatically mechanically disconnectable from the wheel, brushed motors present the problem that a controller failure may result in the motor being stuck at full throttle (a common FET failure mode is shorted on, feeding full power to the motor).

Nothing short of cutting power completely to the controller will stop that, so whatever power cutoff switch/etc you have needs to be able to handle the full current load at system voltage, or else it may also fail shorted (weld itself closed upon attempted disconnect from internal arcing).

Brushless motor controllers don't fail the same way--if their FETs fail shorted, the same mechanical arrangment of motor drivetrain presents high rotating resistance instead, equivalent to braking action.
 
ynot said:
With the current enthusiasm for regen brake assist, why not switch the industry back to brushed type motors, which make regen very simple. If a brushed motor is spinning and not drawing current, it is generating current, about as simple as it can be.
Just how much of an advantage do brushless motors give vs brushed motors?

I'd guess that the average 10% efficiency advantage of a brushless motor over brushed, outweighs any advantage in simplifying regeneration via a brushed motor. Since the losses go to heat, it seems like the brushed motor would have a heat disadvantage when both running or regenning. Just guessing. Better range, less heat for brushless.
 
ynot said:
With the current enthusiasm for regen brake assist, ...
Perceived "enthusiasm for regen brake assist" ....
and that is among people with heavier, more powerful E-bikes and or those regular traveling in hilly areas. Those of us in the flatlands really do not care much for the extra complexity.
 
E-HP said:
ynot said:
With the current enthusiasm for regen brake assist, why not switch the industry back to brushed type motors, which make regen very simple. If a brushed motor is spinning and not drawing current, it is generating current, about as simple as it can be.
Just how much of an advantage do brushless motors give vs brushed motors?

I'd guess that the average 10% efficiency advantage of a brushless motor over brushed, outweighs any advantage in simplifying regeneration via a brushed motor. Since the losses go to heat, it seems like the brushed motor would have a heat disadvantage when both running or regenning. Just guessing. Better range, less heat for brushless.

Used to build small wind turbines for yachties, and came up with a wonderful clean looking fiberglass housing that complimented the look, problem was the generator got so hot inside the housing, that the generators permanent magnets came unglued.
Have watched Justin's video waxing strongly about regen & since one of the motors on my bike is brushed. (rather old bike which I like) got me to thinking how easily I could set up regen at least for the rear wheel.
10% is easily overcome with a little more battery and a little bigger motor.
 
My youngest daughter has a scooter that has NO disk brakes and only has a second thumb throttle on the left side for brake,
initially I was like, holi cit, this sheesh won't stop her etc, but nope that works better than more disk brakes and does well for her, she weights 145lbs.

Regen braking works and most controller don't have the variable regen braking feature and most frames/torque arms are designed for it.
 
gobi said:
... a scooter that has NO disk brakes and only has a second thumb throttle on the left side for brake,
What happens if the second thumb throttle fails or goes intermittent? Full braking? No brakes at all?
 
99t4 said:
gobi said:
... a scooter that has NO disk brakes and only has a second thumb throttle on the left side for brake,
What happens if the second thumb throttle fails or goes intermittent? Full braking? No brakes at all?

This is a fair point. Regen braking is controlled electronically. Lets say you have a BMS cutout due to high current acceleration. Controller looses power too, meaning no regen, and if that's the only brake you have setup...
 
Sure, that's a fair point if the proposition is to eliminate physical brakes and rely solely on regen. I think his point, however, was not that we should follow the lead of that scooter maker and do that, but that inasmuch as they can get away with that and it works, then it's a testament to the usefulness of regen braking. Considering that people sweat over the difference between various kinds of disk brakes etc.

It's a shame that no one has figured out how to clutch the freewheel in geared hubs, so the only choice here is between "no regen" (most) and "no freewheel" (GMAX.)
 
Theres' at least one thread here someplace that talks about doing that, but I can't remember if they ever accomplished it.


However, since most of the "clutches" in geared hubmotors are roller/ramp type, all you need to do to lock one is interrupt the roller's ability to go back into the slot, to prevent it from freewheeling. How much torque it can take in this mode, you'd probably have to test--it's probably not as much torque as in the unlocked mode in forward direction, but it could be more if it forces the roller against the surface harder than the ramp would.

This patent's images
https://www.freepatentsonline.com/20170096030.pdf
show something that pushes pins into the space the pawls of a typical single-speed freewheel, to keep them from retracting. If a device that does this, either mechanically or electrically controlled, can be designed to fit inside a geared hub, it could be used to lock/unlock the clutch at any time.

Alternately, the electrically-actuated clutch design from an automotive air conditioning compressor, etc., could be adapted to completely replace the clutch in the geared hubmotor.

Or a differently designed clutch altogether, like a mechanically-activated dog-clutch, could be used. (the patent design is something along those lines but only as an adaptation of the existing freewheel; this would instead completely redesign the clutch around the idea).


The reason to completely replace the clutch with different design is that at least some of these clutches are just not made well enough to handle the torque of the motor itself, when run at two or three times the power level they are nominally "rated" for--they tend to crack between the notches for the ramp and the outer edge, which then keeps the rollers from pushing against the ramps and transferring torque as designed. Braking torque can be more suddenly applied and higher peak torque, so could be more likely to break clutches.
 
ynot said:
With the current enthusiasm for regen brake assist, why not switch the industry back to brushed type motors, which make regen very simple.
Regen is easier and more controllable with brushless motors (and a motor drive) than with brushed motors.
 
donn said:
Sure, that's a fair point if the proposition is to eliminate physical brakes and rely solely on regen. I think his point, however, was not that we should follow the lead of that scooter maker and do that, but that inasmuch as they can get away with that and it works, then it's a testament to the usefulness of regen braking. Considering that people sweat over the difference between various kinds of disk brakes etc.

It's a shame that no one has figured out how to clutch the freewheel in geared hubs, so the only choice here is between "no regen" (most) and "no freewheel" (GMAX.)

Well, there are motorcycle manual clutches. It can be combined a fixed gear left-sided middrive, so you can have both.
Problem is, they are an overkill for a bicycle - where 'intermittent assist' (accelerations, hills - and regen braking, right) makes most sense.

Admittedly, 'virtual freewheel' (very low power assist to compensate for cogging/iron losses) most does make sense and actually regen braking DOES make up for it - usually, when riding in city and with hills.
However, during long-distance riding on mostly pedal power - no, it does not, especially if your motor is particularly powerful.

There are two options here:

A really *low* power DD motor or one stage reduction fixed gear middrive using RC motor, just enough for 'low assist level when pedalling' and scrubbing speed to decelerate with conventional brakes to come to a complete stop, and very low cogging losses combined with conventional geared motor. In fact, a rear geared motor with left-sided midddrive replacing rear cog seems like the best solution. You can have very considerable assist and regen with 700gr e-board motor and 24v battery at medium to high (think ~50 kmh) speeds, while allowing very brisk accelerations and hill climbing from low to mid speeds when using a combination of two motors and pedalling.

However, if you need efficiency AND high braking power, there is possibility of combinting a middrive e-board motor with a 50cc moped centrifugal clutch (it also serves as a gearbox).

You *will* take a considerable hit to efficiency when using it for high-torque starts, but when fully engaged it should be as efficient any other middirve, and as long as you don't use it to brake to a complete stop and within torque range nessesary for engagement - you can use it both AS assist AND regen (most centrifugal clutches dont care one way or the other), but by overcoming engagement static friction with a brief, but powerful burst of regen braking you can force disengagement and go into 'fully freewheeling' mode.

You just need smart torque-controller like VESC and very simple MC like arduino to handle the logic.
 
harrisonpatm said:
99t4 said:
gobi said:
... a scooter that has NO disk brakes and only has a second thumb throttle on the left side for brake,
What happens if the second thumb throttle fails or goes intermittent? Full braking? No brakes at all?

This is a fair point. Regen braking is controlled electronically. Lets say you have a BMS cutout due to high current acceleration. Controller looses power too, meaning no regen, and if that's the only brake you have setup...
https://gotrax.com/collections/gotrax-refurbished-products/products/certified-refurbished-g2-electric-scooter

I hear you, as a dad I was breaking sweats,
I think Gotrax shared your concerns as ALL the new models have a brake lever.
 
Back
Top