Hub Motor as Generator - 3 phase rectify vs. Regen

I've done a bit of research in my effort to build a series human/electric hybrid drive (the electric chain, if you will). I've found a variety of information about the hub motor as a generator. Most of what I've read from hobbyists (wind power enthusiasts, bike tinkerers, etc), says that the output of a brushless DC hub motor is 3-phase AC. So I would need a bridge rectifier and a DC-DC voltage converter (Buck-Boost) to get the output voltage to charge the battery. However, from what I've read from hub motor manufacturers, the BLDC motors that have regenerative braking capabilities just feed directly back into the BMS (or sometimes the battery directly). If so, then I wouldn't need to use the rectifier or converter at all - presumably all of that is being done by the regenerative braking controller.

If the hub motor can just feed directly back into the battery in regen mode, then shouldn't I be able to hook the pedals directly to the hub motor (as generator), and go to town (so to speak). If the controller is in regen mode, then it should give me a load on the generator (to push against), and the voltage/current out should charge the battery correctly. I would much rather use that system than trust my own ability to correctly design the electronics for a generator/charging system.

Also, wouldn't the regen controller have a built in diode setup - to prevent the battery from running the generator as a motor when the battery voltage goes above the voltage from the generator? (when I get tired of pedaling ).

One more question about the regenerative braking. Is the braking load variable? In other words, when I squeeze on the brake level (before the mechanical brakes engage) is there a variable level of "electronic" braking being applied?

Thanks,

Myles

I'm no E.E., but I have seen the perpetual motion machines ideas come and go here and none have panned out yet.
Parasitic loses? I'm sure the Experts here will elaborate.

When I was a child, we would visit the Smithsonian quite often and I was always impressed with the Foucault Pendulum, which at the time, dominated the lobby of the National Museum of American History. What a wonderous thing, the motion of the Earth could keep this giant ball swinging.
I was rather crest-fallen when my dad told me that their were electro-magnets in the cieling to give it a ""boost" now and then.

Um, actually, I'm not trying to do anything to break the laws of Thermodynamics (already tried that for my 8th grade science fair project).

The idea is to transfer the energy to the drive motor via electroni wires instead of a mechanical chain. I completely understand that a chain is 95% efficient, whereas a generator-motor system would be closer to 70% efficient. I'm willing to live with that for other reasons. I just need help figuring out what hardware I need to get for the system, and whether the regenerative braking system of a BLDC hub motor will do the same thing as the 3 phase bridge rectifier & DC-DC buck boost converter.

Myles

OK, like I said, the experts will be along. While you are waiting, you might try the search function, specificly, the search within a search. Hubmotor, then regen, etc. and the posts will appear.

mytach said:

Um, actually, I'm not trying to do anything to break the laws of Thermodynamics (already tried that for my 8th grade science fair project).

The idea is to transfer the energy to the drive motor via electroni wires instead of a mechanical chain. I completely understand that a chain is 95% efficient, whereas a generator-motor system would be closer to 70% efficient. I'm willing to live with that for other reasons. I just need help figuring out what hardware I need to get for the system, and whether the regenerative braking system of a BLDC hub motor will do the same thing as the 3 phase bridge rectifier & DC-DC buck boost converter.

Myles

Click to expand...

I think you may be very lucky to get 70% efficiency. You want a small direct drive hubmotor spinning at a decent clip as your pedal motor, connected by a controller in regen mode going to a smallish battery, and then any motor to propel the vehicle that suits your application. It will work fine, it will not drive the generator when you stop pedaling, a proper regen configured controller will not overcharge the battery, it all should work fine, but I imagine ~50-60% efficiency will be the real-world efficiency of the system. Remember, the efficiency of every component gets multiplied in the string of power transfer from your legs to the eventual traction motor propelling the vehicle.

Feet to pedals to chain to hubmotor generator drive~98%
Hubmotor ~80% (efficiency hit from regen boost timing)
Generator to controller in regen ~95% (it's not as efficient in regen as in drive, because it has to switch at different times to boost the voltage)
Battery charge efficiency 99%
Drive controller efficiency 97%
Drive motor efficiency 85%

This efficiency stack-up yeilds 60%, and it's pretty optimistic.

What are the real world reasons for this?

cal3thousand said:

What are the real world reasons for this?

Click to expand...

Creative fun could be a good enough reason. The ability to buffer energy into a battery is also kinda neat, sit stationary pedaling for 5 minutes and then release a 15second burst of high power to sprint up a hill with the power you stored.

Personally, I would just put the space the pedal/gen/controller setup occupies into more battery, and have a charger you plug into an outlet for energy. However, I could see it as someone viewing it like a stationary exercise bike, that rather than just wasting the energy in boring friction, stores it so you can do little blasts around town with the pedaling work you did in your living room.

There are several reasons. First, though the primary power will come from the battery, I don't want to be stranded when the battery runs out. Second, with a generator to battery system, I can pedal while I'm coming to a stop, and even while I'm sitting waiting for the light to change. Third, I'm not sure what the local regs are on a purely electric vehicle, so if I'm pedaling at least it looks like it's still a bike

So far the vote is that it should work with a motor in regen mode? Does anyone know about the variable braking force?

Myles

There is at least one trike here on ES that already uses a pedal-generator, but I have forgotten which member built it. It's quite a nice trike though.

Thanks, amberwolf. I did a quick search and found several threads with similar topics. LowRider did a velomobile style trike with a series hybrid system. Was that who you were talking about? Looks pretty cool.

I heard back from the Hub Motor manufacturer that the regenerative braking force is constant (either On or Off) when you squeeze the brake lever. That's not what I was hoping for. I was hoping that it would be a potetiometer in the lever that would signal the regen control to brake harder or softer. But they did say the regen braking force was programmable (through USB), so I guess that's better than nothing.

Myles

mytach said:

Thanks, amberwolf. I did a quick search and found several threads with similar topics. LowRider did a velomobile style trike with a series hybrid system. Was that who you were talking about? Looks pretty cool.

I heard back from the Hub Motor manufacturer that the regenerative braking force is constant (either On or Off) when you squeeze the brake lever. That's not what I was hoping for. I was hoping that it would be a potetiometer in the lever that would signal the regen control to brake harder or softer. But they did say the regen braking force was programmable (through USB), so I guess that's better than nothing.

Myles

Click to expand...

one thing about regen breaking, is that its strength is rpm dependant. so while you wont have a direct control over it, you could potentially use a simple gear system to vary your 'load', at the sacrifice of a small efficiency hit.
I think there are controllers out there though that have variable regen... just a question of finding 'em.

There's no doubt you can make a series hybrid drive like this work, but as you've already guessed the efficiency isn't great. The main problem is spinning the motor that's used as a generator up to sufficient speed to get reasonable efficiency, as the pedal crank rpm is unlikely to be more than around 100 to 120. You could adapt a hub motor to do this, but they typically run around two or three times faster than this for reasonable efficiency, so ideally you'd want to have a mechanical drive to gear up from the cranks to the generator, complicating things.

Getting a controller to act as a boost switched mode controller with a variable output (to charge the batteries) is easy enough, just use regen, preferably with one of the controllers that allows a regen throttle, so you can use that input for control. You should be able to get a combined pedal-drive-generator-regen controller efficiency of around 80% without too much hassle, I'd have thought.

The drive part would be standard, just a battery, controller and motor. The efficiency of this would probably be around 80 to 85%, if decent batteries with a low Peukert number are used.

Overall efficiency would then be the product of these two figures, so around 64% to 68%, reasonably close to your 70% estimate.

Efficiency isn't everything though, as I believe we discussed in another thread here some time ago. The ability to put more power in than is taken out, when going down hill, riding slowly, or even when stationary (if it's a trike) would give a net gain in effective efficiency, so, depending on the sort of places you're likely to ride you may well find that it's more practical than it at first seems. Certainly there are some attractions to getting rid of the chain, the main challenge to overcome is getting a clean way to drive the generator fast enough from the pedals.

mytach said:

Thanks, amberwolf. I did a quick search and found several threads with similar topics. LowRider did a velomobile style trike with a series hybrid system. Was that who you were talking about? Looks pretty cool.

Click to expand...

This one comes to mind: http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=3423

I expect you already have this reference: http://www.hupi.org/HPeJ/0015/0015.html

Miles, yes, I did see that thread and I read through it briefly. The article by Andreas Fuchs on hupi.org was the original catylst for the project. I had considered a parallel hybrid, but I didn't like the complexity. The concept of the series hybrid was floating around in my head and I found that article in some of my research, which confirmed that I wasn't as crazy as I thought (at least not in that area ).

Jeremy, yes it will be a trike and the situation that you described fits what I'm hoping to do perfectly. This is actually an interim step toward a velomobile style version. But I figure I need the experience of actually getting my hands on the parts. I'm a mechanical engineer by profession, so I like to hold things in my hands. The pedal generator RPM is definitely an issue. However, because I couldn't figure out an easy way to mount the crank to the hub motor (it would be turning the axle not the casing), I decided to mount the motor slightly above the crank and attach it with a small chain. A few additional percentage losses, but worthwhile to bring the motor RPM up to the appropriate range. I'm thinking a 1:5 ratio should work well enough.

Sn0wchyld, I think in a typical motor as generator set-up you would be correct, the voltage would vary as the RPM input changed. But with a BLDC direct drive hub motor with built-in regen, the controller must take care of boosting the voltage to the battery pack level (or slightly above) for regenerative braking to actually work. I'm hoping to use that controller to make my life easier (no DC converter or anything). I'll have my fingers crossed just in case though.

Thanks for the feedback and second opinions on the endeavor. It's too bad I have to wait a few weeks for the parts to get here (and for my birthday in June to so I can talke my wife into giving me the Trike I want).

Myles

Sounds a reasonable arrangement. Gearing up the generator will be a good idea, but if you're using a hub motor as a generator then I'd be careful to not gear it up too much, if it's running at around 300 rpm at a comfortable cadence for you (maybe around 80?) then that should be about optimum - spinning it faster will increase the losses a bit, both in the generator and in the step up drive. I'd suggest between 1:3.5 and 1:4 would be about right, 1:5 is probably a bit on the high side.

You only need a tiny controller for the generator regen, a cheap 6 FET one would be more than adequate, as it's never going to handle more than maybe 250 watts or so, and will be running at around 80 to 100 watts much of the time that you're pedalling (unless you're super-fit). The 6 FET Xiechang will regen OK, and is as cheap as chips, too, so would probably make a pretty good "boost rectifier". Over charge from the generator isn't going to be a problem, I'd have thought.

The rest is just straightforward ebike stuff, from the battery through the motor controller to whatever motor you decide to fit.

Thanks, Jeremy.

But I don't think I need to create a simple regen controller. The hub motor that I'm planning to purchase has a programmable regen controller already built in to the motor. So if the circuitry exists in the controller to "fix" the power coming back out during regen, then I really shouldn't have to do anything but hook it up to the battery like it was a regular motor. Basically, I'm just tricking the generator into thinking it's a regular motor but it's ALWAYS running in regen mode (because I'm power it via the pedals).

Myles

Myles :

Here some information how to use an BLDC motor plus a tipical controller (3 phase mosfet bridge) in regen mode :

http://ww1.microchip.com/downloads/en/devicedoc/regenerative braking of bldc motors.pdf

I agree the total efficency will bee poor, but it is interesting when you want to use this motors as part of a wind turbine.

Regards

Rodrigo

http://endless-sphere.com/forums/viewtopic.php?f=3&t=44331

I couldn't find the older build from here on ES. It was a young man with a BMX configured as a recumbent. The generator was in front of the steer tube, and when he came to a stop at a street light, he would keep pedaling.

One of the problems in my view, is that when ever energy is being transferred from one place to the other, some of it is converted into heat (warm wires). So, you lose a lot of your generated power from the pedals to the battery, and then you lose a big chunk again when you transfer the power from the battery to the motor.

In the generation of power, RPMs count. Thats why significant and well-designed wind-turbines generate high-voltage, and they do that at very high RPMs. The best possible wind-generators have a lot of in-efficiency, but when you start talking about lower voltages and lower RPMs, the efficiency gets exponentially worse.

The majority of electricity coming in from the grid to charge a battery is usually generated by very-high RPM/high voltage steam turbines. You are trying to get efficient generation at low RPMs and low voltage. Even though I admit the distance from your pedal generator to the battery is short, the engineering model is very bad choice possible.

hi! you still here?!

sorry for the necro-posting, but i'm just about to build a pedal-power generator around a hub motor and am trying to establish best method of controlling charge - i wondered how your work with brake regen as a charge circuit went and if you have any insight to offer?

https://endless-sphere.com/forums/viewtopic.php?f=41&t=102834

This guy also uses a motor generator for ease of pedaling: https://endless-sphere.com/forums/viewtopic.php?f=6&t=100348