|
Post by 90GTVert on Dec 1, 2020 13:19:53 GMT -5
So the "mark the variator to see if you get full travel" theory is not a major factor anymore to reach high max speeds? Because Brent never had full travel when going WOT, right? In past tests the results work out. When I see more marker removed, I have better top speed at lower RPM. It's still giving an indication of the capability of the CVT to move the belt around. If the weights are too light, it's likely going to have less range on or off of the throttle and if the weights are sufficient to max out marker removal you will probably still see that you don't gain speed by going up any more on roller weight. That said, there's 1 comment that tells me my CVT is all wrong and a CVT should never do this. I have not filmed other CVTs under load. The only other time that I recall filming a CVT on road was the same OR kit on the 103 and it acted in the same manner. I have filmed CVTs on the stand on multiple scoots and they do what you would assume at first thought from what I recall. I don't see how you would expect the belt not to travel more when you let off of the throttle at speed though. The clutch is a long way from being able to disengage, so the only way for RPM to settle down is if the ratio does shift. If that didn't happen, it would act more like if you left a manual transmission in a gear and let off of the gas. With that thought, when I have ran some really stiff contra springs and/or light rollers, it is much closer to that effect where I get a lot of engine braking with RPM staying high even getting out of the throttle. It's something that you should be able to control to some degree, but I tune for WOT acceleration runs and the super light weights and/or stiff contras don't work out for me. I can make a stiff contra work, but then the weights have to be heavier and we're right back in the same spot. Torque driver ramp angles, pulley face angles and variator roller ramps should all play some role here as well. As eluded to by Hippo's coworker; it's more complicated than given credit for. The good news is, they can be sort of beautifully simple for basic tuning.
|
|
|
Post by 190mech on Dec 1, 2020 18:39:26 GMT -5
Yeah right,,Simple like a 2stroke!!
|
|
|
Post by hippo008 on Dec 2, 2020 16:58:05 GMT -5
Genie. Magic. Some engineer at Nissan 30 years ago found a magic lamp and wished for a reactive transmission. Each CVT comes packed with a miniature genie installed to work the pullies. The weights, belts and springs are just offerings to him to do his job.
|
|
|
Post by 90GTVert on Dec 4, 2020 16:33:47 GMT -5
|
|
|
Post by captincvmn on Dec 4, 2020 16:36:37 GMT -5
Looks supportive
|
|
|
Post by hippo008 on Dec 4, 2020 17:41:07 GMT -5
Brent, I'm not sure about the cvt belt travel situation you laid out. I would like to offer a counter opinion.
The CVT only responds to engine rpm, the cvt doesnt control it. When you're riding WOT the engine rpm is high because you made it high with the throttle action. When you let off the throttle there is not as much centerfugal force pulling the roller weights out from the center of the variator. The contra spring force starts to balance with the amount of pull the variator is creating and the belt should travel out on the rear pulley and in on the variator.
In my mind there isn't much reason for the belt to travel back out on the front pulley and in on the rear with less rotational force applied to the system. I'm no expert but IMO something is wrong with your specific cvt setup at WOT. The contra doesn't seem to want to balance with the outward force applied on the ramp plate in the variator.
Does this make sense? Did I crack the CVT code here? I don't know. Food for thought is all.
|
|
|
Post by islandscrub on Dec 4, 2020 20:07:12 GMT -5
Thought I'd pop back on the forum after watching the latest CVT video. The fact the belt sort of hovers around the middle under load totally blew out my assumptions of what's happening under the cover. If you're ripping it @ 14k under load I can't imagine how high it would be if the belt was at full travel ratio.
|
|
|
Post by islandscrub on Dec 4, 2020 20:12:45 GMT -5
BTW, that support spacer makes a whole lot of sense, and would hopefully solve the situation where you can't get the variator ramp off because the teeth are worn. Maybe even eliminate the wear altogether!
|
|
|
Post by 90GTVert on Dec 4, 2020 20:43:02 GMT -5
The CVT only responds to engine rpm, the cvt doesnt control it. ...When you let off the throttle there is not as much centerfugal force pulling the roller weights out from the center of the variator. The contra spring force starts to balance with the amount of pull the variator is creating and the belt should travel out on the rear pulley and in on the variator. In my mind there isn't much reason for the belt to travel back out on the front pulley and in on the rear with less rotational force applied to the system. I'm no expert but IMO something is wrong with your specific cvt setup at WOT. The contra doesn't seem to want to balance with the outward force applied on the ramp plate in the variator. Let me start by saying that I'd never deny being dumb as hell. I don't know that anything I say here is correct, aside from empirical evidence through my own experiences. I would argue that you can't say the CVT only responds to engine RPM and doesn't control it. If the CVT only responds to engine RPM without control of it, the engine will do as it pleases at all times. When you let off of the throttle it will return directly to idle. You can't forget the clutch/gearbox/wheel though. As long as the clutch is engaged, the gearbox won't allow the CVT and engine to do as they please. It all has to be balanced. The clutch doesn't really "care" what the engine is doing. People like to relate it's engagement to engine RPM, but it's really all about the RPM of the rear pulley. For that reason, the clutch won't disengage (at least with my setups) till maybe 10-15MPH. Till then, the gearbox is driving the clutch bell and subsequently the rear pulley. If you get out of the throttle, it has to change ratios to react to the combination of the gearbox in motion (which wants to remain in motion) and the engine (that would return to idle pretty quickly at that point if acted upon by no external force). What you get is the best compromise that it can provide (as setup) between gearbox motion and engine RPM. So, I do agree that the CVT reacts, but it also controls this balance. Now let me go back to this point specifically... When you let off the throttle... the belt should travel out on the rear pulley and in on the variator. Should it? If this happens, your engine is forced to maintain high revs when you get out of the throttle because the CVT and gearbox aren't allowing it to slow down. Forget whatever the rollers and springs are doing and just think about the ratio and how it would affect the engine and gearbox. If the CVT does as you describe and tries to downshift immediately, it would actually cause engine RPM to rise or at least stay steady. That or the belt has to slip so the engine and gearbox aren't sort of locked into changing rotational speeds quite so hand in hand. I think that most of us have only seen a CVT work on a center stand, with the wheel in the air. It usually looks like we imagine and as you describe. I would think that even though the gearbox and wheel and tire have the same mass and perhaps RPM as they will see on the road, the momentum/inertia is different when you have 500lb worth of scooter and rider linked to that tire's motion. Suddenly the engine doesn't have the same influence. If it wants to rev down, something's gotta give. Is it going to be 500lb worth of vehicle, or wouldn't it be easier for the CVT to change ratios? Again, just my thoughts. I'm a high school dropout so there's certainly no engineering degree to backup anything that I say. It could all be nonsense. One other thing. This plays into what I do have, which is time spent tuning CVTs and documenting results. Some people are telling me that I need a stiffer contra spring and lighter weights so that this doesn't happen. I will stand behind the statement that this is not a practical solution. I'm sure I could take those measures to change how the CVT reacts. You would be likely to see at least a less dramatic "upshift" when the throttle is released at high speed. The problem is, then everything else is out of whack. Then it wants to rev to 17,000RPM to accelerate and it's useless. It's like the old saying, cutting off your nose to spite your face. Great, it downshifts as everyone wants to see... but I can't ride it. It would have to be a system issue. Something is wrong with the entire setup. Torque driver angles, roller paths, drive pulley angles,... everything would have to be considered. Suddenly one does need a doctorate to figure out how to get a scooter down the road.
|
|
|
Post by captincvmn on Dec 4, 2020 21:14:56 GMT -5
Basically as long as testing data (which you’re very good at logging and charting) says you’re going fast on this setup vs a previous setup then I think it’s all good. Well, except for the 200 mile belt life. But you video as you ride, you’ve compared breaking points, you’ve tried to catch it breaking. Not really sure there is a way to figure out exactly the five W’s of the breakage. I’m sure it’s present in your mind as you ride though, wondering when it’s gonna let go. Just like watching for wildlife which can be less predictable than tin-tops.
|
|
|
Post by hippo008 on Dec 4, 2020 23:00:15 GMT -5
Brent I think our disagreement lies with how we define the control aspects of the CVT. In my mind everything is controlled by the variator.
The variator pushes the belt up into higher ratios as RPM increases or allows the belt room to drop when RPM decreases. All other aspects of the CVT are just controls on the variator. The way I see it the contra spring and rear pully act as the other side of the pendulum. They apply resistance to the variator to control how fast it can open, close and how much force the variator must apply to the belt in order to change the gear ratio.
All other aspects of the scooter like gears, tire size, static load and dynamic load are external factors that while important won't change the fundamentals of how the system functions.
|
|
|
Post by islandscrub on Dec 5, 2020 0:00:25 GMT -5
When you let the centrifugal force off the variator (throttle down) at the time the rear clutch shoes are still engaged with the bell, the force exerted on the variator is via the rear clutch rpm, therefore the wheel speed/gear ratio. This is a thought one pint in, so if I'm wrong just scold me.
EDIT: The ABSENCE of cetrifugal force allows the spring to push the variator back, thus giving you a better coasting gear ratio.
|
|
|
Post by islandscrub on Dec 5, 2020 0:32:01 GMT -5
The question is, does this phenomenon occur across all weight/spring combos? I feel like you'd tune different if you rely on belt travel vs pure engine rpm, or that not even a thing?
|
|
|
Post by 90GTVert on Dec 5, 2020 1:01:31 GMT -5
The question is, does this phenomenon occur across all weight/spring combos? I feel like you'd tune different if you rely on belt travel vs pure engine rpm, or that not even a thing? I don't really tune for belt travel. I may look at/for belt travel at some points, but for this particular scoot's CVT I have one thing that I really care about. WOT acceleration from 0-30-40-50MPH. For that tuning, basically I just need the RPM to stay as near it's peak output as possible. Peak output in this case not defined by any dyno data, but rather a ton of test and tune runs with RPM and ET recorded so I can see what RPM range nets the best performance. I agree though. If you were primarily concerned with belt travel, you'd end up with a much different setup than I have. Heavier weights and less drive boss length (shimming). You'd end up with lower RPM at top speed, though not necessarily going much or any faster, and I'd dust you when the light turned green. The contra spring, to me, mostly comes into play if you want to change belt control. If the belt slips, you need more pressure from the contra spring (and heavier weights). If you can't get into high gear with any weights that allow a good RPM range, you need a softer contra spring (and lighter weights). Regardless of what any video comment gurus may say, you have to balance the contra and rollers. There is no option to maintain an RPM range while changing only the contra spring or only the weights and I don't even understand how anyone that has actually worked on a CVT before could even make that suggestion... but it's pretty common.
|
|
|
Post by islandscrub on Dec 5, 2020 1:26:13 GMT -5
Brent,
Right, in general that's what you want, hanging at peak rpm and power throughout all speeds. I suppose I just expected that to be using up more of the variator!
|
|