Bizarre, yet reliable and robust CVT transmission design.

Joined
Oct 28, 2008
Messages
14,539
Location
Manhattan Beach, CA, USA
transmission2.jpg



This transmission design initially seemed like one of the most silly transmission designs I'd ever seen, and you wouldn't think it could become efficient or reliable given the nature of it's method of operation. Yet, leave it to Honda to make it as reliable as a normal transmission, and efficient enough to enable the improved control over the engine operation RPM to more than make up the balance in efficiency lost over a standard transmission design.

Not really my cup-of-tea, or something I would personally want to use, but a clever and throughly real-world-proven design for a CVT that may get some brain-juices flowing in folks for innovative mechanical power transmission solutions for EVs.

Here is a little write-up info about it's method of operation:

http://dirtbike.off-road.com/dirtbike/feature/shiftless-wonder-hondamatictm-automatic-transmission-12134.html
 
Funnily enough, the same transmission principle (variable volume swash plate hydraulic drive) was used in a military application many years ago. It was very reliable, but not that efficient, as the hydraulic losses were quite high. Honda seem to have done that thing the Japanese do exceptionally well - optimised the system design - to make this work as well as a conventional transmission.

I've often wondered about using the same principle as Toyota use on their Hybrid Synergy Drive to make a pure electric 'CVT'. The HSD is just a simple, in line, differential, just the same as has been fitted to the axle of cars for a hundred years or so, but without the right angle drive. The mechanical gear ratios are fixed and cannot be changed. The clever bit is the use of two motor generators, as the output shaft speed, torque and direction is controlled by varying the speed, torque and direction on one of the motor generators, whilst allowing the Atkinson cycle ICE to run over its very limited efficient rpm range. In effect, the differential becomes a CVT because torque can be varied on one of the two input shafts via the electronic control system, shunting power bidirectionally between the two motor generators via the electronic control system.

If you substituted a third electric motor for the ICE, then you could build an entirely electronic variable drive. The efficiency would be as good as you could make the control electronics, probably a bit better than hydraulics or a belt type CVT system, but lossier than direct gear/chain/belt drive.

The Toyota HSD is another example of a Japanese company taking ideas that have been around a long time and using them in a new system, with some very innovative optimisation, to create a very clever product.

Jeremy
 
Dude I am a bike mechanic for a great living and its not that efficent or that relible. The unit it self is I guess relible but.... It needs a lot of oil in the motor to Work and it uses a angle sensor that always fails amongst other things. I have replace 3 of the "Hondamatic units" In the Honda Rubicon ATVs and they are ~1300 for the one part pluss gaskets pluss piston rings ect well you are in there and it takes a long mother of a time to change it in a ATV!
The other thing is the 500 cc Rubicon it comes in feels way down on hp compared to a 500 semi automatic (Centrifugal clutch and standard tranny)!
Its a cool techknoladgy but not worth it and its a flop as far as I have seen all the owners hate them cause of the problems!
 
Jeremy was right about this configuration having been around a long time. I "think" this piston arrangement with swash-plate was used in steam-powered torpedoes in WW-one (1914-18). Not boiling-water steam, some kind of chemical reaction (IIRC). As of the 1980's it was still being used in torpedoes using internal combustion (combination fuel and liquid oxidizer)

This type of arrangement has also been used in hydraulics for a very long time. With the swash-plates angle set to zero, the entire arrangement spins, but no oil is being pumped, so very little heat being generated. A less expensive hydraulic pump has a fixed angle and constant flow. A control valve diverts oil to the work-load, or back to the oil tank (or half-and-half). They may be simpler and cheaper, but they run hot, and cannot be run constantly.

I have seen hydraulic wheel-drives (with an engine and pump in the center of the vehicle) in a slow 6WD off-road truck. It had high ground clearance (hydraulic motor inside each wheel). Also allowed each wheel to be individually raised or lowered...but slow. On the highway it was transported on a flatbed trailer.
 
Along similar lines what happened with the dual electric motor e-bike drive system from a couple of years ago that was supposed to be an open source deal? It was 2 motors with very different capabilities to work together resulting in a CVT/motor drive in one that was supposed to gain significant efficiency.
 
If you substituted a third electric motor for the ICE, then you could build an entirely electronic variable drive.
You better take a planetary gear and adapt a motor to the planet assembly and another motor to the ring gear or the sun gear. Either way, by controlling both motors in direction and torque the third part of the planetary would spin at any rpm you like.

-Olaf
 
liveforphysics said:
...
Not really my cup-of-tea, or something I would personally want to use, but a clever and throughly real-world-proven design for a CVT that may get some brain-juices flowing in folks for innovative mechanical power transmission solutions for EVs.

Here is a little write-up info about it's method of operation:

http://dirtbike.off-road.com/dirtbike/feature/shiftless-wonder-hondamatictm-automatic-transmission-12134.html

Necroing a very old post I know, but I'm looking at transmissions atm, trying to find a simple, reliable 2 or 3 speed with neutral and the ability to engage any gear at standstill.
Sadly/expectedly the link no longer works, but will search more.

But!
Speaking of brilliant gearboxes, THIS is probably the most brilliant gearbox I have ever seen liveforphysics!
You know the old 2 cones CVT: Pity you cant put gear teeth on it right, or... can you...!?? :)
https://www.youtube.com/watch?v=neWhdcC30cc

The site:
http://www.new-cvt.com/animations.php
where you will find more video of a prototype working at speed.

It's more suited to an ICE designed for performance in a very narrow rev range and I could see something like a dragster oblitering records with one, if the design can handle the power. I believe it could..?

It could also replace a std bicycle de-railer, but where to put it..?
I think the gear pattern on the cone could be put on/translated to a disk, with the driving gear on a shaft drive.
Hope you like it too!? :)
 
The thing is, I've now done enough days of dyno testing EVs to realize, any/all mass given towards a multi-speed transmission is instead re-applied towards making the motor stator iron and copper and magnets larger, the end result is higher continous power delivery and efficiency, simplicity, longevity.

Remember, only the iron-copper-magnet mass is performing the transformation from electrical power from the battery into useful shaft work. Everything downstream of that energy transformation device is causing you parasitic losses on the eay to turning your wheel, no matter how clever the design.

For the reason, ones focus is best spent finding or making a motor with the desired magnetics to accomplish your desires in the torque required over a given RPM range.
 
My friend has a Nissan Juke with a CVT, he can tow exactly nothing.

Light is good
Smooth shifting with lots and lots of speeds for good fuel economy.
but cant tow nothing

Most people dont care when they are out buying a car. They just like the looks, and price.
Until surprise surprise, you did not have the kodiak boat before but now you do and time to look for a new vehicle.
Fun fun times

Be cool to have all the benefits of a Bizarre, yet reliable and robust CVT transmission design, yet be able to tow. Most normal 6cyl's vehicles can tow 5000lbs.
 
calab said:
My friend has a Nissan Juke with a CVT, he can tow exactly nothing.

Light is good
Smooth shifting with lots and lots of speeds for good fuel economy.
but cant tow nothing

Most people dont care when they are out buying a car. They just like the looks, and price.
Until surprise surprise, you did not have the kodiak boat before but now you do and time to look for a new vehicle.
Fun fun times

Be cool to have all the benefits of a Bizarre, yet reliable and robust CVT transmission design, yet be able to tow. Most normal 6cyl's vehicles can tow 5000lbs.

Most CVTs rely on friction caleb.
Both the variable diameter V pully with V shaped belt and the 2 Cone thing.
That means that; too much torque and they slip like crazy, or you have to pull the belt so tight that efficiency goes to hell.

NB that what makes this Gearbox absolutely genius is the fact that this guy has figured out how to put gear teeth on a cone and get seamless, instantaneous changes with zero slip and the efficiency of a manual gearbox!!! :shock: :D

Do check out the video, just for ...'self enlightenment' firstly.
It really is "Holey Da Vinci, Newton and Einstein!!' genius!
Then consider the advantages of instantaneous gearchanges under power vs a std derailer.
The 58 gears of the prototype too...
:)
 
The thing is, I've now done enough days of dyno testing EVs to realize, any/all mass given towards a multi-speed transmission is instead re-applied towards making the motor stator iron and copper and magnets larger, the end result is higher continuous power delivery and efficiency, simplicity, longevity.

Remember, only the iron-copper-magnet mass is performing the transformation from electrical power from the battery into useful shaft work. Everything downstream of that energy transformation device is causing you parasitic losses on the way to turning your wheel, no matter how clever the design.

For the reason, ones focus is best spent finding or making a motor with the desired magnetics to accomplish your desires in the torque required over a given RPM range.

You make good points liveforphysics. :)
BUT
seeing this guy put teeth that work on what looks like the cone CVTs we know is more about those rare self enlightening;
"WOW! Holy Da Vinci, Newton and Einstein!"
moments..!

ie: It's not a CVT; it's a 58 speed (in the prototype. Could be more or less) toothed cone, with the gear changes of a CVT (kinda) and the efficiency of a manual gearbox. (no slip)
It's well worth watching simply to expand one's own... 'self'. Do take a look! :)
(My reply to calab above expounds on things some too)
http://new-cvt.com/

I quite agree that larger hub motors, with field advance etc are the way to go... until you start talking about:
* Potholes: There are millions of pot holes large enough to bath in here (New South Africa) so unsprung weight isn't just a handling/grip disadvantage; destroyed tires and rims is de-rigueur around here.
* Larger, heavier vehicles. Especially rear wheel drive ones. (simple power transmission) and cost/availability.

I'm researching with an eye to an enclosed velomobile type vehicle.
Mostly for personal security. (Higher murder rate than the Vietnam war here atm)
But also for safety, weather proofing and aerodynamics.

So more of a small, aerodynamic, economical tandem 'car' type conveyance, capable of around 120km/h, using off the shelf motors, Fat Tires/rims, etc.
Something Raht Racer-ish, but better looking, sturdier, leaning and tandem.
That would likely be easier and cheaper to accomplish with a 2 or 3 speed gearbox/es , capable of regen and coasting, don't you think?


(I'm also thinking; possibly a parallel ICE as; thx to Load Shedding in the... 'New South Africa', electricity is off as much as it's on and getting worse. So besides an economical conveyance, this half baked vision of mine might provide some power when/for home. Solves the gearbox issue too, for one of the wheels at least)
 
Logic11 said:
I'm researching with an eye to an enclosed velomobile type vehicle.
Mostly for personal security. (Higher murder rate than the Vietnam war here atm)
But also for safety, weather proofing and aerodynamics.

So more of a small, aerodynamic, economical tandem 'car' type conveyance, capable of around 120km/h, using off the shelf motors, Fat Tires/rims, etc.
Something Raht Racer-ish, but better looking, sturdier, leaning and tandem.
That would likely be easier and cheaper to accomplish with a 2 or 3 speed gearbox/es , capable of regen and coasting, don't you think?

I'm working on a similar project. Link in my signature.

Why limit yourself to two or three speeds? A full suspension trike with a Leafbike motor and all three wheels built into light-duty motorcycle wheels with Mitas MC2 tires would suit you right. You could have 7-speeds in the rear and 3-speeds up front, giving a wide gearing range. More speeds with a Schlumpf drive.

AND unlike the Raht racer, it's quite possible you could come up with a design light/efficient enough to pedal to faster-than-bicycle speeds with the motor turned off, while still being sturdy enough to handle bad roads at speed with the motor in use.

A Steintrikes Wild One would likely be sturdy enough to handle bad roads, or maybe a KMX with aftermarket suspension depending upon how sturdy the current generation spindle designs are. I ride on bad roads, but I'm uncertain if they're as bad as South Africa's roads. My 3rd generation KMX spindle lasted more than 60,000 miles before failure, but my cruising speeds were closer to 20-25 mph before motorizing the trike and 30-35 mph after motorizing the trike, with rare but often daily occasions of exceeding 50 mph down hill. The spindle is mechanically the weak link in a KMX trike. I have no experience with the Steintrikes Wild One, but looking at the design and given what I know about engineering, it appears to be among the sturdiest off-the-shelf trikes you could buy and would make an interesting base for the vehicle you describe, although given its suspension travel and component clearances, designing an efficient body around it would be a challenge.
 
I never looked into it prior, may look into it now but someone does need to come up with a cvt that you can tow with, even just a couple dirt bikes or a kodiak blow up boat with hard bottom. Someone said cvt's are like bicycle gears then my buddy gets f'd over buying the Nissan Juke but its his own fault for 1) buying a cvt vehicle and 2) doing exactly zero research
Like who does that? Spend $10k+ and do zero research
:lol: :lol: :lol: :lol: :lol: :lol: :lol:
I think he paid $10-15k for his ride.

Friction is a clutch in a manual transmission too.
Skip cvt's
unless
Bizarre, yet reliable and robust CVT transmission design.
That is what the market really needs, reliable and robust cvt's.

I will look into it and that video, if I remember.

Have to remember that the makers engineer in a failure rate for more profits. Can never go wrong with old tech, cheap parts but paying extra in fuel. Once you get into the ancient stuff, no parts are available and you have to make it yourself with 3d printers and expensive scanners. Its still pretty awesome where tech has brought us today.

Tremec has 7 and 8 speed manual transmissions, not automatic, but you can always add a box unit (over-drive or under-drive) for 14 and 16 speeds ;) :thumb:
https://www.tremec.com/menu/passenger-vehicle-transmissions/
North Americans dont do manual transmissions, Europe sure does because of gas pump prices.

box units are not a bad price, especially with fuel prices the way they are.
https://www.summitracing.com/search/part-type/overdrive-units
 
Even the best CVT just takes motor shaft power and wastes some of it as heat.
It's easy enough to size the motor to give you the desired torque across the desired RPM range. Everything you need to add in series after your motor just means you need to re-choose a motor better suited for the application.
 
Don't skimp on the motor, thats good advice no matter what the application is, like ebikes or ev's, skateboards, stand up scooters. Just means your wasting money if your motor is not powerful enough, if your lucky when swapping that your batteries can stay.
 
liveforphysics said:
Even the best CVT just takes motor shaft power and wastes some of it as heat.
It's easy enough to size the motor to give you the desired torque across the desired RPM range. Everything you need to add in series after your motor just means you need to re-choose a motor better suited for the application.

:eek: ok; DON'T watch the vids! :lol:

Aren't you curious how a gear can be made to mesh with with a bunch of larger diameter gears with no let up in transmitted power..!? I'm sure I saw somewhere that you where a 'gearhead' before going electric?

Yep; I agree gearboxes add inefficiency, yet they are in cars, rather than a 20 liter single cylinder 'thumper' with a huge stroke.
Why? Is it simply due to the fact that ICEs have no torque at 0 rpm? (Elec motors have sucky efficiency at 0 rpm)
Or does the fact that the engine and gearbox, being smaller and lighter than a huge power source = a more efficient, faster accelerating and top end vehicle overall..? :)

Why does one get them epicyclic geared hub motors?
I see there's a new patent out with 2 sets of sun (ratcheted) and planetary gears (different sizes) and you change motor direction to select the gear you want.
 
Seems like we're cooking along at a gearboxes thread per month at this rate. Every point made here has been made a dozen times previously in all those other threads.

I think Luke is right in almost all EV cases but there are some cases were a gear shifting system does make sense, cases where you have limited parameters and extreme operational ranges. It just so happens that Ebikes sometimes fall into that category which is why I think it always seems like such a good idea to us. Many Ebikes are very power limited and per weight and drag derailleurs are extremely efficient but both of those become moot when you can pick any motor and controller you want for a vehicle and those transmissions start becoming very heavy and draggy to deal with the high torque of a high power electric motor.

Although that "CVT" isn't a CVT, it's just a normal gearbox with an odd shifting system that may or may not run into all sorts of issues when you actually try to shift it under power.
 
I am a gearhead, I've broken so many transmissions and clutches and axles running drag car events and road racing.

As a true gearhead, I flipping hate transmissions with a firey passion. They are my number one cause of DNFs, and the number 1 money hole your drag cars keep throwing $5-10k at a time at to try to get something that holds together, and nothing at any cost does.

Ironically among drag racers it's well understood the fastest dragsters are always single speeds, the next slowest are 2spd power-glide based, and the very much slower use more multiple ratios to be wasting shifting time and horsepower into friction to try to compensate for the inadequacy of the engines powerband.

My gearhead friends who couldn't give a frock about air pollution, do still dream of EV power because they also waste there paychecks on trying to keep a gearset with meshing teeth on it together. Just no clutch and transmission to regularly replace is more than enough to cover the battery cost penalty of being an EV.

Remember, any idiot can get something to function by adding complexity, but today seeing that approach is just a disgusting shameful mess to me, including a CVT. When you spend some time dynoing electric motors, you will see how any mass given to CVT only means less performance, less range, more failure modes under all operating conditions vs just starting with the right sized motor.
 
nicobie said:
I have been involved in some type of racing most all my life and I think Luke is right about power transmission. The less gears and whatnot the better.
My brother was the first to use an automatic transmission (1968 Mustang) on the local 1/4 mile drag strip. Some thought he'd lost a wing nut, but he soon proved that shifting gears going down the drag strip was wasting energy. Had more than one trophy to prove it.

Those old enuf to remember the late 60s local drag racing remember that it was not uncommon for someone to miss a gear shift as well as forget to roll up the window :lol: while speeding down the 1/4 mile.

Anyone here remember seeing John Deere demonstrate its first automatic transmission (1960s) at a State Fair on Machinery Hill. As a young teen i remember the demonstration when the driver shifted that JD 4020 into reverse while going forward. Everybody just looked dumbfounded not knowing what to say. It was impressive, but what's the purpose other than to show-off the rugged transmission. The farmers didn't know what to make of it going forward than quickly slowing down and going backwards. That diesel engine purred like a beast.
 
liveforphysics said:
I am a gearhead, I've broken so many transmissions and clutches and axles running drag car events and road racing.

As a true gearhead, I flipping hate transmissions with a firey passion. They are my number one cause of DNFs, and the number 1 money hole your drag cars keep throwing $5-10k at a time at to try to get something that holds together, and nothing at any cost does.

Ah..! :) yes I get you now. I have no experience in drag racing, but can imagine the mess an under spec'd transmission turns into when it 'goes' on a dragster.

liveforphysics said:
Ironically among drag racers it's well understood the fastest dragsters are always single speeds, the next slowest are 2spd power-glide based, and the very much slower use more multiple ratios to be wasting shifting time and horsepower into friction to try to compensate for the inadequacy of the engines powerband.

?? Even top fuelers have to start from a standstill while the engine is at some ungodly RPM, so how do they do that?
I remember reading that engaging 2nd gear, while almost passed out from G-forces, was a challenge, so always assumed some sort of 2 speed transmission was used?

"...wasting shifting time...":

That's just it!
This transmission wastes zero time shifting from one ratio to another. Shifting is completely seamless, like a CVT, but with more ratios than you can wave a stick at, no slip, and straight cut manual box efficiency! :)

liveforphysics said:
Remember, any idiot can get something to function by adding complexity, but today seeing that approach is just a disgusting shameful mess to me, including a CVT. When you spend some time dynoing electric motors, you will see how any mass given to CVT only means less performance, less range, more failure modes under all operating conditions vs just starting with the right sized motor.

I get the extra complexity, weight and friction losses.
Problem I see, from my limited research so far, is that lighter motors just don't have the RPMs to reach the speeds I'd like.
Enough torque sure, (especially if multiple hubs are used, but not the RPMs..???
Then there's the myriad of damn potholes here, that won't play well with the unsprung weight of a big hub motor..?
I note that you do use chains on your 'Death Bike?' rather than a hub. Most efficient and simple, but still a 'gearbox'. :)

Thx for your patience... thus far... :) Enjoying your input...
 
calab said:
I will look into it and that video, if I remember.
Do! :) It's well worth a look!
I think this biologist may have had a.. 'grain fungus' experiment going when he came up with this! :lol:

calab said:
...Have to remember that the makers engineer in a failure rate for more profits...

This is just a 3D printed prototype atm.
I don't think many people ...'get it', to the point where is basically non existent on everyone's radar.
And its badly named for search engines IMHO.
 
Logic11 said:
That's just it!
This transmission wastes zero time shifting from one ratio to another. Shifting is completely seamless, like a CVT, but with more ratios than you can wave a stick at, no slip, and straight cut manual box efficiency! :)

Ah but that is the problem with this gearbox design, in order to shift you must move the gear from one location to another in less than half a rotation, so how exactly are you going to pull that off and not lunch the gearbox instantaneously at 7K RPMs. It's a cool concept with probably some niche use cases but unless you still use a clutch system of some sort it doesn't seem applicable to real world vehicle use. Even at much lower speeds after reducing the motor RPM with another reduction stage it's still going to be quite a feat.
 
scianiac said:
I think Luke is right in almost all EV cases but there are some cases were a gear shifting system does make sense, cases where you have limited parameters and extreme operational ranges. It just so happens that Ebikes sometimes fall into that category which is why I think it always seems like such a good idea to us. Many Ebikes are very power limited and per weight and drag derailleurs are extremely efficient but both of those become moot when you can pick any motor and controller you want for a vehicle and those transmissions start becoming very heavy and draggy to deal with the high torque of a high power electric motor.

Quite!
The potholes here wont play nice with heavy hubs (unsprung weight) that would give me my desired top speed... I think..? (Limited research so far)
I understand one can wind the stators to suite a rpm range and use field advance etc, but still; 120km/h looks to be too much?
Even for 3x large diameter. fat bike powered wheels/hubs.

scianiac said:
Although that "CVT" isn't a CVT, it's just a normal gearbox with an odd shifting system that may or may not run into all sorts of issues when you actually try to shift it under power.

'Odd shifting system' is one way of looking at it. :)
No one has EVER come up with a way to seamlessly, instantaneously shift ratios with no clutches etc before...
That makes this transmission genius IMHO!

How well it would work in the real world is anyone's guess atm. But I think, if properly designed and made, it will work like a bomb!
And the bicycling world is a good place to test things like this as speeds and pedaling torques are way lower than in the car etc world.
I do hope huys like Justin see and grock this.
 
Back
Top