CVT semi manual pulley adjustment

[coachgeo]

I may be wrong but I think this is how it works on a suzuki Bergman 650cc CVT scooter.

on the DRIVEN pulley the sheave with rollers in this case would have teeth on the outer edge. An eletric motor with a small gear on it very similar to a typical engine starter engages with these teeth. With the electric motor one can manually turn the sheave causing rollers to climb the torque ramp thus causing the spring to tighten and thusly move the sheaves closer together which forces the belt to ride further toward the outer circumfrence of the pulley thus changing ratio.

I may be off base here cause Im still a good bit baffled how the driven/secondary pulley ramp climbing rollers work. Search Bergman CVT

[coachgeo]

Yep, Im wrong. Though the above idea is an interesting concept in manually adjusting the secondary pully.

Anyway, the Bergman's adjustability is in the primary/drive pulley

see http://burgmanusa.com/bkb/650+CVT+Info

PS- This Bergman setup might be a good thing to get from a scraper for use in diesel bikes IF you got ability to create a computer and or some other form of electrically controlling the pully width.

[DieselFly]

How about the CVT from the Aprillia Mana 850 GT. It can run in tree different modes as a 7 speed sequential or in full auto CVT mode.
I cut and pasted the specs. When the new Mini came to the U.S. BMW had to program set shift points into the CVT program. Market reasearch showed U.S. drivers did not like the CVT concept and wanted something that shifted gears. So The Aprillia unit would be the best of both worlds. One might also not the engine is a Can-Am from BRP powering the Aprillia. Which makes me wonder if CVTech has their fingers in the tranmission end of things.

Gear box: Sequential with manual or automatic mode selectable by the user. 7 ratios in manual mode. 3 mappings (Touring – Sport - Rain) in Autodrive mode. Gear change by pedal or handlebar control. The user can switch from automatic to sequential mode and vice versa at any time.

Clutch: Automatic.

Primary drive: Belt.

Final drive: Chain.

Sean

[coachgeo]

Myself dont care if it "feels" like it is shifting. Just wondering if the ability to down shift for "passing", for quick "oh sh*t" cager or animal avoidances etc.

Also off road..... is ability to select gear position an advantage or is auto mode CVT? Or is Both? if so when like "sand"?

[DieselFly]

Good questions Coach. So who is going to go to an Aprillia dealer for a test drive? There is no dealer around me plus there is not a lot of riding weather left here.

Sean

[DieselFly]

I have spent some time reading test ride articles on the Aprillia Mana 850 as well as Youtube test rides. What I get is its CVT is different but not unliked. Here is one of those articles http://www.motorcycle-usa.com/29/2533/M ... eview.aspx

[freefuel]

who makes a semi manual CVT for a custom build? I like the idea of being able to run the engine at a fixed speed and use the CVT to govern the bikes speed.

yeah I know it will sound like a lawn mower droning away as I ride, I don't care what it sounds like.

[tappy]

I don't know anyone that makes one but I've started a design for one. Will be a while yet before I get around to making it though.

[sbrumby]

Just on this point I have a lathe that is a student lathe from a school. Originally it was a 3 phase. It has a CVT drive from the motor to the gearbox, this was originally opperated from a control board with a key switch which opperated
a small electric motor 3 phase that expanded or contracted the pulley. The reason it was like this was so the teacher had the key to prevent the child reving the chuck and job to warp speed. Anyway I dont have 3 phase so I replaced the drive motor. The speed altering motor with a simple manual speed adjuster, works perfect. So I dont see any reason why you cant have a system on a bike that allows you to be in control of the revs / speed.

[freefuel]

cool to hear about the lathe, I'd like to find or make a manual adjusted driven pully that will work with Commet style belts and drive pully (AKA rear pully) so I can make use of common replacement wear parts.

I'd like to see if it could be cable or lever actuated with a fail safe, in that if operator input is lost the unit will default to it's lowest speed or total loss of power transmission.

[tappy]

Based on the design work I've done thus far, it's possible, but there are issues - chiefly clamp forces.

If you under-clamp the belt for a given torque then it'll slip, you lose efficiency, and you reduce belt life.
If you over-clamp the belt you put a lot of hysteretic work into it, lose efficiency, and reduce belt life.

I've gleaned a bit of data & rules of thumb from existing systems & tuning guides, combined it with simple pulley mechanics / calculus and come up with some numbers. The system I've designed is for a Hatz 2G40 - a twin cyliner 990cc engine very similar to the Ruggerini / Lombardini twins. It makes about 24hp at 3600rpm, and 54Nm at about 2200 rpm.

That torque figure is key. In the lowest CVT ratio, when the belt lap angle of the driving pulley is smallest, I need about 3KN (300 Kg force) to clamp the belt to stop it slipping. If I'm still applying the same engine torque but in a higher ratio, the required clamp drops off a bit - in the highest ratios it's only about 1.5KN.

If I'm just trickling the engine at say, 10Nm, then I only need 600N (60Kg force) in low ratio and 300N in high ratio.

If you're trying to apply and hold 3KN with your foot or a hand operated cable then you're going to get tired very, very quickly. So clearly some kind of worm-drive or other system is needed so that it applies and maintains the force, and you just apply a position.

In this case, if you're trying to change CVT ratio with the throttle wide open you're still going to need either a very strong foot, or a very generous lever / reduction ratio in the control system. This will mean that when you want to short shift (or down shift rapidly in an emergency) you'll be hitting that lever about 30 times in a few seconds. This does not sound like a relaxing ride...

Alternatively, you could make a system with pre-defined ratios. Now what happens is that you roll off the throttle to reduce the torque to near zero, stomp on the lever and the CVT sheaves shift over to the next position. Once in position they are locked, and you whack the throttle open again.

So now you've largely lost the "Continuously" part of "Continuously Variable Transmission", but can at least have quite a few ratios and design a system with a good high over-drive ratio, so that the losses of the belt system are more than made up for by keeping the engine in its sweet efficiency spot.

Whether you use a manual mechanism to move the front sheave pair or the rear sheave pair you then need a mechanism to adjust the other set of pulleys and make sure there's enough clamp force being applied. The obvious choice for this is to use torque cams - rather like the driven shear pair of many belt CVT systems already do. This way, when there's not much torque being transmitted, the belt isn't clamped so hard, so there are fewer losses (as long as there's enough clamp to limit slipping). You'd probably combine this with a bit of spring force so that when you throttled off the belt didn't lose grip, then re-grip inconsistently and cause belt wear etc.

The thrust bearings used to hold the sheaves in position will cause a little bit of loss, but it's pretty small and ordinary sealed bearings should cope so it can't be that bad.

The main reason I looked into this was so that I could get a decent over-drive ratio for economical cruise, and avoid the big centrifugal variator system hanging off the LHS of the engine. With the system I describe above, there's nothing to stop me putting the manual bit on the driven sheaves and the torque reactive cams on the engine sheaves. This means I haven't looked at the possibility of using off-the-shelf sheaves, as typically they have cams on the rear sheaves. But I suppose I could always get a couple and turn them down to a smaller diameter for the front. Sounds like a plan forming...

[coachgeo]

....

The main reason I looked into this was so that I could get a decent over-drive ratio for economical cruise, and avoid the big centrifugal variator system hanging off the LHS of the engine. With the system I describe above, there's nothing to stop me putting the manual bit on the driven sheaves and the torque reactive cams on the engine sheaves. This means I haven't looked at the possibility of using off-the-shelf sheaves, as typically they have cams on the rear sheaves. But I suppose I could always get a couple and turn them down to a smaller diameter for the front. Sounds like a plan forming...

????

Lot of this made since till got to the last paragraph. Your probably right......... just myself...... Im getting confused by unclear to me terminology. for example you say:

...typically they have cams on the rear sheaves...

Huh? Rear sheaves ?: are you speaking of the inner most sheave of the Primary or are you talking about the primary itself......... or are you calling the secondary the "rear sheave"?. CVT's that use cams....... the cams are in the Primary (the pair of front sheaves) which is the one on the engine and not the the rear sheaves (called the secondary). Not all Primary's use cams. While some use cams that swings out due to G forces to change the width of the sheaves on the primary clutch. Others use weights and sliders and cause the sheaves to change by G forces a different way.

Also you state

...But I suppose I could always get a couple and turn them down to a smaller diameter for the front. ...

Huh the front (or what is called the Primary clutch/Pulley) already have cams.... (or weights) and making them lighter (or heavier) is common way to tune them. The secondary has Ramps...... no cams?

[coachgeo]

for other thoughts...... here is how GM does their semi manual adjustments of CVT........ via Hydraulics. The demonstration uses air pressure.... which opens up another idea for the DIY'selfer with way more tools and access to machine shops than the average bloke.

http://www.youtube.com/watch?v=c47caRqbbnE

[freefuel]

OK now I need to find the drawings for that Saturn car. after that demo it seems almost too easy to manually control that GM CVT trans. I'd mate it to a Kubota V2203 diesel engine and run set the RPM to it's sweet spot. then adjust the transmision to regulate the road speed.

[tappy]

Hydraulics work fine as the fluid is incompressible so dictates position, not force.
Air is compressible however so acts like a spring - it can apply force, but not position. It works OK in a simple demonstration like that where there's no torque being transmitted and the air is allowed to just whack the sheaves to the far stop. But imagine the following:

You're riding along at a particular CVT ratio, at a fixed speed, no incline so the torque transmitted is constant.
Now you roll off the throttle. This reduces the torque transmitted so the torque-cam on the driven sheaves fights back a bit and shifts you into a higher ratio.

Many of the effects might be beneficial:
Whacking the throttle open will tend to shift you into a lower gear, so you'll accelerate harder, and as you roll off a bit the CVT will shift up a bit.
Approaching an uphill gradient may tend to make it down-shift

Some not so:
When going DOWN a hill the reduced torque might make it shift into a higher gear, removing engine braking or even speeding you up
Any panic situation where you quickly shut off the throttle will speed you up - the CVT will shift up and the large momentum of the engine's flywheel will be transferred straight through to the back wheel

It might act like an interesting sort of "pre-selector" CVT, but making it controllable might take a bit of mucking about..

Controlling BOTH driving and driven sheaves with compressed air would allow you to make it do exactly as what you ask, but only with a fairly complex sense & control system that could sense what the sheaves were doing and adjust air pressure to keep them where you'd demanded.

[BertTrack]

Tappy,

With the Track i noticed nothing of a flightwheel inertia speedup.

When i let go of the throttle it shifts up instantly (too fast to notice a time period during which etc.) and since there isn't any added power coming from the engine it will remain there.

Since i'm not adding power it's not accelerating from the moment i let go of the throttle.

In the Tracks case it will run about 50% over idle while braking on the engine. Where the force of the weights in the primary are counter balanced by the spring in the secondary.

[tappy]

Interesting - maybe the pneumatic system would be OK.
It would certainly be less messy then an hydraulic system if it sprung a leak!

[Diesel Dave]

Interestingly as nothing is new in the world of motorcycles, Rudge's first TT winning bike had manually controlled CVT.

[coachgeo]

Based on the design work I've done thus far, it's possible, but there are issues - chiefly clamp forces.

If you under-clamp the belt for a given torque then it'll slip, you lose efficiency, and you reduce belt life.
If you over-clamp the belt you put a lot of hysteretic work into it, lose efficiency, and reduce belt life.

I've gleaned a bit of data & rules of thumb from existing systems & tuning guides, combined it with simple pulley mechanics / calculus and come up with some numbers. The system I've designed is for a Hatz 2G40 - a twin cyliner 990cc engine very similar to the Ruggerini / Lombardini twins. It makes about 24hp at 3600rpm, and 54Nm at about 2200 rpm.

That torque figure is key. In the lowest CVT ratio, when the belt lap angle of the driving pulley is smallest, I need about 3KN (300 Kg force) to clamp the belt to stop it slipping. If I'm still applying the same engine torque but in a higher ratio, the required clamp drops off a bit - in the highest ratios it's only about 1.5KN.

If I'm just trickling the engine at say, 10Nm, then I only need 600N (60Kg force) in low ratio and 300N in high ratio.

If you're trying to apply and hold 3KN with your foot or a hand operated cable then you're going to get tired very, very quickly. So clearly some kind of worm-drive or other system is needed so that it applies and maintains the force, and you just apply a position.

In this case, if you're trying to change CVT ratio with the throttle wide open you're still going to need either a very strong foot, or a very generous lever / reduction ratio in the control system. This will mean that when you want to short shift (or down shift rapidly in an emergency) you'll be hitting that lever about 30 times in a few seconds. This does not sound like a relaxing ride...

Alternatively, you could make a system with pre-defined ratios. Now what happens is that you roll off the throttle to reduce the torque to near zero, stomp on the lever and the CVT sheaves shift over to the next position. Once in position they are locked, and you whack the throttle open again.

So now you've largely lost the "Continuously" part of "Continuously Variable Transmission", but can at least have quite a few ratios and design a system with a good high over-drive ratio, so that the losses of the belt system are more than made up for by keeping the engine in its sweet efficiency spot.

Whether you use a manual mechanism to move the front sheave pair or the rear sheave pair you then need a mechanism to adjust the other set of pulleys and make sure there's enough clamp force being applied. The obvious choice for this is to use torque cams - rather like the driven shear pair of many belt CVT systems already do. This way, when there's not much torque being transmitted, the belt isn't clamped so hard, so there are fewer losses (as long as there's enough clamp to limit slipping). You'd probably combine this with a bit of spring force so that when you throttled off the belt didn't lose grip, then re-grip inconsistently and cause belt wear etc.

The thrust bearings used to hold the sheaves in position will cause a little bit of loss, but it's pretty small and ordinary sealed bearings should cope so it can't be that bad.

The main reason I looked into this was so that I could get a decent over-drive ratio for economical cruise, and avoid the big centrifugal variator system hanging off the LHS of the engine. With the system I describe above, there's nothing to stop me putting the manual bit on the driven sheaves and the torque reactive cams on the engine sheaves. This means I haven't looked at the possibility of using off-the-shelf sheaves, as typically they have cams on the rear sheaves. But I suppose I could always get a couple and turn them down to a smaller diameter for the front. Sounds like a plan forming...

Well tappy as usual I am wrong as hell. My wrong conclusions sit out in public like egg on my face a few post above. Wish manufactures would use similar names thru out. in the below link I discovered there are cams ALSO found inside some secondary clutches. Ill study these images and associated part names and re read what your say here and try to get it into my thick skull somehow what your meaning to convey.

http://www.manualsdir.com/manuals/16551 ... ml?page=38
http://www.manualsdir.com/manuals/16551 ... ml?page=39

[coachgeo]

Interestingly as nothing is new in the world of motorcycles, Rudge's first TT winning bike had manually controlled CVT.

Interesting how Rudge did it. Maybe it will inspire someone here

The History of CVT

The idea of employing a belt drive system with variable diameter pulleys is as old as about 1908, when the Rudge company developed their ‘Multi’ motorcycle. The Rudge Multi incor­porated a CVT system using a variable-diameter drive pulley fitted to the engine, and a fixed-diameter pulley mounted on the rear wheel, both are coupled by a leather belt. By rotating a small hand wheel positioned next to the fuel tank, the running diameter of the drive pulley could be varied to alter the gear ratio. At the same time, the position of the rear wheel was also slightly altered to maintain the drive-belt tension. This system was so successful that once the Rudge Multi had to be banned from the Isle of Man TT race in order to provide other motorcycles some chance of winning.

source: http://what-when-how.com/automobile/con ... utomobile/

More detail from another site.

Rudge Multi

The Multi gear

The Multi gear is one of the truly inspired devices of the veteran years. Doomed by the advent of the Sturmey Archer CS gearbox and its imitators, but inspired. Stay a moment to figure out how it worked.

Belt drive was cheap, smooth and well accepted by the motorcyclist of the pre-WW1 years in the UK. The major problem was the fixed gear ratio, so that there was always compromise between outright speed on the one hand, and flexibility and hill climbing ability on the other. Most later veteran engine pulleys could be adjusted on the roadside by winding the flanges closer together (to give a "speed" gear as the belt rode higher) or apart for the hills. Of course this change required an adjustment in the belt length to suit the new pulley diameter, and it was not uncommon to carry two belts of different lengths.

Enter the Multi gear. In the model illustrated below, from the 1914 catalogue, the protruding cylinder on the end of the engine shaft houses a clutch. Quite a neat device with around 50 small-diameter metal plates and a single coil spring. A similar clutch could be fitted to the single speed Rudge of the day to give the Free Engine model.

RudgeMulti.jpg (32.85 KiB) Viewed 29678 times

The Multi gear, 1914

But the clutch is not a necessary part of the Multi gear. Pullin's TT-winning Rudge, for example, has the Multi gear but not the clutch. So what exactly is the Multi gear?

The long lever in the tank-side gate is the gear change level. As it pivots about the engine mainshaft, it slides on four ramps on the crankcase and cause the flanges of the front pulley to close up (level forwards) or open (lever back). The pulley halves are splined together. So far, it is not so different from the adjustable pulley. But what about the belt tension problem? This is the inspired bit. If you study the figure above, you will notice a linkage between the front gear and the rear wheel, via a bell crank.

rearRudgeMulti.jpg (17.94 KiB) Viewed 29674 times

Rudge Multi rear wheel

The rear wheel is a testament to the fine wheel builders that Rudge Whitworth were at the time. It is in essence two wheels, built on separate, concentric hubs. Wheel number 1 - the "usual wheel" - carries the rim and tyre together with the inside half of the rear belt pulley. This part of the wheel is fixed relative to the frame in the usual manner, and uses 40 radial spokes. Now imagine a second hub, sliding over the hub of the first wheel in between the spoke flanges. Laced to this second hub by 40 more radial spokes (yes, that means 80 spokes in the rear wheel alone!) is the outer half of the belt pulley.

If you're still with me you will have the plan by now. As the gear lever is pushed forwards, the flanges of the front pulley close up, causing the belt to ride higher. Simultaneously, the outer half of the rear pulley is sliding outwards, causing the belt to fall lower into the rear pulley. Hey presto! the gear ratio becomes taller, and the belt tension remains constant. Infinitely variable gear ratios - at least between 3 1/2 and 7 to 1 - and constant belt tension. Good enough to last well into the twenties.

Copyright© Leon Mitchell 1998

Source: http://users.senet.com.au/~mitchell/bik ... /multi.htm

rudge rear wheel? one side of pulley fixed to wheel other side (outer side) moves to allow belt to ride high or low to change ratio

[tappy]

The Rudge system is fine if you have no rear suspension. With suspension the constant wheel movement would kill the belt.

[coachgeo]

The Rudge system is fine if you have no rear suspension. With suspension the constant wheel movement would kill the belt.

Agree. The interesting part that applies more to our projects IMHO though lies in the manual ramp arrangement of the Primary pulley. There might be something to that idea even today. Would not a standard Secondary still automatically adjust to a manually adjusted Primary?

[tappy]

Yes it would - I thought that's what we'd been talking about?!

[coachgeo]

Yes it would - I thought that's what we'd been talking about?!

ok... glad to be on the same page now