When cutting this out on our machine, the belt is pretty tight and won’t make a full revolution around the gear. I took .003" extra off and it fit better but not great. I’m sure it will work but I want to make the second one better
Any advice? Should I make my own tooth profile? Any advice for machining it?
I did make a different pulley with a profile I generted on the router using a drill to cut the pulley teeth so they were round and accurate and it seemed to work better than this one
Can’t speak to the profile you’re working with, but in the off chance it helps at all, I did come across this HTD pulley generator recently that I’ve found works quite well. Easy to export to STEP or other formats too so you can edit them later. I’ve printed a few and they seem to all work well.
The HTD profile is actually pretty simple. Just a few key dimensions. So you could double check the MKCad generator by sketching the profile using the math described here: Timing Pulley Design Tutorial
As noted in the thread, Ty later updated the MKCAD pulleys to use the correct geometry, which is as pictured above. It’s fairly easy to tell the difference between the old geometry and the new-- the new is as posted in this thread, the old looks more like this:
I’m not sure why this wouldn’t reproduce sufficiently on a given machine, but honestly I didn’t have an issue with the old geometry or the current geometry on my printers, and I’ve never tried to reproduce either on a CNC.
Adding a bit to what’s been posted… The MKCAD cad profile is only very trivially wrong now, and I’m confident if machined w/ accurate tolerances it would work fine.
However, most teams are unable to router or print to accurate enough tolerances, so offsetting the face .002-.005 (diametrically I believe) seems to be enough to cover those manufacturing errors.
This .005 number seems to pair with what both Vex and AM seem to do on their 42T and other large plastic pulleys of increasing the size of the root to give more room for the belt tooth. The pulley tooth looses no height with this change
This is a 42T 5mm MKCAD pulley with no offset, and a Vex 42T Pulley.
For comparison, here is the same two pulleys but with a -.005 offset for MKCAD. You can see that the bottom of the roots are now roughly tangent, but the MKCAD tooth is -.0025 shorter. This seems fine and is what 2135 ran for 2021 offseason and 2022 season, but taken to the extreme (numbers larger than -.01 maybe) I’d expect belts to slip under lower load and wear faster.
As the least involved MKCAD member, I’d advocate that MKCAD makes no change to these files as they are accurate enough to use, and that teams need to determine the appropriate offsets to make for their own manufacturing methods.
Thanks for catching this. In my port to the better profile from the previous thread, I never updated the fit adjustment parameter. I’ve updated the document to expose two variables from the equations that derive the profile.
ExpansionFactor seems to adjust how wide the valleys in the pulley are
CompressionFactor seems to adjust how deep the valleys in the pulley are. At the moment, they can only be made shallower.
Here’s a few screenshots comparing a vexpro 36T 5mm HTD pulley and the MKCAD configurable pulley, with these new adjustments. Red is the intersection between the pulleys:
