However if we want to speak in terms of pulley travel, I am restraining the movement of the pulley. By controlling the engine RPM. By manipulating the accelerator. With my foot.

That works in an automatic car transmission, because of the torque convertor, which the CVT in this case doesn't have. Also in an automatic transmission they have one-way or overunning clutches which allow the power to dissipate when not engaged. The CVT, or at least the old ones I'm familiar with didn't do that, you had to come down in RPM below the engagement level to slow down. Maybe the newer cart ones don't work like that, I don't know. You could run them at a slow speed by just never getting to full throttle, but to back off you needed to let the RPMs get below engagement.

This thread was born in discussion of that device and remains so. I do not understand why you think the only solution is to physically limit the pulley travel, or why any other ideas you consider off topic.

It's your thread I'm just saying how I see things. Never said it was "off topic" just that it has changed. I'm giving things I know will work from my experiences with snowmobiles.
Then why not govern the engine speed from the pulley movement? Like they do with a fly ball governor.

 

why not govern the engine speed from the pulley movement? Like they do with a fly ball governor.

The more I think about it, the more I converge into two options:

• Use mechanical means: a fly ball governor is probably the simplest.
• Use electronic means: add an encoder or resolver to the wheels, and then have circuit activate a valve that will restrict fuel (or air) intake when a certain speed is reached.

 

That works in an automatic car transmission, because of the torque convertor, which the CVT in this case doesn't have. Also in an automatic transmission they have one-way or overunning clutches which allow the power to dissipate when not engaged. The CVT, or at least the old ones I'm familiar with didn't do that, you had to come down in RPM below the engagement level to slow down. Maybe the newer cart ones don't work like that, I don't know. You could run them at a slow speed by just never getting to full throttle, but to back off you needed to let the RPMs get below engagement.




It's your thread I'm just saying how I see things. Never said it was "off topic" just that it has changed. I'm giving things I know will work from my experiences with snowmobiles.
Then why not govern the engine speed from the pulley movement? Like they do with a fly ball governor.

What you describe is certainly different than how mine works. I wish I had experience with the old style you describe, so I we could understand where the other is coming from. The best I can do is describe the operation of the one I have. Please don't take it the wrong way if the following sounds like I'm talking down to you. I'm going to explain it as if explaining to someone who has never heard of a CVT; maybe that level of detail will tease out where the difference lies between this one and the ones you're familiar with. I'm actually curious what the difference is.

The weights in the primary pulley are slung outward by centrifugal force. The faster the RPM, the more force. The weights are not allowed to sling directly out (perpendicular to the shaft) as they would like to; rather, as the weights travel outward, they are forced along an incline back toward the engine. On the other side of the weights, is the moving half of the pulley. So as the RPMs increase, centrifugal force increases, the weights move further out and the sliding half of the pulley moves closer to the engine. The sliding pulley half's movement is opposed by a spring. Below about 1500rpm the spring is stronger than the centrifugal force of the weights, so the pulley remains open. Between ~1500rpm and ~3000 rpm the force of the weights is enough to force the pulley closed by some proportional amount. After 3000rpm the pulley is fully closed. Coming down in speed from say 4000rpm, once crossing past the 3000rpm mark, the pulley begins to open again, as the spring force is once again stronger than the centrifugal force of the weights. Between 3000 and 1500rpm deceleration, the pulley opens back up proportionally, and below 1500rpm it is once again fully open. If RPM is varied up and down between 1500rpm and 3000rpm, the pulley can be seen opening and closing. This is the behavior of it whether or not a belt is on it. When a belt is on it, this closing and opening of the pulley results in the belt riding higher or lower in the pulley.

The secondary pulley is a simpler device. It has a spring holding it closed. When the belt is tightened (by the closing of the primary pulley) the belt is drawn deeper into the secondary pulley, forcing its halves apart and compressing its spring.

Now there is some complexity in the relationship between the primary and secondary pulley and this is where I'm a but shaky in my understanding. The secondary pulley is not exactly as simple as I described. When it is forced open, it is not forced directly open; there are helical lands that force it apart, and with a twist. This twist I believe transmits some torque feedback back to the primary pulley. I believe that if there is too much torque on the secondary, the helical angle of those lands applies a compression on the pulley halves, assisted by the spring, which forces the secondary back toward the closed position, which forces the primary pulley back toward the open position, in some sort of choreographed automatic downshifting dance.

There is no one-way bearing on this CVT but other more expensive ones have them. At idle, the belt rests against the rotating shaft, so the belt cannot be tightened too much or else the friction of the belt on the shaft is enough to spin the secondary pulley and make it creep while at idle. I believe I could get even more ratio range if I were able to make the belt tighter. This is what the one-way bearing allows you to do, I believe.

So the culmination of the above is an operation very much like an automatic transmission. Press gas to go, press harder to go faster, press lighter to go slower. And you don't need to come all the way to a stop (or all the way to idle) to slow down. If you are accelerating, you can change your mind and immediately decelerate, and vise versa.

 

The more I think about it, the more I converge into two options:

• Use mechanical means: a fly ball governor is probably the simplest.

Not to be rude, but just to save you and anyone else the time, we can take this and any other mechanical means off the table. I am not interested in making any special mechanical adjustments each time there is a new driver. This function will be controlled by the PLC, one way or another.

• Use electronic means: add an encoder or resolver to the wheels, and then have circuit activate a valve that will restrict fuel (or air) intake when a certain speed is reached.

A valve.... I think you might be onto something there. This could be much simpler than the cable sheath de-schnoodler that I'm currently cooking up. Truth be told I am having issues with it and it would be happy to give up... eh, I mean choose a better solution. I have seen some setup before, I believe on a pressure washer, where there was some sort of an unloader apparatus on the pump linked to a secondary butterfly valve in the carb, almost like a choke, so that if you weren't using pressure the engine would throttle down until pressure dropped again. If I could fit a butterfly valve between the carburetor and air filter, it could be servo-operated and offer all the positive aspects of drive-by-wire without introducing the possibility of runaway operation. I think I like this. A lot. Thank you. I will go tonight and see what can be done.

 

Maybe I could just highjack the existing choke instead of adding another butterfly valve. "Choke mode" could just as easily be an electronic function as a pull knob.

 

Not to be rude

My feelings are much harder to hurt than that, my friend...

As for the electronic idea... maybe a resolver or encoder at the wheels won't be needed if your cart is equipped with a speedometer that the circuit could interface with instead.

 

My feelings are much harder to hurt than that, my friend...

As for the electronic idea... maybe a resolver or encoder at the wheels won't be needed if your cart is equipped with a speedometer that the circuit could interface with instead.

I will be picking up pulses from somewhere, not sure yet where exactly. Perhaps I will drill & tap a hole in the transaxle to count gear teeth. I would rather get it from somewhere else though. Not a lot of good options.

 

Perhaps I will drill & tap a hole in the transaxle to count gear teeth

That's probably best... using a simple and sturdy inductive sensor to do the task... those things are extremely resistant to dirt and contaminants.

 

The weights in the primary pulley are slung outward by centrifugal force. The faster the RPM, the more force. The weights are not allowed to sling directly out (perpendicular to the shaft) as they would like to; rather, as the weights travel outward, they are forced along an incline back toward the engine. On the other side of the weights, is the moving half of the pulley. So as the RPMs increase, centrifugal force increases, the weights move further out and the sliding half of the pulley moves closer to the engine. The sliding pulley half's movement is opposed by a spring. Below about 1500rpm the spring is stronger than the centrifugal force of the weights, so the pulley remains open. Between ~1500rpm and ~3000 rpm the force of the weights is enough to force the pulley closed by some proportional amount. After 3000rpm the pulley is fully closed. Coming down in speed from say 4000rpm, once crossing past the 3000rpm mark, the pulley begins to open again, as the spring force is once again stronger than the centrifugal force of the weights. Between 3000 and 1500rpm deceleration, the pulley opens back up proportionally, and below 1500rpm it is once again fully open. If RPM is varied up and down between 1500rpm and 3000rpm, the pulley can be seen opening and closing. This is the behavior of it whether or not a belt is on it. When a belt is on it, this closing and opening of the pulley results in the belt riding higher or lower in the pulley

Thought about this all afternoon while running my mill. And came up with also explaining with RPMs.
Lets say the engagement RPM is your 1500RPM. So to get the cart to move you need to go over 1500RPM an unknown amount, to over come ground friction and weight. But then lets say after it gets to the desired speed it only takes 1200RPM to maintain that speed. Where would you set your throttle limiter? Which as I understand you, is what you want to do, but maybe I'm not understanding you correctly.

Think about how your car works, in lower gears(dive pulley apart) it takes more RPMs to get moving. As speed increases and gears shift(drive pulley coming closer together) it takes less RPM because of the gear ratio. Think about how cruise control works in a car. Your going a long at the desired speed, and come to a grade, the engine gets more throttle to try and keep the set speed. The very simpleness of the CVT doesn't work like that, it counts on the driver to change the throttle to keep the speed. The operator/driver replaces the cruise control.

Your explanation of the CVT you have is pretty much like the old ones, but I'm thinking the main difference is how a cart interacts with the ground compared to a snowmobile. The snowmobile was pretty much on or off,pulling or coasting/sliding over the surface. The belt in contact with the ground had bars(don't know the real term) that acted like a one way driver, if power was on it dug into the ground, with power slowwed down they bent over to slide on the ground.

 

then have circuit activate a valve that will restrict fuel (or air) intake when a certain speed is reached.

The engine is not going to like that. Leaning out the mixture(cutting fuel alone) will cause a burnt valve or piston at the worst and backfiring at the least. Adding more fuel and less air(choke or what ever) will carbon up the plugs fast and build up carbon in the cylinder.

 

Thought about this all afternoon while running my mill. And came up with also explaining with RPMs.
Lets say the engagement RPM is your 1500RPM. So to get the cart to move you need to go over 1500RPM an unknown amount, to over come ground friction and weight. But then lets say after it gets to the desired speed it only takes 1200RPM to maintain that speed. Where would you set your throttle limiter? Which as I understand you, is what you want to do, but maybe I'm not understanding you correctly.

Think about how your car works, in lower gears(dive pulley apart) it takes more RPMs to get moving. As speed increases and gears shift(drive pulley coming closer together) it takes less RPM because of the gear ratio. Think about how cruise control works in a car. Your going a long at the desired speed, and come to a grade, the engine gets more throttle to try and keep the set speed. The very simpleness of the CVT doesn't work like that, it counts on the driver to change the throttle to keep the speed. The operator/driver replaces the cruise control.

Your explanation of the CVT you have is pretty much like the old ones, but I'm thinking the main difference is how a cart interacts with the ground compared to a snowmobile. The snowmobile was pretty much on or off,pulling or coasting/sliding over the surface. The belt in contact with the ground had bars(don't know the real term) that acted like a one way driver, if power was on it dug into the ground, with power slowwed down they bent over to slide on the ground.

I think this might be the disconnect between us. What I'm trying to describe is a dynamic throttle limiter, not a fixed throttle stop. It would be in a PID loop, reacting to changes in vehicle speed. Below the setpoint it would be totally out of picture, allowing the operator to use the full range of throttle. Above setpoint (or on approach to setpoint) it would function exactly as a cruise control does.

 

The engine is not going to like that. Leaning out the mixture(cutting fuel alone) will cause a burnt valve or piston at the worst and backfiring at the least. Adding more fuel and less air(choke or what ever) will carbon up the plugs fast and build up carbon in the cylinder.

I don't think we have to worry about leaning out, at this point all we are talking about is a butterfly valve.

Is the existing butterfly valve in the carb not doing the same thing as we are discussing? My understanding is, the throttle is connected only to a butterfly valve. It is not connected to any kind of fuel regulator. The fuel dispensed into the cylinders is a function of the amount of airflow through the butterfly valve (venturi). Close the valve, less air, therefore less fuel. Is that wrong? I do not claim to be a small engine expert. If that's right, then how would adding an upstream secondary butterfly valve effect the air:fuel ratio?

 

The engine is not going to like that. Leaning out the mixture(cutting fuel alone) will cause a burnt valve or piston at the worst and backfiring at the least. Adding more fuel and less air(choke or what ever) will carbon up the plugs fast and build up carbon in the cylinder.

Ok, here's another idea.

How about a screw actuated through step motor opposing the throttle, adjusting backwards and forwards as necessary to limit the speed?

Something like this:

​

 

Close the valve, less air, therefore less fuel. Is that wrong?

That's the way I understand those types of motors work. Their fuel demand is proportional to the air intake. A diesel engine, on the other hand, works the other way around.

 

Ok, here's another idea.

How about a screw actuated through step motor opposing the throttle, adjusting backwards and forwards as necessary to limit the speed?

Something like this:

View attachment 212859​

Ok I didn't want to go into detail about my current failure (everyone can be a hero online only admitting their success) but it is relevant to your suggestion so here goes.

What I am in the middle of making is a stepper-controlled linear actuator to (in effect) move the anchor point of the accelerator cable sheath nearer to or further from the accelerator lever. I did the math on speed and force needed; worst case, it needs to overcome the spring tension (30lbs) of the accelerator lever at full throttle and I decided on 2s as the timespan to go full stroke. I built my actuator (mostly) and discovered some dirty details about steppers that I didn't know. Due to the driver I bought, it has no balls in the midrange of its speed. I need a more expensive driver like a gecko and I don't want to buy it unless I know for sure it will fix my issue.

So while I do like your suggestion, maybe even more than my own idea (it would give tactile feedback to the operator), it would be even more demanding of my actuator. I don't think I could do it with a stepper unless it was a real monster. It will have to overcome whatever force a human can put on it, which might be a grown man who doesn't realize what's going on and fighting the thing with a good 300lb stomp or two. He might win, for good.

I really want the butterfly valve to work. That's the best idea yet in theory.

 

It will have to overcome whatever force a human can put on it,

Not if you counter-stress it with an adequate sized spring. The spring would take the extra travel that the human would apply to the throttle while the actuator is stopping it from moving any further.

 

Well I did a bit more reading about the operation of carburetors, chokes, etc. and it seems I had an incomplete understanding, no surprise there. The choke (and therefore any other valve placed upstream of the carb) does effect the air/fuel mixture. It creates a vacuum which draws more fuel and make the engine run rich. So I think in order for a valve to work in this application it would need to be inserted between the carburetor and the intake. Unless that causes some other problem that I don't know enough to consider.

 

Is the existing butterfly valve in the carb not doing the same thing as we are discussing? My understanding is, the throttle is connected only to a butterfly valve. It is not connected to any kind of fuel regulator. The fuel dispensed into the cylinders is a function of the amount of airflow through the butterfly valve (venturi). Close the valve, less air, therefore less fuel. Is that wrong?

The choke (and therefore any other valve placed upstream of the carb) does effect the air/fuel mixture. It creates a vacuum which draws more fuel and make the engine run rich.

That is what I said. It makes no difference in any size engine with a carb.

So I think in order for a valve to work in this application it would need to be inserted between the carburetor and the intake. Unless that causes some other problem that I don't know enough to consider.

Doing that, a down stream butterfly, will also make a lean mixture and works the same no matter where in the intake it is.

I don't know your kids ages but wouldn't it be both better and easier to teach them how to drive this? Not trying to be rude but that's what makes the most sense to me. Maybe by you being a passenger and having them try different speeds while driving they will learn what they can handle. Unless they are very young.

 

A diesel engine, on the other hand, works the other way around.

A diesel has no intake butterfly. It adds more fuel for more speed/power output. Before digital injection came about it used a valve the worked like a distributor in a spark ignition to add the right amount of fuel to each cylinder. It uses the fuel to regulate the speed and power. It can do this because of both it's higher compression ratio and no spark plug to make an early ignition of a too lean mixture.

 

That is what I said. It makes no difference in any size engine with a carb.

Doing that, a down stream butterfly, will also make a lean mixture and works the same no matter where in the intake it is.

Ok. I don't understand why that is, but I recognize you know more about this than I do so I'll give it a rest for now.

I don't know your kids ages but wouldn't it be both better and easier to teach them how to drive this? Not trying to be rude but that's what makes the most sense to me. Maybe by you being a passenger and having them try different speeds while driving they will learn what they can handle. Unless they are very young.

Not an unreasonable suggestion and I probably could do what you suggest; my kids are good about following direction and anyway they would probably be too scared to go even as fast I consider to be their maximum safe speed. But as I said earlier, I have certain lines I draw. Every parent does. I don't make them wear knee pads to ride a bicycle but neither do I let them handle firearms unsupervised. Letting them operate a vehicle capable of 60mph unsupervised is in the same category in my mind. I might as well hand over the keys to my truck.

Their ages are 13, 8, and 6. I don't think the 6 y/o will be allowed to drive it until she is 20. She is too willful and not afraid of anything. The 13y/o is pretty responsible. I think I will let her have the option to go fast once she has a bit of experience. It will cut down on the amount of instruction and practice she will need in couple of years when she gets her learner's permit.