``` The Gearing Advantage ```

safe

1 GW
Joined
Dec 22, 2006
Messages
5,681
The Gearing Advantage

:idea: Let's start with a few assumptions...

First let's say that you are racing in something like NEDRA where the voltage is limited for your class and you have to run a DC motor of some kind. You've done everything possible to create or buy a motor that delivers good performance. You've decided that you want range as well so you've made certain that the efficiency peak is equal to the power peak in your motor. You did this because this represents the best that a motor can ever achieve. While there are any number of alternative configurations (like allowing a very high current limit) for this study we are just trying to establish a "baseline" upon which we can compare results. Every change that can be done to the motor simply adjusts our baseline... the final result should still be the same.

For this theoretical experiment I used a little tiny 250 watt motor and pumped up the voltage to a ridiculous 96 volts. The current limit is set at 30 amps. Are these the only numbers I could have used? No, but again we have to agree on some baseline and so for this case this is it. (this motor would most likely burn up at this voltage)

Looking at the chart below we see the difference in power output at the range of speeds that the motor can cover. It's top speed is something like 56 mph for this demonstration and so the fixed gear bike simply has the bike stuck in 8th gear all the time. The multiple speed bike has 8 speeds and can use the appropriate gear for any given speed.

The results are interesting. For the exact same motor there is an overall 20% power gain using gears as opposed to using a fixed gear. This does not even begin to get into the issue of ability to climb a hill which is obviously going to favor the multiple speed solution. I'm also not dealing with efficiency which also should point towards an advantage for the multiple speed bike.

These power figures are "power available" at those speeds... it doesn't mean you need to use it all, but it means you have it there for you if you want it.

A reminder... this motor was configured for maximum efficiency... so the overall efficiency could not likely be improved upon significantly. (so all the solutions that involve increasing the current limit just end up creating more heat and driving down the range)

:arrow: This is as they say: "As Good As It Gets".
 

Attachments

  • the gearing advantage.gif
    the gearing advantage.gif
    11.7 KB · Views: 3,032
Yet another "gears are great" debate?
We already have numerous threads very specific to this topic, like this 30-page monster you started:

http://endless-sphere.com/forums/viewtopic.php?t=150&highlight=gears+hubmotor+debate
Safe said:
After a lot of number crunching and modifications of formulas (most recently the PWM Current Warping effect) I think I'm at the point of being able to present definitive evidence that gearing does have advantages that non-geared machines do not have and for the same amount of power output (the legal limit) one can go more, climb higher and run faster than a similiarly equipped non-geared machine.

The debate begins!

I've said about all I can say on this topic already. It seems to me you have too.
 
:arrow: I've compared many things in many ways.

There's boosting the voltage.

There's boosting the amperage.

There's rearrainging the way the amps are delivered. (MCL)

There's introducing gears.

There's forced air cooling.

To really get the full flavor of any given technique you need to find the right way to slice and dice the data to really expose the benefits or disadvantages of any technique. This time I'm doing a straight forward comparison of identically prepared motors and looking at power. In previous comparisons I've looked at gearing range (steepest slope verses top speed) and things like efficiency. I've even done virtual drag races to compare the effect of gears or no gears.

So to my knowledge this is the first time I've focused entirely on power. There is a clear and simple 20% power gain when you add gearing.

It's not earth shakingly new news, but it's still a first to my knowledge... :oops:

But maybe I'm going senile and have forgotten... :lol:
 
yawn.
 
At this point it's "common knowledge" that gears are a "good thing". But a long time ago it was still somewhat controversial. What's really going to make this exciting is when someone really pushes all the limits and does all the right things and produces a bike that impresses people. People like to see things in practice and not just theory.

I'm seriously thinking of going DOWN in motor size (like the 250 watt class) and doing all this crazy stuff at once to see just how far the little 250's can be advanced. I really love tricking out small motors. Overvolting, overamping, MCL, forced air cooling, multiple speeds... if all of that is done at once it's going to open some eyes. (and it might wake up a few Europeans and Australians that have been living with their 250 watt motors running stock for too long)

Someday... :roll: (probably Project #003 the way it looks now)
 
Peak Efficiency vs Overall Efficiency

People often get overly excited over the "peak efficiency" number for a motor. The value of the "peak efficiency" is only useful when the motor is actually running in the exact rpm that the best efficiency actually takes place. For something like a Fixed geared bike the only time you get even near your "peak efficiency" is when you are up near your maximum speed. What about on a hill? Well... that literally "sucks" current and produces lot's of heat, the efficiency at low speeds is rather poor. One could improve matters a little of they "throttle fiddle" and drop the throttle so that the efficiency powerband drops and that will improve things a little, but that also means you are losing power. So the Fixed geared bike just can't win. :wink:

The charts shows how not only are we seeing more peak power across the entire range of speeds (from 0-55 mph in this rather extreme case) but the efficiency is better by an overall average of 20%.

With an advantage of 20% in efficiency any small losses due to the transmission (like 5% to 10%) become fairly trivial since you are after all getting an increase of 20% in peak power too. While the "monster motors" out there are so large that they can afford to ignore power and efficiency gains, for anyone using a smaller (750 watt size or less) motor the advantages of using gearing is pretty overwhelming:

:arrow: More Power - 20% better

:arrow: Better Efficiency - 20% better

"The Gearing Advantage" is hard to beat... 8)

In fact, with numbers like these it's very possible that a less efficient brushed motor could exceed the overall efficiency of a significantly more efficient brushless motor just on the gearing advantage alone. The "best of all worlds" would be gearing and a brushless motor combined. (and there are products that address this need)
 

Attachments

  • the gearing advantage (eff).gif
    the gearing advantage (eff).gif
    12.1 KB · Views: 2,914
The Worst Motors Improve the Most Dramatically

Contrary to what you might think it turns out that the lower the efficiency of the motor configuration that you start with the greater the advantage that gearing produces. I was kind of shocked actually and expected the results to be less favorable. So my guesses are sometimes wrong, oh well, it ended up making my overall argument even stronger. :)

:arrow: So what is this?

This is a generic MY1020Z3 running at 36 volts and a standard (BCL) 40 amp controller. The top speed for such a vehicle using some mildly aerodynamic figures is 42 mph. So these charts represent how the overall motor and gearing (or lack of gearing) behave as a unit in the real world at different speeds on flat land. The fixed motor is basically the same thing stuck in 8th gear all the time just like a hub motor would do.


24% boost in efficiency

28% boost in power


Impressive results?
 

Attachments

  • gearing advantage - 750W - 36V - 40A - 42mph (Watts).gif
    gearing advantage - 750W - 36V - 40A - 42mph (Watts).gif
    20.8 KB · Views: 3,217
Heat is also Reduced

It should come as no surprise that heat is also reduced. In this case the reduction is a full 33%. Only in first gear and on a steep hill might you have any problems with heat...
 

Attachments

  • gearing advantage - 750W - 36V - 40A - 42mph (Heat Watts).gif
    gearing advantage - 750W - 36V - 40A - 42mph (Heat Watts).gif
    21.6 KB · Views: 3,204
Less Stress on Rear Hub

As a result of spinning the motor faster (the result of having extra gears) the rear hub does not get the same high level of torque applied to it except in first gear. So as long as you are not in first gear you are not likely to damage a multiple speed hub even though you are getting significantly more power out of the motor. So it's another "win" only this time it's for the rear hub. A true hub motor doesn't really care about this, but for people wanting to use internally geared hubs it's good to know and have calaculated your loads. This is a very conservative design.

The Sturmey Archer 8 Speed has been designed so that first gear is the direct drive gear, so it can handle the highest loads... exactly what you want to have happen. So "it's all good" as the saying goes... 8)


P.S: I've got to add that I'm not accounting here for rider misuse. There's nothing to prevent a rider from riding in 8th gear all the time, so all of these charts assume a rider that has a reasonable level of expertise in knowing what gear to be in at any time. You can't prevent people from being stupid. (however, with the MCL controller concept you actually can "prevent the stupid"... so anything can be made idiot proof, but I'm not attempting to deal wth that here for now)
 
safe said:
Less Stress on Rear Hub
The Sturmey Archer 8 Speed has been designed so that first gear is the direct drive gear, so it can handle the highest loads... exactly what you want to have happen. So "it's all good" as the saying goes... 8)
[/b][/color][/size]
Well ignoring the fact that your statement is backwards (the motor wants to spin at 2-3000RPM and there is a 10:1 reduction to first gear) I still do not see the advantage over a high pole count motor. Rerating gives you an effective infinite number of gears.

Dan
 
cadstarsucks said:
Well ignoring the fact that your statement is backwards (the motor wants to spin at 2-3000RPM and there is a 10:1 reduction to first gear) I still do not see the advantage over a high pole count motor. Rerating gives you an effective infinite number of gears.

:arrow: You're new here and have not followed all the previous discussions.

This was basically in response to a source of anxiety long ago where people posed the question:

"Wouldn't an electric motor tear a multispeed rear hub to pieces?"

...at first no one here could figure out what the rear hub torque would be (the value that the chain itself introduces to the rear hub) and so people were saying that the electric motor would not work. Well, this shows that not only should this work (peak hub torques for some of the multispeed hubs are allowed to be as high as 100 Nm before breakage) but that the hub torque should be rather conservative. In order to measure the rear hub torque I've already accounted for all the gearing and use the highest value BEFORE the hub. If you were "into" bicycle hub terminology you would know that they tend to calculate the maximum hub torque based on first gear. The hub I'm using as an example is the Sturmey Archer 8 Speed hub. It's first gear is direct drive and all the other gears go up from there. In effect I'm taking the "worst case scenario" of the hub torque ENTERING the hub and not how it exits in the form of rotation of the rear wheel. The torque will obviously go down if the gearing goes up inside the hub. (pretty smart those Sturmey Archer people huh?)

In essence the point here was that you actually apply less torque on the rear wheel compared to a fixed geared motor/chain/rear sprocket combination. Dirty D on his fixed chain motor machine will apply more torque to his rear wheel... however... since he has no gears to break it really won't matter. The bottom line is this issue is nice to know, but not as earth shaking as the power, efficiency and heat revelations.

XRK8_A.jpg


All DC Brushed and Brushless motors have a defined powerband that has good efficiency at high rpms and poor efficiency at low rpms. Using gears in effect "corrects" for the flaw in the powerband of a DC motor.

The exception to the DC Brushed and Brushless is the Induction motor (discussed on another thread) which is sort of "upside down" comparatively and gives exactly the "right stuff" for what we want. This is the reason that the Tesla Roadster uses an Induction motor and NOT a typical DC motor.

There are some weird "hybrid" designs floating around that are trying to do things like alternating the firing order of the coils in order to simulate gearing. So even within the "electric only" crowd there is an admission that there is a problem that needs fixing.
 
Whooda thunk a fertilizer-spreader could learn to type?

Safe, the powerband of an electric motor is dependent on the design of the motor: efficiency and power can be designed for any RPM.

"A brushless motor can properly operate at very low speed, too. The lowest speed which can be achieved is exclusively defined by the resolution of the seek sensor you are using: with an std encoder at 4096 impulses per revolution, 16000 position per revolution can be resolved and the rotation is uniform even widely under 1 r.p.m. Generally speaking, the lowest speed allowed for the rotation to be kept perfectly uniform, is the one where the frequency of the encoder exceeds the passing band of the system, usually 30-50Hz."
http://www.rcv-srl.it/index.asp?id_sezione=18&id_lang=2&random=vero


"Kollmorgen GOLDLINE® Direct Drive Rotary (DDR) series motors offer a high performance, zero maintenance servo solution. The Kollmorgen GOLDLINE® DDR achieves very high torque density through the combination of large diameter, short length, and a high number of magnetic poles.

This DDR can be used as a flexible indexer, providing programmable, rapid, indexing far exceeding the throughput of conventional mechanical or variable reluctance technology indexers. It can also be used for replacing mechanical transmissions such as gearboxes, timing belts, and rack and pinion reducers.

5.3 to 339 N-m (3.9 to 250 lb-ft) continuous torque accommodates a wide range of application requirements
Speeds up to 500 RPM meet most high torque - low speed application requirements
Power ratings up to 2040 watts meet high power demands of most applications
Rugged multi-speed resolver feedback provides reliable precise operation
Optional high resolution Sine Encoder feedback devices provides exceptional accuracy and repeatability below one arcsec
Direct load connection eliminates maintenance of gearboxes, belts, or pulleys

Zero backlash and compliance provides more responsive system performance
Dual bearing option available for high moment load applications
IP65 and 67 shaft sealing offers resistance to dust and water jets
CE compliant, UL recognition for global acceptance
Built-in thermostat for over-temperature protection
Provides stiff mechanical platform for attachment of indexing wheels


http://www.danahermotion.com/products/product_detail.php?parent_id=23

:roll:
 
safe said:
cadstarsucks said:
Well ignoring the fact that your statement is backwards (the motor wants to spin at 2-3000RPM and there is a 10:1 reduction to first gear) I still do not see the advantage over a high pole count motor. Rerating gives you an effective infinite number of gears.

:arrow: You're new here and have not followed all the previous discussions.

I see... it makes sense that there might be a 1:1 from a pedal but not from a motor. It is easier to make a gear motor than a multipole and, as I recall, your pics have been of gear motors so you have, in effect, two transmissions between the rotor and your wheel.

While transmissions can help particular motors by allowing them to run where they like to, they also add load to the motor.

Dan
 
Did you read the technical parts of the first link?

"The torque motors, or low speed motors, are standard motors developed with particular windings having steady high Ke and Kt. In order to fully understand the potentiality, take as an example an "ideal" brushless motor with yield equal to 1 and cos = 1 (practically, good approximation). Under these conditions, as the motor is fitted with permanent magnets and therefore in constant field, the tension to the head of the motor is proportional to the speed through the Ke constant..."

They eventually get to their "bottom line" which is that they have produced a motor that works best at 300 rpm rather than 3000 rpm. They've essentially done the same thing that a hub motor does. No surprises... they've just eliminated the need for a gearing down of the 3000 rpm down to 300 rpm. They've eliminated the fixed gear part, but not addressed the powerband part.

So you have presented an example that doesn't really change anything. For ANY DC Brushed or Brushless motor the formulas remain the same. The only way to really change anything is to have a dynamic permanant magnet (an "Inductive" magnet) so that you can advance or retard the "timing" of the motor. That's how Tesla brought about the "Second Industrial Revolution" and changed the electrical world forever. (look up "the Second Industrial Revolution")

:arrow: The Bottom Line:

You haven't proven anything... and certainly haven't disproven what I've presented about the standard DC motor...
 
safe said:
All DC Brushed and Brushless motors have a defined powerband that has good efficiency at high rpms and poor efficiency at low rpms. Using gears in effect "corrects" for the flaw in the powerband of a DC motor.

:roll:
 
TylerDurden said:
This 300 rpm motor is cool :p (or something to that effect)

I think you've gotten confused as to what that motor does. It does have it's peak efficiency at 300 rpm, but it only revs to about that same amount. You can't run those motors up to 3000 rpms, so what you are thinking is "low rpm performance" is really all the rpms you get.

So that's the "gotcha" that you didn't see... :wink:


Back to "The Gearing Advantage":

gearing_advantage__750w__36v__40a__42mph_eff_percent_103.gif


gearing_advantage__750w__36v__40a__42mph_watts_103.gif


gearing_advantage__750w__36v__40a__42mph_heat_watts_158.gif
 
What you are failing to understand, is that electric motors may be built for power/efficiency at any specific RPM.

Gears MAY or MAY NOT be an advantage, depending on the specific motor and specific application.


:lol:
 
What you are failing to understand, is that electric motors may be built for power/efficiency at any specific RPM.

exactly, that is the problem at any specific RPM, vehicles need to operate at a wide range of speeds so either the motor needs to be efficient over a wide range of speeds, or very efficient at one speed with gears.
 
Dirty D you do seem to "get it" very quickly on most topics. 8)

Yes, you can go and buy a motor that works well ONLY at 300 rpm. You can also buy a motor that works well at ONLY 3000 rpm. It's trying to deal with the "real world" of wide ranges of speeds that things get more difficult. When you add multiple gear choices to a motor that otherwise has a narrow desired speed you effectively expand the width of the powerband. When you compare the hub motor to the multispeed bike you get all these really cool advantages as far as power with the gears.

The actual benefits of gears depend on the specifics, but on average you can expect about a 20% overall power improvement, a 20% efficiency improvement and a reduction in heat creation that's going to be at least 25%.

Gears are worth their effort to add... they "add value" in large enough numbers to compensate for any frictional losses they might introduce.

But hub motors are still lower maintenance, so they do have their place. All the complexity of gears makes life more difficult. I mean can you imagine a bicycle with gears? Absurd! No one would ever want to bother with such things. The only bike a man ever needs is a fixed gear bike... no freewheel, that would too much of a luxury. Maybe a coaster brake? And do people really need a seat? No! Go without it. And handlebar grips and tape? Why those are all decandent luxury items too! Maybe the ideal is a unicycle? And on and on and on... :lol:

Compared to a motorcycle the bicycle is still a model of simplicity. Your ability to handle a more complex machine is a personal choice. Some people (like myself) can handle the extra "tech" and have fun toying with it. Others will see their bike encounter it's first problem and throw their hands up in the air and give up. Some people own motorcycles and don't even know how to tune them, preferring to hire a mechanic to do their repairs. So it comes down to what you personally can realistically handle. Do you like "tech"? If so, then maybe gears are a good choice for you because they add performance.
 
dirty_d said:
What you are failing to understand, is that electric motors may be built for power/efficiency at any specific RPM.

exactly, that is the problem at any specific RPM, vehicles need to operate at a wide range of speeds so either the motor needs to be efficient over a wide range of speeds, or very efficient at one speed with gears.
Which is why I like rerating... buy for low speed and "cheat" for high :)

Dan
 
cadstarsucks said:
Which is why I like rerating... buy for low speed and "cheat" for high :)

What you say "sounds" like gibberish. What are you saying? If you have a motor that is designed to run optimally at a certain speed you can't do much to change that. It comes down to the design of the motor. Most DC motors use permanant magnets and the magnets themselves determine the preferred operational rpm. That's the "big concept" about Tesla and the Induction motor because that motor design does not use permanent magnets at all and instead "induces" a magnetic field in what is basically just a hunk of metal. The Induction motor's only problem is that it needs to fully saturate the core that its trying to magnetize and there are difficulties doing that well on small motors. (something I'd like to learn more about)

:arrow: Do you "get" what I mean here?

The permanant magnets "define" the motor. You can't do anything about the magnets once the motor is built.

Get it? Got it? Gunna get it? Got it yesterday? Will get it tomorrow? :wink:

Or are you here to goof off?
 
safe said:

The actual benefits of gears depend on the specifics, but on average you can expect about a 20% overall power improvement, a 20% efficiency improvement and a reduction in heat creation that's going to be at least 25%.

Power is an instantaneous measure. Your charts in this thread and your statements such as above claiming a ~20% gain are based on integrating the power curve over rpm's, and hence are not comparisons of extra available power. As such, these statements are grossly misleading. At any single point on the curve, the idealized geared system may be 5% more powerful, or 50% more powerful, depending on the chosen point.

As I showed before via the hubsimulator, a 406 at half throttle and half speed is about 6% less efficient than the same motor at full throttle and full speed. For most riders on most fixed gear systems over most terrains, the time spent at very low throttle settings and very slow speeds is minimal. Hence, in real life there is nowhere near 20% efficiency to be gained by gears over the full powerband. More like 3% if you can implement a very good geared system. Not worth the trouble for any but the most impractical DIY'ers like yourself, or riders with 'special needs' like 25% grade volcano climbers.

406_80v_pt_908.jpg

406_80v_908.jpg


Beyond that I'm pissed at myself because you (damn bastard) managed to suck me into yet another one of your never-ending evangelizing-gearhead threads. :roll:
 
safe said:
What you say "sounds" like gibberish.

That is because you are an idiot.

The method he refers to was discovered by a real engineer with a real PhD, from a real university. NOT like your "Physcal Sciences" Assoc. , from some mail-order diploma-mill.

There is real math behind the re-rating process, unlike your bogus spreadsheets that leave-out half the needed elements like iron losses and switching losses, let alone simple labels on yer friggin axes.

:roll:
 
i don't know why everyone is getting so hostile, but what safe is saying really does make sense, the extra power to be had manifests itself as more torque at the lower vehicle speeds its pretty much common sense.
 
xyster said:
At any single point on the curve, the idealized geared system may be 5% more powerful, or 50% more powerful, depending on the chosen point.

One assumes that they WANT to get more power. So at full throttle (the only throttle setting that matters when full power is the goal) you do get more than the overall average of 28% improvement at times as you climb up in speed. Overall the net gain is a clear and simple 28% (for the MY1020Z3) but on an instantanious level you are correct the value can be much higher. (about double the power at some rpms) That's why you can visually read the chart and see WHERE the performance enhancement is and where it isn't. You can even see the fine shifting points where the advantage varies slightly.

I don't know it it can be presented much better.

:arrow: The "partial throttle" argument for efficiency doesn't hold water because the goal is to go fast. Would an Indy Car seek to limit their throttle while trying to increase speed? Definitely not. So you have to pretty much throw out the "partial throttle" thing as "neat" but "does not apply" to the power and performance issue.


gearing_advantage__750w__36v__40a__42mph_watts_103.gif
 
Back
Top