Variator Performance

21 posts in this topic

Posted (edited)

Disclaimer: I am in no way an expert and do not claim to know everything about this topic. Please take everything in this thread with a pinch of salt and a shot of scotch (I don't bother with the salt). This is my opinion and experience with variator modification and tuning. Please feel free to correct any misinformation at any time, we are all learning together. The discussion at the moment does not discuss modification to the rear pulley and mainly refers to the front pulley/variator.

 

To get the most from the discussion, I think the best thing to do is separate the term performance into 2 categories, 1. Modification and 2. Tuning

For this topic, I will consider Modification to be making geometric alteration to the function of the variator such as cutting, grinding, shimming, choice between rollers/sliders, belt length etc.

I consider the tuning of the variator to be any factor that changes the rate at which the transmission variates (that might not be a proper word) such as changing roller/slider weight and choice of contra spring.

Basically, the variator controls the ratio of the transmission.

As all may know, at idle the belt sits down into the front pulley(variator) and rides high on the rear pulley, we can call this the low range. This configuration/ ratio allows for the torque of the engine to be multiplied allowing us to take off from a stop or climb steep hills.

As engine rpm increases, the rollers/sliders begin to accelerate to the outer diameter of the variator ramps. Lets skip this part for a second and only think about the end result of the high range, where the belt rides high on the front pulley (variator) and sits down into the rear pulley. this gives a ratio that will allow for higher speeds.

What we want to achieve: Personally I think we ask too much and are unrealistic when it comes to variator performance but here we go. We want a really low range for the best acceleration, a setup that allows the variator to variate at optimum engine RPM and a high range that gives us the maximum top speed capable. Aftermarket variators such as Nibbi/ NCY/ Dr. Pulley/ Malossi/ Polini/ Kidnme/ K&S/ all the others have spent time designing packages that are much better than stock at low and high ranges. They do this by varying the geometry of the variator. 

Misconception 1: Lighter roller/slider gives more acceleration and heavy rollers/sliders give more speed. This is not really true as the weights only affect the rate at which the system variates. Light rollers/ sliders require greater radial force (centrifugal force) to push the variator open. The kinetic energy is lower than using a heavier roller/ slider due to the reduction in Mass. Since kinetic energy is dependant and proportional to mass and velocity, if we reduce the mass but require the same energy so we can move the rollers/ sliders to push the variator face, we need to increase the velocity. Meaning more RPM is required. 

So simply put, using lighter rollers/ sliders requires higher RPM to variate than heavier ones. Now that thats out of the way lets get to that Contra spring as that affects the rate at which the system variates. Firstly, the contra spring provides a load on the rear pulley to resist the rear pulley from opening.

When the rollers/sliders push against the variator plate, it forces the  front pulley to narrow, this causes the belt to rise in the front pulley while simultaneously increasing the belt tension causing the belt to be pulled into the rear pulley forcing it to open. The Contra spring resists this force, once the force (tension in the belt) is greater than the Contra spring load, it will cause the spring to compress and open allowing the belt to pulled towards the center of the rear pulley giving the higher range of the transmission. 

So simply put, using a stronger contra spring (aftermarket 1500rpm, etc.) forces the transmission to variate at a higher rpm than a weaker spring (stock contra spring) if all other things are left untouched.

Edited by bcyprian25
2 people like this

Share this post


Link to post
Share on other sites

Posted

Nice writeup 25:ok:

Share this post


Link to post
Share on other sites

Posted

Thanks, I have a lot more to add, but didn't want to just start throwing modifications out there without base info. Sorry it took so long, the info I wrote is a combination of research and a lot of just watching and figuring out the way the mechanical components interact with each other, like the description of the contra spring and what causes the rear pulley to open. Just had to make sense of it.

I have to do a bit more thinking before writing to give a good explanation of the way the ratio changes depending on load for climbing a steep hill etc. I know there is a lot of write up out there, but I prefer to understand it in my own way, maybe it might help some others see it better also. I'll try to break it down as best I could.

Share this post


Link to post
Share on other sites

Posted (edited)

Good stuff.

I'm with you on some of your misconception 1 and on some not so much.

Lighter flyweights do make more acceleration. It has to do with a different misconception.

Most if not all of the CVT discussion in these scooter forums is oriented around max performance, WOT, highest top speeds. But the CVT is designed for more than that.

I'll use an example using made up numbers:

Engine rpm at clutch engagement - 3000 rpm

CVT "tuned" engine rpm - 7000 rpm (This is a maximum rpm that the CVT trys to limit by upshifting.)

CVT controlled engine rpm operating range - 4000 rpm (7000-3000)

The operating range is a big part of the CVT that tends to be ignored by these scooter forums. I'll wager there are scooters all over the world that the operators seldom use full throttle or run the engines up to "tuned" rpm. They aren't getting to top speed of course but the machines are upshifting to some extent and the folks are puttin around just fine.

If you impose the ground speed through the drivetrain up to the drive pulley, you can look at it as engine rpm - mph. Lightening the flyweights equates to a higher engine per mph. That's the same effect as the lower gearing, better engine torque that you describe.

So at takeoff your rpm's are closer to the 3000 than the 7000 (not much hp) and the lighter flyweights allow the engine to use that lower gearing effect to increase rpm's - mph a bit.

The speed part I agree with. Different flyweights effect the upshift of the CVT. Once the CVT up shifts to full shiftout then the flyweight's job is essentially done and different flyweights won't matter.

 

Edited by SD12
1 person likes this

Share this post


Link to post
Share on other sites

Posted

Ya good write up, explains more for me, this is my first scooter, I'm an old sprocket and chain kind of guy.thanks

1 person likes this

Share this post


Link to post
Share on other sites

Posted

Good stuff.

I'm with you on some of your misconception 1 and on some not so much.

Lighter flyweights do make more acceleration. It has to do with a different misconception.

Most if not all of the CVT discussion in these scooter forums is oriented around max performance, WOT, highest top speeds. But the CVT is designed for more than that.

I'll use an example using made up numbers:

Engine rpm at clutch engagement - 3000 rpm

CVT "tuned" engine rpm - 7000 rpm (This is a maximum rpm that the CVT trys to limit by upshifting.)

CVT controlled engine rpm operating range - 4000 rpm (7000-3000)

The operating range is a big part of the CVT that tends to be ignored by these scooter forums. I'll wager there are scooters all over the world that the operators seldom use full throttle or run the engines up to "tuned" rpm. They aren't getting to top speed of course but the machines are upshifting to some extent and the folks are puttin around just fine.

If you impose the ground speed through the drivetrain up to the drive pulley, you can look at it as engine rpm - mph. Lightening the flyweights equates to a higher engine per mph. That's the same effect as the lower gearing, better engine torque that you describe.

So at takeoff your rpm's are closer to the 3000 than the 7000 (not much hp) and the lighter flyweights allow the engine to use that lower gearing effect to increase rpm's - mph a bit.

The speed part I agree with. Different flyweights effect the upshift of the CVT. Once the CVT up shifts to full shiftout then the flyweight's job is essentially done and different flyweights won't matter.

 

This is some good write up as well SD, this is exactly what I hoped for, some good discussion. I agree that most scooter owners dont run at WOT. I on the other hand have a bad habit of doing what I'm not supposed to, so I ride a bit harder than I should. The explanation of the lighter weights will give more acceleration is correct as you mentioned, the misconception I dont like is when guys think that because its a lighter weight, they wont get as much top speed as if they used heavier weights. Dont think I explained enough, lol. 

The part you explained about the ground speed and condition is exactly what needs to be spoken about for the shifting of the transmission or "torque sensing". I'm having a hard time continuing the write up right now, we are getting flooded again, so I got to head out of work early today. I'll try to continue the modification section over the weekend. 

Please feel free to add the shifting characteristics and mechanical principal of the CVT dependent on load demand/ road characteristics, I know a few of you can explain it well.

I think more discussion like this is what the scooter community needs

Share this post


Link to post
Share on other sites

Posted

BCY you had me at Scotch hahaha that's how we learn, read and read some more, I've played with my polaris quad, there's differant settings with the springs to change how long it will Rev before the belt climbs, 

1 person likes this

Share this post


Link to post
Share on other sites

Posted

This is some good write up as well SD, this is exactly what I hoped for, some good discussion. I agree that most scooter owners dont run at WOT. I on the other hand have a bad habit of doing what I'm not supposed to, so I ride a bit harder than I should. The explanation of the lighter weights will give more acceleration is correct as you mentioned, the misconception I dont like is when guys think that because its a lighter weight, they wont get as much top speed as if they used heavier weights. Dont think I explained enough, lol. 

The part you explained about the ground speed and condition is exactly what needs to be spoken about for the shifting of the transmission or "torque sensing". I'm having a hard time continuing the write up right now, we are getting flooded again, so I got to head out of work early today. I'll try to continue the modification section over the weekend. 

Please feel free to add the shifting characteristics and mechanical principal of the CVT dependent on load demand/ road characteristics, I know a few of you can explain it well.

I think more discussion like this is what the scooter community needs

I want to honor your request to keep this thread on the variator. I am pretty knowledgeable on the Torque Sensing function of the driven pulley and I can throw in my 2 cents if you intend to make a thread covering that.

I tend to use analogies a lot on this sort of stuff and I like to do some comparisons to snowmobile CVT's and even car auto transmissions. I try not to go too deep so as not to cause confusion but I think it can help. The snowmobile world has quite a bit written up on this and can be very helpful if you learn up on the basic functions and understand the differences. Even the owner/service manuals might include a pretty nice theory of operation.

One of the differences to snowmobile CVT's is that they tend to make a lot of use of the  mid range of throttle and performance.  With the larger pulley diameters (more range), higher hp, faster speeds and the road conditions (scooters on smooth pavement vs. sleds off road / varied terrain), I think even the no-fear squids would be surprised at how much they are actually off the throttle on a snowmobile. Not sure how much of that compares to a scooter with the performance range being much smaller on a scooter between no-throttle and full-throttle.

1 person likes this

Share this post


Link to post
Share on other sites

Posted

Here's a debatable opinion involving the GY6 type variator.  Roller vs sliders.

The rollers fall into what I would call bad engineering design. Using a roller as a wedge between two ramps.

I haven't used sliders. I think the rule of thumb is that a slider generates the cg force of a roller approx. 2 grams heavier. I think that is because the sliders work way more efficiently than rollers. More of the cg force gets transferred to the pulley sheave whereas some of the rollers cg force gets wasted in scuffing and binding.

I believe that while a slider works as a wedge and forces the ramps apart reasonably well, a roller is also used as a wedge in trying to force the ramps apart but doesn't work very well. The roller would have to roll on one ramp and skid backwards on the opposing ramp. At some point the roller eventually just stops rolling and skids on both ramps. That is where the flat spots come from.

The flat spotting isn't consistent either. I have had roller sets where one roller had flat spotted down to the metal and another of that batch had much less flat spotting, maybe several small flat spots. I think that is because one roller locked up sooner than the other.

1 person likes this

Share this post


Link to post
Share on other sites

Posted

Sd12 I think your observations are spot on. Ive posted these before but look at a couple sets of the rollers just skidding along to a flat spot then maybe skid to make another flat spot, as this happens I can feel the tune go away and time to service her.

IMG_1183.jpg

DSCN1439.jpg

And my one experience with Sliders, I'm pretty darn meticulous and that one rolled over or I installed it that way.

I felt this tune go away thats why i pulled it down.

DSCN0634.jpg

Share this post


Link to post
Share on other sites

Posted

On another site I was asking about putting sliders in my smax and they were totally against it, ever Dan was, so I just left them with rollers, I have never found any ware on the interweb what the stock weight is of smax rollers, I was thinking of 1 to 2 gram heavier just to achieve the same speed but maybe a bit lower rpm, so I ride every day a stock bike hahahaha so little information on the yami smax.

Share this post


Link to post
Share on other sites

Posted

I agree fully with you guys, the last set of rollers I had turned into a flat spotted slider with a big loss in performance, I was pissed. It happened in under a month. So I always run sliders. I've even modified the sliders at one point. The face that is closest to the perimeter of the variator, Ive shaved about 20 thou off of it as thats what comes into contact and stops the travel of the slider outward. I only saw an increase of about 3-4 km/hr. I'll continue the thread with some actual modifications a bit later today.

1 person likes this

Share this post


Link to post
Share on other sites

Posted

I watched a Utube video of someone grinding out ramp ends of the variator sheave. The theory seemed to be that if any restriction was removed there, then the flyweights would go farther outwards and upshift farther.

Something in the CVT is going to be the mechanical limit and if happens to be the sheave ramps, then yeah the CVT shift range could be extended till it gets to whatever would be the next limit.

Thing is, centrifugal force is very powerful and when added to these engine type rpm's, its something to be aware of. And messing around at the outer perimeter of the sheave is the worst place.

I did some hack math. Using an example of a 5" diameter sheave and 8000 rpm.

Any thing at the perimeter of the sheave would have a speed of 2094 f/s (feet per second).

For some perspective, a 38 special would have a muzzle speed of  1000 f/s. A 9mm at 1258 f/s and a M1 at 2000 f/s, all approximates.

If the thing did grenade, the cover is there and would be a huge safety net, more than most CVT machines have. Course from the cutch bell back is a direct connection to the drive wheel and anything that might plug that up could lock up the wheel. And the folks that turn the covers into some version of a ankle-biter are removing that safety net.

I'm not much of a safety cop, but that sort of hacking is really upping the risks. Even a unnoticed crack could be catastrophic with those forces.

1 person likes this

Share this post


Link to post
Share on other sites

Posted

Thanks for keeping this easier  to understand. I think in most cases, the rider would like to get to the engines' "usable power band" as quickly as possible,(and stay there) not necessarily WOT. How does one accomplish that? By using sliders? Heavier or lighter rollers? By varying clutch spring or variator spring  tensions? Is it possible to adapt a larger or smaller variator from, say, a snowmobile or Jetski-and would there be any advantage? What effect-if any- does clutch "tuning" have on the variators performance? There's a lot to learn but I don't know where else to find some answers.

Share this post


Link to post
Share on other sites

Posted

If you want max performance then max hp and max throttle is key, no way around it. Lots of good tuning info on the forums for that. You can get very good advise on which parts may work for your situation. You could take one machine and tune it differently for different riders, the more different the riding styles, the more different the tune.

Using a larger pulley diameter system may improve things a bit, probably not a lot. These systems are kind of sized to fit. Snowmobile pulleys are quite a bit larger, both diameter and mass / weight. That larger diameter has a huge benefit. Snowmobiles have a ground speed range from 0 to triple digits using a single speed just like the GY6 system. But they also have the hp to handle those speeds. A 50cc or 150cc wont.

If you were to custom build a CVT system using something more like a scaled down snowmobile design I think you might see some big improvements.

Share this post


Link to post
Share on other sites

Posted

SD you sure do know your CVT systems and appreciate all the feedback from everyone, this is exactly what we need. To add to your previous post about grinding ramps and slider "channels" I also use it to manage the mechanical limits of the CVT. Also note that every almost every setup is different and has to be hand finished accordingly. I guess I should cut to the chase and actually put mods in.

The first step I take is removing the variator faces and putting them together, I prefer to take it off the scoot. I place them face to face centered and put the belt on. Almost all the time for unmodified systems, the belt will not sit to the top of the tapered faces. here we see our first obstacle, this will limit our maximum attainable final drive and potential top speed. Conversely, with the CVT fully installed, if the belt doesn't rest on the boss (deepest between the faces), (not sure if this is ever the case) we are not able to achieve the lowest attainable ratio that will give the greatest torque. 

Ok so after typing up a storm with modifications I did, it is confusing to say the least without having a visual guide to make sense of things. I will have to prepare a few photos and add some arrows etc before I continue. My apologies for the delay.

Share this post


Link to post
Share on other sites

Posted

Thanks for keeping this easier  to understand. I think in most cases, the rider would like to get to the engines' "usable power band" as quickly as possible,(and stay there) not necessarily WOT. How does one accomplish that? By using sliders? Heavier or lighter rollers? By varying clutch spring or variator spring  tensions? Is it possible to adapt a larger or smaller variator from, say, a snowmobile or Jetski-and would there be any advantage? What effect-if any- does clutch "tuning" have on the variators performance? There's a lot to learn but I don't know where else to find some answers.

I looked up the Comet brand of CVT's. Their still around in some form and big with the go-cart conversions. Seems like they deal with the China copies also.

Those kits are physically sized somewhat between the snowmobile and GY6 CVT's. The Comet is more of a proper design than the GY6.

There is a 30 series kit that is asymmetrical and would probably be the best option.  You would have to use some of the kit and have to make up the adaptors. That would be an interesting project.

https://www.gokartsupply.com/tavapp.htm

Share this post


Link to post
Share on other sites

Posted (edited)

SD you sure do know your CVT systems and appreciate all the feedback from everyone, this is exactly what we need. To add to your previous post about grinding ramps and slider "channels" I also use it to manage the mechanical limits of the CVT. Also note that every almost every setup is different and has to be hand finished accordingly. I guess I should cut to the chase and actually put mods in.

The first step I take is removing the variator faces and putting them together, I prefer to take it off the scoot. I place them face to face centered and put the belt on. Almost all the time for unmodified systems, the belt will not sit to the top of the tapered faces. here we see our first obstacle, this will limit our maximum attainable final drive and potential top speed. Conversely, with the CVT fully installed, if the belt doesn't rest on the boss (deepest between the faces), (not sure if this is ever the case) we are not able to achieve the lowest attainable ratio that will give the greatest torque. 

Ok so after typing up a storm with modifications I did, it is confusing to say the least without having a visual guide to make sense of things. I will have to prepare a few photos and add some arrows etc before I continue. My apologies for the delay.

Something that makes these sorts of measurements and comparisons difficult is that these CVT's have no sort of "position memory" after handling if you will. If you spin up the CVT and spin down several times it will take a similar position set (level of the sheaves where  the belt rests), but if you were to remove the belt and reinstall, it will likely take a different position than it had before you removed it.

I found from experience that after any sort of adjustment or handling you need to exercise the CVT by spinning it up and down to have any sort of accuracy of position.

I guest there is also a possibility that the driven pulley Torque Sensing pins / slots could be the limiting factor. I've been trying to envision what would happen if the pins were removed and the CVT was spun up for testing. Probably have to have the wheel off the ground because any sort of load would cause the moveable sheave to freewheel. But if the belt actually traveled further in the pulley's then that would suggest that the slots are the limiting factor. Then maybe extending the slots would be a viable option.

Edited by SD12
1 person likes this

Share this post


Link to post
Share on other sites

Posted

Really good stuff guys.

                   :cheers:

1 person likes this

Share this post


Link to post
Share on other sites

Posted

I looked up the Comet brand of CVT's. Their still around in some form and big with the go-cart conversions. Seems like they deal with the China copies also.

Those kits are physically sized somewhat between the snowmobile and GY6 CVT's. The Comet is more of a proper design than the GY6.

There is a 30 series kit that is asymmetrical and would probably be the best option.  You would have to use some of the kit and have to make up the adaptors. That would be an interesting project.

https://www.gokartsupply.com/tavapp.htm

That sounds like a great build, I've toyed with the idea of just adapting a straight up centrifugal clutch and shoot for one ratio, but obviously it has its draw backs compared to a varied setup. I do prefer my odds with a chain compared to a belted CVT though, I'm not sure exactly how much we lose with the belt. I'm no go kart expert but I believe Noram and Bully make some mean clutches that are tuneable. I even thought about either making my own or adapting a manual clutch to run a single ratio. 

Share this post


Link to post
Share on other sites

Posted

Not sure how converting from a CVT to a centrifugal clutch  would work for you, switching from a transmission with clutch to a clutch. Might be fun for doing doughnuts in the dirt but pretty much unusable on the roads.

Efficiency: I read somewhere in the past that CVT's were less efficient than manual transmissions but better than automatics. I think they were also better than hydro-static's. That may have changed because I know automatics have evolved a lot in the last decade or so. Still CVT's aren't bad, especially in small engine machines.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now