Drive X Pack Mule

[joebiodiesel]

I figured I'd post my final results. After a little trial and error, I ended up with a 46" belt. It seems to work fine when run on the stand. I' see how it goes when i take it out tomorrow.
I really don't know why the math doesn't work, but I know where I ended up! I'd like to say it was a good learning experience, but it really wasn't. I still don't know how to get the right belt size without trying different belt sizes.
Joe

 

[control4userguy]

^A few other factors are taken into consideration when sizing a belt. Measuring circumference can go two ways- inner and outer. If you measured for inner and bought that way then the belt will come to you 2 inches longer (for A belt) or 3 inches longer (for a B belt). The other factor is where exactly the belt must rest on the pulleys. Measuring based on pulley diameter alone is incorrect. One must measure from the "working" diameter on the pulley which is where the belt cords lie on the sheaves. This probably would apply to your case though as you had to go longer.

 

[joebiodiesel]

That's sounds like it should work. Now that I have a working belt, let me try measuring the outside diameter of where the front belt rides, any try the math again. I'm sure I'm not the only poor bastard struggling with finding the correct belt size. With any luck it will help someone else.

Joe

 

[control4userguy]

... This probably would apply to your case though as you had to go longer.

I made a mistake. I should have said wouldn't apply.

 

[joebiodiesel]

My memory is terrible. I figured this once already. I already posted regarding my error on the OD of the front clutch.

Here is the math again.
Half the circumference of the top pulley - 10.2
13" Center to Center, twice - 26
Half the Circumference of the drive pulley(to the OD of the belt) - 5.5
That gets me 41.7"

The problem has to be that the belt on a CVT takes a looping path around the drive clutch, and that math works when you have perfectly straight paths. I wish there was a better answer for anyone that has to do this in the future. Trial and error was an expensive and time consuming process.
Regardless, a 46" belt works perfectly.

Joe

 

[control4userguy]

So, what belt is on there now (p/n)? Let us assume a B belt. 42+3+whatever looping you are talking about=46". It does depend on the type belt, the at-rest fitment (I mentioned about this early on, ie; where it sits on the small pulley) and whether or not your 46 inch "fits-perfectly" belt is inner 46 or outer 46.

 

[joebiodiesel]

So, what belt is on there now (p/n)? Dunno. I got it from a carting buddy. We tried several belts of different lengths.

42+3+whatever looping you are talking about=46". Here is the looping path a CVT belt takes.





It does depend on the type belt, the at-rest fitment (I mentioned about this early on, ie; where it sits on the small pulley) and whether or not your 46 inch "fits-perfectly" belt is inner 46 or outer 46.

Belt OD is 45+". Not quite 46". The "perfect fit" is evident by no extra space between the drive pulley and the belt, with no dragging on the ID of the belt. I was lucky in that I didn't even have to slide the engine to have it right where it needed to be. It is also evident in the witness marks on the drive pulley. They go right to the top of the sheave. If the belt was too long, it would not be polishing the pulley near the center shaft, and it would run past the outer edge. If it was too short, it would drag on the shaft and not make it to the top at full throttle.

 

[control4userguy]

The fitment around that drive shaft is what adds to the overall length. Belt calculators aren't designed for "loose-fit" applications like these torque converters. At the end of the day, the most sensible way of sizing a belt for a TC is to use a rope with approx. the same width as the belt you intend to use (5/8, 3/4). Wrap it around as it would if it were a real belt and then mark & measure. As long as the rope is resting in the sheaves as the belt would, it's close enough. Since belts are sized to the inch, you gotta make the call as to how you go about achieving proper tension.

Let us conclude this back & forth by saying the use of a web-found belt calculator probably will not work to your benefit unless other factors are taken into consideration. Arc correction being just one example.

 

[control4userguy]

In defense of the math, I hand-calculated and physically measured for a set of 3" pulleys on a TC backplate. This is a "no-slack" fitment on a centrifugal clutch drive to a jack-shaft. A 22" (#20 part number) fit. The calc returned 22.042" OC.

 

[joebiodiesel]

Cool! I'm sure I'll use those calculations again in the future. Maybe I'll need to buy some chain or something else of that nature.

I also mess with old tractors. My 1928 Rumely Do-All shed the fan belt a couple of weeks ago. It's a flat belt, so I just used my tape measure to get the belt length. I was getting a chuckle out of it thinking that I didn't have to worry about measuring ID or OD to get the right length. :smile:

Joe