Conclusive proof gearboxes are awesome.

[motomoto]

Ok Ok I lost this argument. I am taking the rear sprockets off of my Bafang BBSHD setup and putting on one sprocket
for my trail rides.

 

[spinningmagnets]

ICE powerplants use transmissions like crutches to help them limp around with devices that make 0torque at 0rpm

I think the bigger the motor, and the more Watts the battery/controller can provide...the less gears are helpful (and then become added expense, wasted drivetrain volume and weight, un-necessary added complexity). If we construct a race-series with specific restrictions, we can tailor the rules to give a bias to either non-hub with a transmission, or a DD hub. Using real available equipment, lets compare a MXUS 3000W 45mm wide stator against a BBSHD with a 3-speed Sturmey SX-RK3.

If both are restricted to using identical 48V X 30A = 1440W battery packs, the BBSHD provides more range for the same watt-hours, and if the course includes a steep uphill, the lack of transmission will make the MXUS get hot. Non-hub with transmission wins.

Bump up the power to 5000W on a street course, and if both have the same top speed, the BBDHD will overheat trying to keep up with the MXUS.

If we upgrade the 3-speed trans to a Lightning Rods "big block" at 5000W (both having the same top speed), it easily keeps up with the MXUS without overheating (3-speed trans not really necessary on flat land), but it costs more and and has lots of complex bits in the dual reduction. Big and simple DD hubs remain popular for street ebikes that want to keep up with car traffic (2600W+)

A few years ago, the selection of motors and batteries was much worse. Now there are choices. If a builder has the budget for a big motor and a high watt output battery, and he's running it on the street...I agree on no transmission. But...if you limit my budget to a 1440W battery/motor/controller...then throw in some steep hills...I will want a transmission in spite of the losses.

 

[spinningmagnets]

But why dont the flagship production E bikes, Zero, Lightening, etc,....use hub motors

Higher magnet speed (in the motor) in relation to the wheel RPMs, plus makes the unsprung weight much better for the suspension response. Centralizing the weight also helps cornering.

However, the discussions here are transmission Vs one-speed.

 

[speedmd]

In the tesla example, if you were to make the motor roughly 3 times the diameter, you would have roughly the same low end torque in a direct drive setup. It should also be able to rev and pull way past the current cars top speed with the same technology. No?

 

[nutnspecial]

Nah don't give up Moto (and def don't go single speed on your mid jajaja), we're being productive! Determine terms of argument (which is not simply for/against 'proof gearboxes are awesome' imo), and we get alot to think about! Just in this 2016 'reboot' alone there seems to be very compelling and thought provoking arguments, hopefully continuing to refine and define!

Well said SM, this is the same perspective several of us (me2) have been leaning towards imo. . . .

I'll butt out, as further argument re:teslaVS will likely be technical and complex beyond my current ability to clearly define/express/discuss, for now.

 

[Hillhater]

In the tesla example, if you were to make the motor roughly 3 times the diameter, you would have roughly the same low end torque in a direct drive setup. It should also be able to rev and pull way past the current cars top speed with the same technology. No?

Maybe, if 3 times the diameter can give 8 times the torque ?.....
...but the weight increases disproportionately, and a big motor inevitably needs a bigger battery ( more weight again) , so unless you actually can produce more torque than the geared motor, you will lose performance

 

[Hillhater]

But why dont the flagship production E bikes, Zero, Lightening, etc,....use hub motors

Higher magnet speed (in the motor) in relation to the wheel RPMs, plus makes the unsprung weight much better for the suspension response. Centralizing the weight also helps cornering.

Im asure that they are key factors, but i suspect there is much more to it,
IE, wouldnt the motor have to be bigger and heavier than the high Rpm unit ?
Wouldnt the battery spec need to be changed to suit the hub motor ?
The overall package weight would increase.

...However, the discussions here are transmission Vs one-speed.

Is it, i thought we were debating the need for any transmission on an EV, single reduction or multigear ?

 

[liveforphysics]

In the tesla example, if you were to make the motor roughly 3 times the diameter, you would have roughly the same low end torque in a direct drive setup. It should also be able to rev and pull way past the current cars top speed with the same technology. No?

Maybe, if 3 times the diameter can give 8 times the torque ?.....
...but the weight increases disproportionately, and a big motor inevitably needs a bigger battery ( more weight again) , so unless you actually can produce more torque than the geared motor, you will lose performance

3x diameter does offer 8x torque. Think about it like unrolling the motor, and you see you have both a longer strip of stator/rotor interaction area, as well as applying that force onto a longer lever arm.

However, in the case of the Tesla, it seems like it would be a better option to eliminate having the torque get combined in a differential and instead have a 2x larger diameter independent motor per wheel. Having the differential in an oil bath is a very small loss when you're at peak power, but becomes meaningful at cruise power levels.

But why dont the flagship production E bikes, Zero, Lightening, etc,....use hub motors

Higher magnet speed (in the motor) in relation to the wheel RPMs, plus makes the unsprung weight much better for the suspension response. Centralizing the weight also helps cornering.

Im asure that they are key factors, but i suspect there is much more to it,
IE, wouldnt the motor have to be bigger and heavier than the high Rpm unit ?
Wouldnt the battery spec need to be changed to suit the hub motor ?
The overall package weight would increase.

...However, the discussions here are transmission Vs one-speed.

Is it, i thought we were debating the need for any transmission on an EV, single reduction or multigear ?

The battery has a reduced power demand because the drivetrain has higher efficiency. It's so long past time to be perpetuating the myth it uses more energy to be direct drive. For any power level, the motor that has the most copper and has the most well used iron and fewest added losses is the most efficient option for converting that electricity into motion.

A direct drive motor doesn't need to weigh even a tiny bit more than a tiny high RPM motor, it merely needs to use the electrical materials at the same rate as the tiny high RPM motor. This can mean the motor requires a very high pole count rotor if the needed RPM range is very low, which can lead to issues in manufacturing, and the structure to securely support the tiny thin stator teeth can get to be a substantial engineering project. On paper it should always be capable of being lighter to be single stage, in practice it's almost never done due to mfg and engineering related challenges to use the iron at the same rate some tiny motors at high RPMs do.

 

[Miles]

Ok Ok I lost this argument. I am taking the rear sprockets off of my Bafang BBSHD setup and putting on one sprocket
for my trail rides.

Obviously, you shouldn't. You are starting with a power limited motor which is probably itself not very efficient. Its speed is reduced through two stages (?) to match pedal cadence speed. Its speed is then increased again by the chain drive to the derailleur. You have a rather underpowered and inefficient set up but it does at least allow you to change ratios with little penalty.

 

[Punx0r]

I wish I could find it now, but I recall a past thread where a real life example of (I think) a hybrid car was given with two different gear ratios and the output torque plotted for each. The headline result was the “low” gear produced more torque, but only up to around 10mph, thereafter the “high” ratio increasingly dominated to the car’s top speed. The car already had more torque at low speed than the tyres could handle, so the “low” ratio was useless – as would be a variable ratio transmission.

 

[speedmd]

It was on field weakening. https://endless-sphere.com/forums/viewtopic.php?f=30&t=63467&start=50

a better option to eliminate having the torque get combined in a differential and instead have a 2x larger diameter independent motor per wheel. Having the differential in an oil bath is a very small loss when you're at peak power, but becomes meaningful at cruise power levels.

Yes, agree, much better to have independent motors for each side. Axial motor setup possibly. Getting rid of the reduction gear box would eliminate most of teslas current drive trains long term reliability issues.

 

[toolman2]

I have always heard that each stage of reduction subtracts efficiency. Why exactly is that? If a gearbox or jackshaft was setup to do say 10:1 in 2 stages vs a single stage with 10tooth motor 100toot rear, would the 2 stage be that much worse? Is it the addition of another shaft and bearings that kills the efficiency? I was under the impression that reductions with large size differentials between the driving and driven sprocket/gear are inefficient. So wouldn't the 2 stage have some advantage being that the driving vs driven sprockets would be closer to the same size at each stage?

Seems like as long as you have the same amount of chain on the same qty of teeth then the efficiency loss would be close to the same. Not trying to argue a point I'm just curious about some of these generalizations I have heard thrown around.

Its this simple. The gears oil and bearings and other moving parts all have friction and drag causing losses. A properly designed motor that is slightly larger in size does not have any increase in drag.

Incorrect, any given motor design that is slightly larger in size is also slightly higher in drag.

But to answer the actual question Dan, a single reduction can have as low as 2% loss but only if its below 3 or 4:1 and its to do with tooth count and working radius.
The grinding Luke speaks of with gears would be very much the case with a 6 tooth onto a 100 tooth and could result in a 10% loss and lots of noise, however a 22 tooth to a 60 and done twice (surprisingly, less loss) is far less grind and more like a push. also read up on chordal effect -it gets bad below around 13 teeth.
-This all reminds me of debating with Luke maby 5 or 7 years back when he wanted to use 100-80 hobby king motors at 15-20 000rpm and use something like 6 tooth sprockets, ceramic bearings and wrap the can in carbon fiber to stop it exploding, he refused to accept the existence of eddy current losses that i had actually measured, that might limit the useful rpm. Those days when you didn't listen were special, don't you think Luke?

Anyway Dan in f1 naturally aspirated 20 000rpm days, the transmission sometimes had over 4 stages of reduction to keep each of them down to 2.5:1 each and over 98% efficiency (per stage) with spray bar lubrication etc.

 

[Arlo1]

Its this simple. The gears oil and bearings and other moving parts all have friction and drag causing losses. A properly designed motor that is slightly larger in size does not have any increase in drag.

The larger motor does not need to have any increase in drag. If designed for the system.
Remember the diameter might go up but the RPM will go down.

 

[alan]

In the tesla example, if you were to make the motor roughly 3 times the diameter, you would have roughly the same low end torque in a direct drive setup. It should also be able to rev and pull way past the current cars top speed with the same technology. No?

Larger motors will have lower max RPMs before they self destruct from centripetal forces. The gyroscopic precessional forces can become an issue with large fast rotors as well. In a car such as a Tesla, the wheel speed is easily within RPM ranges of any reasonable motor, but the reduction gear is needed to multiply torque, which pushes the motor near max RPM at max car speed. The motor sizing plus reduction gear is a much lighter weight solution to maximize torque and also minimize motor heating within the speed envelope, than would be a larger motor.

To address a previous point, a well designed two speed tranny can be as efficient as a single reduction gear, built with four gears sharing two shafts, with two gears engaged at a time, and shifting done with a simple sliding of one shaft's two gears. The extra gears and syncros will only add some rotational momentum and weight, but no significant additional frictional losses (except during the shifting). If efficiency across a wide operating range were critically important, then the tranny solution would be worthwhile, if only for a slight increase in efficiency. Max efficiency in a narrower operating range will nearly always favor a direct drive solution.

A lot of the tranny argument above was comparing apples and oranges where the efficiency/power/weight/etc. parameters of more than just the motor and tranny were being implied (such as battery and controller issues), but not clearly stated when absolute statements were being made. It is a lot like arguing which car has better aerodynamics where one side is only considering the drag coefficient and the other side is only considering the frontal area times the drag coefficient. My arguments favoring a tranny was strictly in reference to motor power in/wheel power out efficiency when high torques are needed across wide RPM ranges. That is small variations in motor efficiency across its operating range can be positively compensated by an efficient tranny. However, I agree with the practical arguments for not using trannys when other factors are included, which include most real-world applications.

 

[speedmd]

It has been a long road for me to come to the simple conclusion that eliminating gears (with all their associated evils), gearboxes, gear oils, bearings, intermediate shafts and housings is a winning argument from a engineering stand point. "We have seen the fork" as Jerry Seinfeld said it.

Relatively certain that in time we will have solved all the other issues and it will be simplified to the most basic components.

 

[Arlo1]

[arguments for not using trannys when other factors are included, which include most real-world applications.

All of the very in depth data has been presented in this HUGE going no where thread.

I ride a Zero it runs a belt strait from he motor to the wheel. It works better then a Brammo with a tranny and accelerates faster and has less things to break and is more efficient so is that an Apples to Apples enough comparison for you?

 

[Punx0r]

It was on field weakening. https://endless-sphere.com/forums/viewtopic.php?f=30&t=63467&start=50

Wow! Good memory!

 

[Arlo1]

To address a previous point, a well designed two speed tranny can be as efficient as a single reduction gear, built with four gears sharing two shafts, with two gears engaged at a time, and shifting done with a simple sliding of one shaft's two gears. The extra gears and syncros will only add some rotational momentum and weight, but no significant additional frictional losses (except during the shifting).

Bullshit!

When ever you have additional gears to select they have friction points no matter how you try to build it. The gears not in use will be rolling on bearings at different speeds then the shaft causing drag/inefficiencies and they will have some sort of device like a shift fork to either move the dog type gear or the synchro to engage/disengage them this will also not be spinning with the gear causing more drag/inefficiencies

 

[Hillhater]

3x diameter does offer 8x torque. Think about it like unrolling the motor, and you see you have both a longer strip of stator/rotor interaction area, as well as applying that force onto a longer lever arm.

However, in the case of the Tesla, it seems like it would be a better option to eliminate having the torque get combined in a differential and instead have a 2x larger diameter independent motor per wheel.
..."..........
A direct drive motor doesn't need to weigh even a tiny bit more than a tiny high RPM motor, it merely needs to use the electrical materials at the same rate as the tiny high RPM motor. This can mean the motor requires a very high pole count rotor if the needed RPM range is very low, which can lead to issues in manufacturing, and the structure to securely support the tiny thin stator teeth can get to be a substantial engineering project. On paper it should always be capable of being lighter to be single stage, in practice it's almost never done due to mfg and engineering related challenges to use the iron at the same rate some tiny motors at high RPMs do.

No problem, I understand the square relationship between torque and rotor diameter, but again its a theory for a perfect situation and not so easy to replicate in practice ?
But i have a hard time to think you can double/triple the diameter of any given motor design and retain the same weight as original !
Again, possible in theory ?...but in practice it will never happen if the same mechanical design materials and standards are retained.
Back to our simple Ebike drives where the difference in weight between a 1kW hub and a similar power mid drive is obvious.

 

[Arlo1]

Hillhater I think it is supposed to mean the amount of power the motor can produce per given amount of weight will stay the same so when you are making the motor bigger it is getting heavier but it is able to run more continuous power. This means all weight given up by removal of a transmission will then be used to make a more powerful motor making the system the same total weight and more efficient but because the motor can run at higher levels continuously and it removes the losses of the transmission.

When making a motor bigger think hollow center.

I have personally seen this done and it is proven data can't tell you more because its things I am not supposed to see....

But I am sure others can verify this as well.

 

[liveforphysics]

-This all reminds me of debating with Luke maby 5 or 7 years back when he wanted to use 100-80 hobby king motors at 15-20 000rpm and use something like 6 tooth sprockets, ceramic bearings and wrap the can in carbon fiber to stop it exploding, he refused to accept the existence of eddy current losses that i had actually measured, that might limit the useful rpm. Those days when you didn't listen were special, don't you think Luke?

I was so confused, and you and Miles and Bigmoose and many other were offering excellent advice I was foolishly ignoring.

I've had a dyno and track educate me the hard way since then, and now I appreciate core losses and how to make good use of iron and copper.
I also was once heavily bias to axial flux motors, but now I respect inrunner, outrunner, axial all as equals just each with different packaging options.

 

[alan]

When ever you have additional gears to select they have friction points no matter how you try to build it. The gears not in use will be rolling on bearings at different speeds then the shaft causing drag/inefficiencies and they will have some sort of device like a shift fork to either move the dog type gear or the synchro to engage/disengage them this will also not be spinning with the gear causing more drag/inefficiencies

A simple two-speed tranny can have an input shaft and an output shaft (exactly like a reduction gearset with the same number of bearings, having the same loading based almost entirely on the torque loads and not the mass of the gears) with two gears on each shaft. Say the input shaft has a large gear A and a smaller gear B, and the output shaft has a smaller gear C and a larger gear D. A and B can be spaced apart and C and D can be close together. C and D move together on the output shaft so that either A and C are engaged or B and D are engaged. The synchros and sliding mechanism (throw-out) can be designed such that each part is only in contact with any rotating part during the shift, or spin exactly with one shaft, when the gears are engaged, such that the ONLY additional mechanism in play when the gears are set to one speed or the other is a gear hold lock, which can also spin with the one shaft as a locking fork engaged through the center of the shaft, always rotating with the shaft, able to lock into either speed gear position. This is but one design I know of. There is ZERO additional rolling or sliding friction when the gears are engaged over what would be seen in a two fixed-gear reduction unit. There is additional mechanical complexity and weight, but NO additional frictions except during shifting. Both the shifting and locking mechanism can be done with non-contacting electro-magnets. Also, not applicable in bikes or cars, etc., but if the torque load is always in one direction, then opposite gear bias alone can be used to hold the gears in place with no locking mechanism needed.

 

[Arlo1]

When ever you have additional gears to select they have friction points no matter how you try to build it. The gears not in use will be rolling on bearings at different speeds then the shaft causing drag/inefficiencies and they will have some sort of device like a shift fork to either move the dog type gear or the synchro to engage/disengage them this will also not be spinning with the gear causing more drag/inefficiencies

A simple two-speed tranny can have an input shaft and an output shaft (exactly like a reduction gearset with the same number of bearings, having the same loading based almost entirely on the torque loads and not the mass of the gears) with two gears on each shaft. Say the input shaft has a large gear A and a smaller gear B, and the output shaft has a smaller gear C and a larger gear D. A and B can be spaced apart and C and D can be close together. C and D move together on the output shaft so that either A and C are engaged or B and D are engaged. The synchros and sliding mechanism (throw-out) can be designed such that each part is only in contact with any rotating part during the shift, or spin exactly with one shaft, when the gears are engaged, such that the ONLY additional mechanism in play when the gears are set to one speed or the other is a gear hold lock, which can also spin with the one shaft as a locking fork engaged through the center of the shaft, always rotating with the shaft, able to lock into either speed gear position. This is but one design I know of. There is ZERO additional rolling or sliding friction when the gears are engaged over what would be seen in a two fixed-gear reduction unit. There is additional mechanical complexity and weight, but NO additional frictions except during shifting. Both the shifting and locking mechanism can be done with non-contacting electro-magnets. Also, not applicable in bikes or cars, etc., but if the torque load is always in one direction, then opposite gear bias alone can be used to hold the gears in place with no locking mechanism needed.

Bullshit. The gears not in use will add drag/inefficiency no matter how you make it the gear not in use will not be turning the same speed as the shaft and that will cause drag. Also those extra gears will cause more oil drag or what ever lube you run as they will fling it around more then not having them.

 

[DanGT86]

Edit because Arlo beat me to it.

The point of the no gearbox argument is that if you add up the loss from that frictional drag and rotating mass as well as the weight penalty of the gearbox it would not be as efficient as using that same mass in the motor.

 

[nutnspecial]

There is no argument afaik here that more energy isn't used in power transmission as system complexity increases. From a losses of a naked permanent mag motor (ie brushless hub), to cv power transfer, to a reduction, to a 2spd, etc.
I'm thinking the argument is: current available ev design can completely overcome the necessity for current available variable gearing? But if it's that simple, why do I not remember it said so simply? Arlo, Luke, is there a simple claim?

There is ZERO additional rolling or sliding friction when the gears are engaged over what would be seen in a two fixed-gear reduction unit.

I found Alan's above full explanation helpful and trust it's accurate along with noted 'disclaimer'. The below was actually enough for me,

To address a previous point, a well designed two speed tranny can be as efficient as a single reduction gear, built with four gears sharing two shafts, with two gears engaged at a time, and shifting done with a simple sliding of one shaft's two gears. The extra gears and syncros will only add some rotational momentum and weight, but no significant additional frictional losses (except during the shifting). If efficiency across a wide operating range were critically important, then the tranny solution would be worthwhile, if only for a slight increase in efficiency. Max efficiency in a narrower operating range will nearly always favor a direct drive solution.

Followed by thread specifics I further agree with.

A lot of the tranny argument above was comparing apples and oranges where the efficiency/power/weight/etc. parameters of more than just the motor and tranny were being implied (such as battery and controller issues), but not clearly stated when absolute statements were being made. It is a lot like arguing which car has better aerodynamics where one side is only considering the drag coefficient and the other side is only considering the frontal area times the drag coefficient. My arguments favoring a tranny was strictly in reference to motor power in/wheel power out efficiency when high torques are needed across wide RPM ranges. That is small variations in motor efficiency across its operating range can be positively compensated by an efficient tranny. However, I agree with the practical arguments for not using trannys when other factors are included, which include most real-world applications.

Is this an acceptible statement everyone could agree on here?
In vehicles such as cars, trucks, and bikes, variable gearing will find decreased benefit as operational/conditional range/intensity of demands decrease, and/or power level and/or power efficiency increases

Considering the broadest amount of info and understanding I can, I then would question whether variable gearing could ever be completely ruled out as circumstantially beneficial in such types of transportation before achieving unity, and continue backward from there to looking for evidence for or against.

I wonder if model S could be equipped with a 2nd spd of 1.5:1 or 2:1, offering better competion in the race (previous linked) with mclaren, or it's 1spd original. I suspect it would be possible to increase performance 0-150+mph with little drawback in weight, cargo capacity, range, and launch.

Further, I wonder if the same could be said when it's 2spd VS a 1spd 'ALL E' model S designed for competitive status considering the same 5 criteria. Would also be good to keep track of comparative costs, vehicle handling, and perhaps environmental friendliness???