The Great "Gearing vs Hub Motor" Debate

One Horsepower - Fixed or Geared?

This is a generic 750 Watt Unite motor running at 36 Volts and 30 Amps on a bike with very good aerodynamics. These two charts show that gears really do give an advantage, but not in top speed. The real advantage is in hill climbing torque. If the fixed geared bike is geared to attain the maximum top speed it loses low end torque. If the fixed geared bike is geared for torque it loses top speed. So the chart for the geared bike shows how much more torque is possible (in first gear) compared to the fixed gear assuming top speed is mandatory.

:arrow: Gearing will always dominate whenever a FIXED POWER (one horsepower in this case) is the defining parameter.
 

Attachments

  • one horsepower fixed gear.gif
    one horsepower fixed gear.gif
    11.2 KB · Views: 2,498
Safe said:
Gearing will always dominate whenever a FIXED POWER (one horsepower in this case) is the defining parameter.
Yawwwwnnnn...and here we go yet again:
"dominate" what?
dominate in the area of reliability?
dominate in the area of ease of installation?
dominate in the area of high-power, 2-5kw installs on bicycles?

I know what you mean...but fact remains gears dominate hubmotors in some areas, and in some environments, and under some specs, but not others. Hence the reason this thread is 27 pages long. Hence the reason hubmotors are popular.
 
CRMRCYL_TH.JPG
:!: The 2007 Electric Bike Awards :!:
CRMMCL_TH.JPG



:arrow: First Place "Performance"

The PMG 132 wins in performance in the unlimited power category. There is no motor on the market that can beat it, but riding this legally might be a problem in most areas.

fullbody%20right.jpg


:arrow: First Place "Most Practical"

The 5304 Hub Motor (and others) wins in this category because it has a good mix of overall performance and also has low maintenance.

Dual%20Speed%20Hub%20Motor%20Kit%20FP.jpg


:arrow: First Place "One Horsepower Class"

The Transmagnetics Motor combined with an 8 speed internally geared hub would yield the ultimate performance in the one horsepower class and would be legal in most states. (note the bike below is an example of a geared bike that did 58.6 mph) This type of bike doesn't even exist yet except in the imagination. (the one pictured does, but the "ideal" doesn't)

machine_x_bonneville2.jpg
 
Any more info on that first bike? That would be a great combo if it could be made a bit more bicycle looking to fool the cops.
 
That's where I've seen it before, thanks for the link. I'm still laughing at the 15-25 mile range spec though... 960Wh of lead will give about 600Wh at 20 amps, and even less at high loads if you pull 10+kW.
Put enough lithium on the bike to make it useful, and you're looking at $6000+ :(
 
To my simple mind this is what I've concluded from the great debate of gears vs fixed.

:arrow: Hub motors have a useable wide bandwidth,
so they can get away without the need for any gears.


:arrow:Gears do help out with efficiency but at most all you need is two.

(no need to read any further, the rest is just there to back-up these statements).


Actually this holds true for most electric motors, but BL hub motors are optimized to be broader than most.
By bandwidth I'm referring to their relatively flat, broad power band as shown in this graph.

example_hub_motor_771.gif


Bandwidth is defined as the the two end points at which the power level has dropped by 50% off it's peak point. This applies equally to any kind of power, audio power from an amp, optical power from an LED or LASER, etc.

In this graph, ignoring everything except the black power(out) curve, the peak is 80, so cutting off at the 40 end-points it ranges from (I'm presuming 5mph div?) 12mph to 47 mph giving a useable bandwidth of 35mph. (mph being a stand-in for wheel rpm)

Now compare the bandwidth of this motor with gearing.

example_big_motor_and_chain_143.gif



It's bandwidth is 45 mph, a gain of 10 mph or almost 30%. Not at all insignificant, but not earth shattering either. The hub motors curve is already pretty close to that perfectly flat ideal, but not completely flat. Anyone who has driven an ebike for any length of time can relate that this statement is perfectly true:
safe said:
If the fixed geared bike is geared to attain the maximum top speed it loses low end torque.

If the fixed geared bike is geared for torque it loses top speed.

The obvious conclusion is that all that's required is a hi/lo gearing setup.
Because the hub motor is flatter (compared to something like an ICE) or basically it's flat enuf, at most all you need is two gears to shift that broad band either up or down slightly & you've got the total rpm range covered. Anything more than two is redundant as is borne out by some of the best engineering minds that have determined that the Tesla only come with two speeds.


Having two gears would be nice to have but the question becomes, how much is that extra 30% worth & what does it cost to achieve it. Especially from the manufacturers standpoint where the upper limit of the bandwidth is capped at 20mph by regulation. With that cap the gearing advantage drops down to only 5mph & that remaining bit can be somewhat made up for by shifting the power peak downwards with modification of the controller & motor design. To a manufacturer building the hub motor oversize to compensate is preferable because it saves steps in the assembly process with a lower total parts count & that saves money all along the production & distribution chain. Fewer parts mean fewer things to go wrong & less parts to stock.


The way I look at it is that gears are not desireable, they are a necessary evil. You include them only if you can't get the job done without them. The performance requirements for an ebike are so modest that the powerband of a hub motor ju-u-u-st barely stretches wide enuf to cover the full envelope as is, without gears. Unless you do a lot of hills, you spend such a small percentage of time of your overall driving profile in that lower portion where it's deficient that it's a reasonable compromise to save on cost, weight & complexity. Even for hills, the design of the motor can be optimized as the various flavours of Crystalite do this.


In eletrical terms I view a transmission as an impedance matching network like an electrical power transformer, as are any of the basic machines such as the lever. They convert something you have in abundance to something you need which you are lacking. Typically for a small motor you have plenty of rpm but not a lot of torque, so gears match the load to the power source, converting its high rpm/low torque to high torque/low rpm at the output of the transmission.


Building a hub motor oversize to adequately cover all the bases is cheaper/simpler than a small motor with transmission to do an ideal job of it. Yes, the oversize hub comes with an efficiency hit, but put it this way;

:arrow: If you're able to move a heavy rock by muscle power alone you don't automatically reach for a lever every time, even though it would be considerably easier & more efficient to do so.



The great debate seems to be geared towards (pun intended) an unreality of getting out the absolute ultimate 'ideal' in performance which may be the case in the rareified atmosphere of motorcycle racing but not for the majority of typical users. In that situation multiple gearing of an electric motor does serve a purpose.

Like an ICE, the human motor is very 'peaky', ie. narrow bandwidth. What the 21 or 27 gears do for you is shift that very narrow power band up & down across the rpm range which reqiures a lot of ratios to cover the whole band with a bunch of narrow power band 'slivers'. An Electric motor with its wider power band obviously doesn't require as many ratios.

'Safe' has come to realize that a transmission designed for the human motor is unsuitable for an electric motor & needs a wider spacing. The wider spacing is an acknowledgement of the electric powerplant's wider bandwidth. His transmission has dropped from 12 gears to 6 & if he keeps going, he will ultimately arrive at the same conclusion everyone else has; that you only need two speeds, a high & a low for puttering around town.

Now for racing there is benefit for some additional gears. The power band of the hub motor while relatively flat is not perfectly flat, there is a peak point. Gearing would shift that peak along over the rpm range giving you an ever finer 'ripple' of multiple peak points the more ratios you have. But that peak is shifting largely in the region that was already covered within the 50% cutoff of the fixed drive so the benefit is minimal, but yes very real.

To gain that small benefit, the question again at what cost, how many gears do you want to carry around? The ultimate for a perfectly flat power curve would be a CVT. Even at that, the difference in performance is primarily in efficiency & the (relatively) small difference will manifest after hundreds of laps in a race where small differences make all the difference in winning by a narrow margin. How many of us need ultimate performance all the time?

For day-to-day riding, the efficiency gains of even a CVT hooked to an electric motor may turn out to be just barely noticeable. I fear that people might have too high an expectation of the performance enhancement a CVT will bring to an electric drive system. The electric motor is pretty good already on its own. You're largest gain in performance comes after adding your first gear ratio. After that it's all diminishing returns.

That is my summation of 28 pages of dialog passed thru my own filter of understanding & experience. I've been waiting for someone smarter than I to step up & do it. Didn't look like that was gonna happen & don't need it drag on for another 28 pages. Forgive me if it seems like I'm dumbing it down as that's not the case. Like I said, it's how my simple mind makes sense of it.

Have I got it strate?
Is the debate over?
Do I 'win'?

Whaddaya say we lock this thread & throw away the key?
My two gears worth.
 
Whaddaya say we lock this thread & throw away the key?

You just want the last word, don't ya Turbo? :)
 
Toorbough ULL-Zeveigh said:
Building a hub motor oversize to adequately cover all the bases is cheaper/simpler than a small motor with transmission to do an ideal job of it. Yes, the oversize hub comes with an efficiency hit, but put it this way;

:arrow: If you're able to move a heavy rock by muscle power alone you don't automatically reach for a lever every time, even though it would be considerably easier & more efficient to do so.

:arrow: I actually agree with you for the most part.

My conclusion is that if you have the opportunity to do "whatever you want" that the smartest thing to do is use something like a PMG 132 and simply run it lean. A big motor running on low current gets great performance and has a wide powerband and plenty of torque. To reach a speed of 50 mph you need about 3 horsepower. "Ideally" all electric bikes would be 3 horsepower with a fixed gear. (hub motor or not is a separate issue)

:arrow: But the way the laws will likely ACTUALLY play out is that the upper limit for power will be one horsepower.

So I'm VERY TEMPTED to simply buy a PMG 132 and run it at 3 horsepower (which is actually a legal electric bike in Missouri) and smoke past everyone else. Whether I'm going to cough up $1000 for the motor and another $400 for an Altrax controller is another matter however. I personally am going to simply "compromise" with a 1200 Watt motor and a 5 speed which achieves almost the exact same numbers as the PMG 132 and this only costs me $90 for the motor and $30 for the controller. But in the future I might go for the PMG 132 and the fixed gear.

But my second project.... my "living in the legal real world" project... will attempt to exploit gears and a legal limit 750 Watt motor and pedals and see how far that technology can be pushed.

If there are no bureaucrats making laws then a big fixed motor is better.


The "golden rule" seems to be something like either 2 to 1 or 3 to 1 more power is required for the fixed gear to equal the multispeed. So we basically agree that "somewhere" around a doubling of power is required for the fixed gear to equal the multispeed bike. (or thinking in reversed terms you need 2 gears if you use half the power of a fixed geared bike)

Geared 1 Energy Unit = Fixed 2 Energy Unit

I think these charts really sum up the situation for the one horsepower bike.

Without gears this bike is really lame... (it's not the NUMBER of gears, but the gear ratio SPACING that matters... the "magic ratio" seems to be about 2 to 1 gearing from top to bottom... if you can do that with two gears then that would work)
 

Attachments

  • one horsepower 2 speed.gif
    one horsepower 2 speed.gif
    11.5 KB · Views: 2,935
  • one horsepower 8 speed.gif
    one horsepower 8 speed.gif
    11.5 KB · Views: 2,932
  • one horsepower fixed gear.gif
    one horsepower fixed gear.gif
    11.2 KB · Views: 2,932
Why do you care so much about what's legal? You do realize that no cop can tell whether you're running 750W or 2500W, right? Especially when you can use a powerful motor but a weak controller/batteries to stick to the 750W limit. They're not going to inspect your batteries or controller even if they knew anything about electric vehicles, which they don't.

I understand if you're trying to stay legal just in principle, for a challenge, and in fact I guess you are since you mention your state is less strict. But there's really little point in it.

Blinkers, taillights, etc., are another matter, as those are important if frequently driving with traffic, but limiting motor power seems unnecessary.
 
If it looks like a (750W) duck and sounds like a (750W) duck, it will be assumed to BE a (750W) duck...otherwise you're on you're own.....
 
I've stripped all three of my battery packs down due to heating on the cell tabs. Measuring voltage drop across the connections, I calculated an average connection resistance of 1.15 milli ohms, but the tabs must be heating up quite a bit as the heat shrink around the packs was melting where it goes over each tab. Once I solder copper braid over each tab I'm hoping to see a significant reduction in voltage drop.

While everything is apart, I reconfigured the 36v pack into a partial triangle to better fit inside the frame, and I'm raising the total cell count to 76-77 so it can be 'in boost' all the time 8)

At 4-5000W, the X5 hub gets warm, but nothing alarming, however with the extra power I'll be pushing when the packs are all back together, I'm expecting the hub to start getting hot and needing some flow through cooling. One interesting thing to note is that as I've increased power, my Wh/km and Ah numbers have improved, even though I'm getting to and from work in record times. (34km of mixed city/highway riding in 36 min)
The reason? Extra power lets me get behind cars and trucks more consistantly to draft them. Get behind an 18 wheeler and you barely need 1kW to sail along at 45mph.
 
jackatfsi said:
If it looks like a (750W) duck and sounds like a (750W) duck, it will be assumed to BE a (750W) duck...otherwise you're on you're own.....
But the batteries and controller are totally silent and almost always totally hidden, so you can neither see nor hear the 'duck'. You can see and hear the motor but the motor is only as powerful as its input, decided by the controller.

You can see if it's exceeding 20 mph, but you can't determine if it's 750W or not. Of course even 'legal' e-bikes would, I expect, exceed 20 mph if the user pedals, even though that's illegal (the motor is supposed to shut off if the speed above 20 mph).

Of course it's a stupid law.
 
"Safe" fancies developing a production enterprise of quasi-motorcycle ebikes:

Slow enough to be legal ebikes, fancy enough to be interesting, easily modified to do 50mph.

Kinda like pseudo-assult style firearms that need a receiver change to go full auto.

(As if cops wouldn't look twice if they saw one on the street.)

:?
 
TylerDurden said:
"Safe" fancies developing a production enterprise of quasi-motorcycle ebikes:

Slow enough to be legal ebikes, fancy enough to be interesting, easily modified to do 50mph.

Yeah, that's right.

I've got dreams of developing a sort of micro-road racer that has the power of an electric bicycle, but looks kind of like a motorcycle road racer. Imagine one of these with an electric motor:


IGSA7s.jpg


MadMac.jpg
 
Are there bicycles that allow prone-position pedaling (a face-down recumbent like in the photo above)?
 
My theory is that for hills you stand and for straight aways and downhills you sit behind the fairing. Since aerodynamics while hill climbing is small you can stand up and extend your legs well outside of the fairing without fear of wind resistance.

I built this concept first in 1980 (remember that magazine article?) and it definitely works.

Most of the time you run on the motor... you only pedal on the hills...
 

Attachments

  • pedal road racer.gif
    pedal road racer.gif
    5.7 KB · Views: 2,945
Miles said:
To me, it still makes more sense to pedal most of the time and only run the motor to help maintain speed on the hills...

If your goal is "aerobic exercise" then pedaling is what you want to do. But from the perspective of "getting from A to B" there's no point in pedaling because it just means you expose yourself to more wind resistance and thus more power losses. By staying behind the fairing you cut enough wind losses to equal all the effort you put into it. All that pedaling is essentially a "waste of effort".

There's a stronger argument for something like a recumbent because then you get the advantages of low aerodynamics AND you can still pedal while you are doing it, but the negative is that the position places you on your back and near the ground. You lose the "fun" when you lose the forward facing position too. The "road racer" position is what makes sport riding fun and it allows for better reaction times to road hazards. Plus, there are times when I just sit upright and don't care about wind resistance or performance and just enjoy the road. It's comfortable to just "ride around" with the "road racer" position. And another "secret" is that it's a good way to develop your upper body strength. On an upright bike (like a mountain bike) you don't use your arms very much, but on the "road racer" you support a lot of your weight with your arms. The result is that you get as much of a workout with your arms as your legs.

:arrow: It makes you look more "Arnold"... :wink:

But in England with a 250 Watt limit on power I can see how pedaling makes sense. I'm running a 750 Watt that actually pumps out 1.4 horsepower and at speeds of 40 mph it's all about aerodynamics. The small 0.25 horsepower that the legs could add sort of falls off as insignificant. With a 2-3 horsepower motor (like the 5304 or the PMG 132) the human power becomes meaningless.

 
If you want to go 30-40mph, and have up to 1.4 horsepower available, then why bother with the "small 0.25 horsepower that the legs could add" on hills? Unless you want the "aerobic exercise"...?:wink: It's not going to make that much difference to the speed.

Yes, having a 250Watt / 16mph limit does change the equation, somewhat :)
 
Miles said:
If you want to go 30-40mph, and have up to 1.4 horsepower available, then why bother with the "small 0.25 horsepower that the legs could add on hills?

:arrow: Because our American National Electric Bike Definition defines an electric bike as:

1. 750 Watts.
2. Must have pedals.
3. Cannot exceed 30 mph on motor power alone.

If you calculate the top speed for EXACTLY 750 Watts it gives you 30 mph with an aerodynamic bike. The extra hill climbing ability would just add a litttle extra and make the pedals have some function. What I'd do is sell the thing with a 20 amp controller (that would give exactly 750 Watt peak output) and let people change the gearing and even run a higher amped controller in areas that it's allowed. The idea is to build a "base platform" that people could modify in the aftermarket area.
 
Back
Top