Did the formulas needed on a piece of paper, then tried to create a DC from them to make it useful in SU (they gave me errors (DCs broke), but for fun I then did the “construction” in Geogebra:
(Excel would certainly be a good choice)
This geometry -with wheels of different diameters, though possible, is not recommended as the blade tracking adjustment (by tilting the idler wheel) becomes extraordinarily sensitive and could be unstable.
Could you elabotate on what you mean by unstable under exact conditions (as mentioned above)?
It’s much easier to control blade tracking and make fine adjustments if the wheel is large. Generally the tilt of the wheel only needs to change very slightly to change where the blade runs on the crowned surface of the wheel. The blade needs to run in the right place on the wheel to eliminate drift. If you don’t tune the bandsaw to eliminate blade drift, the rip fence needs to be angled to match the drift (that can be done but it isn’t ideal) and it makes a miter gauge or any jig that might need to run in the slot on the table useless.
As I wrote before, it makes more sense for a two-wheeled bandsaw to have both wheels with the same diameter. It’s simpler to make two wheels that are the same size than it is to make two different sizes. The rest of the bandsaw is easier to construct and align.
1 Like
Seems to me this topic has wandered astray, maybe just out of intellectual curiosity. So far as I can see, the OP never asked about using wheels of different diameters on a single machine, he asked about how to handle changing the diameter of both wheels. So the concerns about stability and alignment may be amusing but irrelevant?
1 Like
The OP also asked about modeling the blade.
Along these lines, modeling the straight portion of the blade from the tangent points of the wheels is fairly easy. However, there are a number of approaches to modeling how it wraps around the wheel itself. If the teeth are not especially critical, then a simple sawtooth function can be used. The parametric form of this in terms of ‘x’ from Wolfram is here: Sawtooth Wave -- from Wolfram MathWorld
Modifying it to map the sawtooth onto a cylinder using cylindrical coordinates gives these three equations in terms of ‘alpha’:
X = cos(alpha)
Y = sin(alpha)
Z = alpha - floor(alpha)
Varying the fixed parameters and casting the equations in terms of ‘u’ and ‘v’ Ruby functions yields:
X = 100.0 * Math.cos(u)
Y = 100.0 * Math.sin(u)
Z = v * (10.0 + ( 5.0 * ((40.0 * u / Math::PI) - (40.0 * u / Math::PI).floor)))
Using these equations in UV-PolyGen, I created a semicircular portion of a sample blade:
Model: bandsaw_blade.skp (628.6 KB)
This is by no means a suggestion to use this approach. As I said, there are many ways to accomplish this. I just thought I would toss in my two cents 
The plugin can be downloaded from here: https://sites.google.com/site/spirixcode/code/uv_polygen.rbz
Building a bandsaw that doesn’t have some sort of vibrational harmonics depends on the rigidity of the frame, the inertial masses of the wheels, and the dimensions / tensions of the blade itself. This paper explains it in full gory detail:
1 Like
You deserve an award for “Tangentially Related External Link with Obscure Speciality Information”. In my experience in this forum, your linked paper is the most difficult to understand (for us mere mortals) I’ve seen on this forum! (And I mean that in a good way!) 
Thanks for the information on modeling the blade. This thread has drifted a bit. My initial inquiry was how to determine the distance between centers of the wheels. That question has been answered and jimhami42 has given me a fine example of how to model the blade. Thanks to all.
Mike White
I have some experience with belt drives (a band saw is a very special case
of a belt drive) and in my experience symmetry of driveR and driveN wheels
makes things infinitely easier. Remember that axial pressure is put on
the blade when in operation cutting and this distorts the blade causing it
to track differently. If the disk driving the bandsaw blade and the idler are of
different diameters I fear that aligning and crowning of the wheels will be
an arduous project.
2 Likes
Thanks @slbaumgartner, @DaveR and @gregor_shapiro.
True. I wanted to use different radii, just in case, and… solve this in Geogebra in a general manner.
I do agree about one and the same radius for both wheels. I’ve seen Matthias Wandel’s videos on bandsaws several times in the past. Almost started to make one myself.
I know about the necessity to have a firm frame to avoid unwanted vibrations and a save band saw tracking. Large wheels help in decreasing bending stress in the band as well. (less bending).
A good surface (rounded) on each wheel helps keeping the band “on track”, taken that both wheels are aligned correctly.
I forgot about same radii and got carried away in the Geogebra adventure.
One thing about different radii, the larges driving wheel would have more grip on a band due to an angle >180 degrees. The idle wheel (in contact less than 180 degrees) doesn’t slip anyway. But this advantage doesn’t weigh up to advantages of one radius for both wheels when adjusting the wheels, band tension etc.[quote=“gregor_shapiro, post:29, topic:36470”]
… and in my experience symmetry of driveR and driveN wheelsmakes things infinitely easier.
[/quote]
I can’t argue with that.