MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now


Parametric Timing Pulley v4

by JamesNewton Dec 3, 2012
Download All Files

Thing Apps Enabled

Please Login to Comment

http://www.sdp-si.comhttp://www.sdp-si.com not only has a great calculator, but they also sell belts! Prices are fantastic, and they provide the full specs for the belt, unlike McMaster-Carr. Sadly, they seem to be discontinuing some of the belt lines. Hope the one you need is still available.

SDP/SI has an absolutely mac-daddy timing belt / pulley center distance / ratio calculator.


Please let me know if you run into any problems by posting the parameters you used here so I can try to duplicate and correct the issue. 

Any other features you want to see? I have one vote for an option to support a lego motor shaft. 

I found that printed pulley was not good, printout with below settings did not fit the belt (too small).
pitch = 2;
notchDepth = .95;
beltWidth = 6.5;
beltThickness = 0.7;

I looked into the code and found an error in two places:

    outerRadius = circumference/PI/2-beltThickness;
    innerRadius = circumference/PI/2-notchDepth-beltThickness;

The above statements should be replaced with:

outerRadius = circumference/PI/2-beltThickness/2;
innerRadius = outerRadius-notchDepth; 

After this correction, the model is much better, printed part fits the specified belt.

Hi Bojans, Thanks for digging into it. One question, when you measured the belt thickness, are you sure you excluded the notch depth? It may be that the description needs to be updated, but I notice your changes basically divide the beltThickness in half. Most belts have teeth that are about half as deep as the distance from the bottom of the belt to the tip of the tooth. In other words, changing the program as you specified would, in most cases, exactly compensate for having entered the total thickness of the belt and teeth, instead of entering the thickness of the belt without the teeth.

Hi James, yes I did.
The thing is, when belt is bent around the pulley, the belt base contracts at teeth side, expands at back side and the middle (where the reinforcement (glass, cloth) remains the same, that is why ther eis division by 2.
This holds for the belt I have (GT2, perhaps it would be different for some other kind of belt I imagine the belt with steel core may behave differently).

Hmmm... yeah, that makes sense. I'm nervous about changing it because it's worked so well for me and other people. E.g. I just printed a huge 6 inch diameter pulley as a test for a CNC conversion and it was perfect, once I gave it the right numbers. I think I will leave the comment and other people can report what they run into. They can effectively do the same thing by dividing their thickness number by 2.

I think a comment in a code about belt internal structure and behaviour would be OK...
Or introduction of one more parameter (position of reinforcement inside the belt base), to adress the issue.
Please note, this issue affects small pulleys much than large ones, in my case it was about 0.7mm difference in diameter.
I first noticed a problem with 120 teeth, 2.5mm pitch pulley, it did not quite match the belt and when I enlarged it's diameter by 0.7mm (by scaling up in Repetier, it was OK.
60-teeth pulley had the same absolute diameter error (required scale to compensate this was different of course).
Looking at the OpenSCAD preview radius, based on belt pitch&thickness and the actual belt bended I saw the difference, so I went into the code and concluded there is an error. Those differences between display and bent belt dissapeared after code correction.

Maybe the next version will add support for a captured nut for the set screws? The one wheel has it:


Highly Configurable Wheel (One Wheel To Rule Them All)