Modern Vespa

Just bought an MP3 400, and since it has rev counter, my knowledge of how the CVT works was turned upside down (I think).
My question is, when I accelerate, the engine immediately revs up. How can this be possible, without belt slippage? I thought the belt shouldn´t be allowed to slip in the pulleys, at least not more than «residual» slippages.
I always thought that you need to increase speed for the rotation in the rollers start to build enough force to «change gear». And the rotation in the rollers just happen when you start to go faster.
So, this belt slippage is normal in the way the CVT works?

It's difficult to express in words what I want to know

Regards,
Phil

Hi Phil: There's also a centrifugal clutch which engages at (roughly) 3,000 rpm. Before that the engine isn't coupled to the wheel. You'll also notice it when you decelerate, you'll feel the engine slowing you down, then it'll start to coast at roughly 20 mph.


Clutch bell in my hand with clutch behind (golden fan blades attached to what looks like a drum brake)

Has I said.. it's difficult to explain

I wasn´t considering the start. I meant when you are already running.
For instance, if you're cruising at 40 and you put full throttle, the engine revs immediately, however, the speed of your bike doesn't had time to change. Since the speed of the driven pulley is the same, the belt speed is the same, and the drive pulley should be the same. However, the drive pulley (connected to the engine) revs up. By doing so, something has to slip.
The clutch is always engaged, so it can't be that (unless there is something wrong with my cvt).

For that matter, is this the usual behavior of your bikes, too?

Thanks,

Phil

Both the variator and the clutch pulley can move in and out - the belt sits in between them in a position that's determined by load, revs, roller weights, clutch springs and driven pulley spring. The latter three are chosen to allow this acceleration to happen 'just right', but can be tinkered with for sportier or smoother take-off and acceleration.

Not so, that is where you have lost the workings. The driven pulley speed is the same, but the belt speed is not. And not because of slippage, but because the belt has moved to a different position between the two cones of the rear pulley.

This may help you understand
http://auto.howstuffworks.com/cvt2.htm

That's a great link Stoots. Never heard of a metal belt before. I wonder if...

Thanks for all the contributions, but... you're explaining me the BASICS. I already know the basics (I hope ). I'm talking beyond the basics. Let's start to think how all the pieces work together, how the complete set works.
Drive pulley, belt pulley and driven pulley work together, with no slippage between them (right?).
What controls the opening and closing of the pulleys? The rollers, better yet, the WEIGHT of the rolers, better yet, the centrifugal force applied to the rollers. And this centrifugal force comes from where? ROTATION.
How do you change the rotation of the drive pulley? Revving up the engine. HOWEVER, this is not enough. Don't forget the WHOLE set. How can the drive pulley rotate faster, if you have it connected to the driven pulley (via belt), and to the wheels? Please, follow me closely! You need, first of all, to increase the rotation of the driven pulley, which in turn will increase the rotation of the drive pulley, which will increase the centrifugal force over the rollers, which will, finally start to close the drive pulley, and, just then, the ratio provided by the cvt would start to change.

This was the way I though the system would work. However, has I said in the first post, the change in revs (by means of accelerating the drive pulley) are much faster than the change in velocity (driven pulley).
And this is the point where I think I'm losing the whole picture, or, on the other hand, maybe something is wrong with my bike.

This is hard to explain

Regards,

Phil

The Burgman 650 has an Aluminum Composite belt....

http://www.globalsuzuki.com/gcm/bigheart/002.html

the SECVT's innovative high-performance Dry Hybrid Composite Belt consists of 204 pieces of H-shaped high-strength aluminium blocks which are embedded with a pair of heat-resistant rubber tension members and covered with resin. The exceptional durability enables the belt to be run dry without the power loss characteristic of an oil bath, yet it also keeps the SECVT system lightweight and compact.


Dave

No big mystery, the gear ratio is changing rapidly as the engine encounters load. Just like a big downshift with a manual transmission.

HOW? HOW? Explain me the HOW this is done, the forces, the control.
I downshift in a manual transmission because my brain collects information about the engine revolutions, sound, vibrations, etc, and ME, MYSELF AND I decide that the thing needs a downshift.
Explain me HOW the CVT knows that it needs to downshift.

Thanks,

Phil

Since it's beyond my capacity to understand, I'm going with magic. Yeah, that's it. Magic.

I think it may be an order issue.

It is an increase in centripetal force by increased rpms on the driving pulley that causes the belt to move out (away from the center) which in turn causes and increase in speed. At least that is how I understand it.

So:
1. increase power
2. rpms begin to climb
3. diving pully closes from centripetal force on the rollers
4. gearing goes up
5. speed increases

It just happens so smoothly with a well designed cvt that there is no chance for slippage.

That is not always the case. I remember the cvts in our old ski doo snowmobiles. They slipped quite a bit (usually on take off) and I would go through a lot of belts every winter...along with bogey wheel springs but that is a different story.

here is a good one that shows part way through it holding a constant speed then adding power.

The answer you are looking for is the torque drive

It's a part of the rear pulley system.

Edit:
With increased torque, the torque drive widens the gap of the rear pulleys allowing for increased engine speed and power. It's an ingenious simple system allowing downshifting without slipping the belt.

It is the torque as it is applied to the drive pulley (variator) on acceleration. The torque needed to force the belt higher in the drive pulley is greater than that of the torque driver spring in the rear (driven) pulley. The torque driver spring is what pulls the driven pulley movable faces together upon deceleration or removal of torque from the drive system, thus shifting the the CVT into lower ratios. With a strong enough torque driver spring it is quite conceivable that you could cause stalling or belt slippage as the engine would not have sufficient power to pull the driven pulley faces apart.

Dave

Problem: How do the 2. rpms begin to climb (in drive pulley), without the 5. speed increases (driven pulley)?
Remember the WHOLE connected system.

Thats what the torque drive do.

Now, THAT'S something new! Torque!
That's something that I was thinking about empirically, but it's difficult to visualize. I haven't yet understood completely how this works, but in my mind, the fact that you increase the force (rpms) in the drive pulley «appears» to potentially have the effect of causing the belt to dive in the pulleys.
I need to put some thinking about it, during my commute home

In the mean time, could you elaborate a little more?

my mind is glowing

I was writing the previous post when you replied
I'm already in the «mood» to accept the effect of torque in the CVT. The way of doing so keeps a little blurry in my mind, but I'm putting some thougts deeply into the system.

Could you elaborate a little more on how this causes the ratio to change?


This can't be simple MAGIC

The "torque slave" assembly on the driven pulley has pins in angled slots on the shaft that I've always assumed allow it to respond to changes in torque in changing the opening of the real pulley halves. I've never seen a definitive description of how this works, but I'm surmising that it opens the pulley when the torque demand on the rear wheel increases, lowering the gearing. In other words, the gear ratio is sensitive to what is happening at both ends of the drive train, the engine rpms AND the torque demand at the rear wheel.

I guess you could call that magic.

After a weekend of thoughts (and a 500km (about 300miles) trip in my MP3), here it goes. I still doesn't get the whole picture, but maybe with everyone help we can figure it out. I also made some drawings. See bottom.
Also, I couldn't imagine how the spring in the driven pulley could detect the ratio. However, I could see the rollers having a crucial part in this.

For a 1st approach I considered the following simplifications:
1) CVT in higher ratio possible (belt can't go deeper in the driven pulley) - this way, the force generated by the driven pulley spring is not taken into account.
2) the belt only pulls in the top. The bottom part of the belt is left «hanging» - like a bike chain. I don't know if this is a simplification, or if this is really what happens. That's why the bottom part of the belt has that slang

Now, let's try to explain.
Pic 1: I considered that the drive pulley just work in the top quarter. That's the place where all the forces are applied and transmitted, from the pulleys to the belt and from the belt to the pulley.
Pic 2: What happens when you accelerate? You generate a momentum (torque). I represented some of the forces of this momentum. The «Inertial Force» means the force generated by the driven pulley, which try to resist the increase in acceleration.
Pic 3: Those 3 forces (F1, F2 and F3) are the part of momentum which is transmitted to the belt.
Pic 4: I just decompose those forces in their vertical and horizontal components. It should be decomposed in the same direction as the inertial force, but all the drawings were made when I insert the inertial force (sorry).
Pic 5: In the active area, we have a sum of forces applied which are of opposite direction than that of the «inertial force».

So, following my initial idea of the belt sinking in the pulleys, when you start in a static equilibrium of the system, the torque generated is equal to the inertial force. When you accelerate, the torque force is higher than the inertial force.
This is where I can «see» the effect, but I can not explain it very well. It seems obvious that the more force you apply to the pulleys, more the belt has to sink, because the centrifugal force generated by the rollers is not enough to counteract the torque anymore. By doing so, the vario changes.

I'm just trying to visualize how this works. A lot of variables were let outside of this approach, and I'm leaving some parts of the system outside. For instance, you may say that the same effect happens in the active area of the driven pulley, and both «sinks» will compensate and nothing will happen.

Well... is this just a pile of rubbish or makes sense to someone?

Phil

P.S. Couldn't find anything useful in all the internet searches I made. The belt CVT starts to seem a very complex and understudied thing.

I agree that these are difficult concepts to describe -- especially when one isn't sure he understands them fully himself -- but you may be overthinking this. Vector diagrams are all well and good, but I prefer to approach it intuitively.

Here is how I think it works: When you are cruising along with the transmission in a relatively high gear and demand downshifting from the transmission by suddenly rolling on more throttle (to climb a hill, for example, or to pass), I think the angled slots in the driven pulley hub respond to the increased pull on the upper portion of the belt (the part that is actually doing the pulling) by forcing the driven pulley closed (with the help of the big torque spring), effectively instantly lowering the gearing. The engine rpms stay relatively low until this shift happens, allowing the driving pulley to be forced open simultaneously. Then, once the shift has taken place, the variator takes over again just as in accelerating from a stop and the engine rpms will increase until the load balances the engine torque available.

Why else would those angled slots riding on pins exist in the sliding driven pulley hub if this were not the case?

I wish oopsclunkthud would weigh in on this. I'm pretty sure he knows all this stuff cold... don'tcha Patrick?

That's the problem with this CVT. A lot of information everywhere about the basics, but no information whatsoever about how this really works.
I never took apart a CVT, and I just once saw mine without the cover, when I replaced the air filter.

These angled slots you speak about, it's the first time I eared about them.
Can you put here some pictures?

Phil

Exactly. The slots act like a screw. When there is excess torque (from opening the throttle, it twists the pulley relative to the slotted cylinder. This screws the pulley in, lowering the effective gear ratio.

When the torque decreases the spring counteracts this motion. Excess torque decreases as the bike's speed start to increase, or if you roll back the throttle. Everything relies on equilibrium, with each factor (centrifugal force on the rollers, torque on the torque server, rear pulley spring) all constantly adjusting to each other.

It is quick, stepless, smooth and effective. An elegant mechanical solution to a critical issue. Without a torque server, you couldn't maintain speed up a hill, passing would be frightening and general performance would be very poor.

P.

+1 Spot on I think. I would add more point, in order for the torque server to work the belt must be able to slip. Notice that only 1/2 of the torque server moves on slots thus there is a twisting motion relative to the fixed side. This adds an additional factor. The belt has to pull / push the torque server to its new position while overcoming the friction created by the fixed half of the server.

Got it. Understood the concept. I'm trying to imagine how all this works together. Can't believe I never saw it clearly explained, despite the «angled grooves» appeared in one of the best explanations of CVT I discovered: http://www.teamcalamari.com/zuma/variator.html

Well, now I have new terms to search

Phil

That's a great way of describing it, Paul. I could envision it in my mind's eye, but your description is clear and concise.