We used center-to-center distances to calculate the length of belts and they all ended up way too loose, As an example, for a C to C distance of 9.73" and 24t pulleys, we calculated a 124t belt. It was too loose. We replaced one pulley with a 30t pulley and it was better, but that calculated a 9.54" C o C distance. Are the belts supposed to be loose, do we need to subtract a few teeth after the calculation to compensate? Any idea what we are doing wrong?
Could you post your exact calculations and pictures of your setup?
distance between centers is shown at 9.843 you said your CC on your robot was 9.73… thats a difference of .113… which equates to .226 of belt length so you’re almost .25 short in your belt distance.
A couple things I could think of.
How are you manufacturing this center to center distance?
Is the shaft that the pulley is on supported on both sides?
CAD design, CNC milled.
Yes
A thing to remember with belts for every X in CC distance changed your belt is seeing 2X that change
W couldn’t find a 123 t belt. A 122t belt has a CC of 9.65. Should we always pick the smaller option?
Where are you using these belts? can you add a tensioner to the system?
And no you shouldn’t just always pick the smaller option, you need to design things properly and that means understanding your system and the available products.
If you don’t have the ability to have a tensioner somewhere can you switch to sprockets and chain?
also i reccomend Vbeltguys.com
here is a 123t 5mm htd belt
9mm option
15mm option
I think this is the biggest mental change going from “poke two holes, add chain, take out a link, add a half-link if you must” to belts where that isn’t an option. So thank you for emphasizing that!
Having faced the same problem (belt calculators seem loose), we have a technique that seems to work well for us
- Calculate radius from pulley center to top of tooth using DP, which for HTD5 belts is (0.0627*T/2 - 0.03). For T=24, this would be 0.7220
- Set up a CAD drawing using the radius and the desired distance between centers. Measure loop length. For center distance of 9.73in would be a loop of 609.51mm
Closest standard belt would be 122. And since the belts are usually only a few dollars each, order multiples and sizes slightly larger and smaller.
If you’re using Solidworks (not sure about other packages) path length dimensions can be a huge help in belt powered mechanism design. Drive them with a equation that forces pitch circles and c-c distance to stock belt/pulley sizes.
So the 9.73" was calculated for a 123T belt? Just want to clarify.
Adding to what Eric said, if the belt size you calculated isn’t available, you need to recalculate the C to C for the next available belt size, either smaller or larger.
My team’s robot used some tensioners last year, I’ll try to find some pictures tonight to give y’all a start.
What does “too loose” mean?
Are you ratcheting / skipping teeth, or does it just seem loose?
Agreed with other posters above - you should design your center distance around the available product you intend to use, rather than picking the product with the closest center distance and expecting it to work without modification.
It’s easy to add a quick-fix tensioner if you have to - just put a bolt and some delrin in there to press on the belt slightly.
Also remember that HTD-5 pulleys have a negative offset of .7mm so if you have lets say 40 teeth that would make a normally calculated circumference of 200mm with an actual (2x.7 x pi = 4.4mm less which gives you 4.4mm of extra belt. So that is 4.4/25.4 = .173 inch of extra belt right there above and beyond what was mentioned above.
Former Gates engineer here.
Synchronous (timing) belts do not stretch because of Kevlar or fiberglass reinforcing strands. To keep proper tension you must have the centerline distances correct and a sturdy structure . Or a tensioner pulley, or a way to slide your driver or driven shaft to tighten the belt. But make sure it can’t come loose from vibration or impact in competition.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.