Belts 9mm or 15mm?

Only our Rookie year did we use chain, in 2011. Since then we have used belts with a 99% reliability of our drive train. In 2014, we went to 9mm wide belts figuring that a factor of safety of 5 was enough, as we were trying to fit them in 1 x 2 tubing. The small sprockets and the narrow belts just didn’t hold up in defensive play in Rebound Rumble. We quickly became experts at changing a complete drive train side in 5 min, going through 32 belts that year.

Since then we went back to the 15mm belts and use 1.5 x 2.5 tubing to fit them and the sprocket in. Both 2015 & 2016 we had NO drive issues nor repairs for the lifetime of the robot.

So in short, if you are going to use them for drive and plan on doing alot, 15mm wide are the only way to go.

I am curious as to how you got a safety factor of 5 while fitting belts inside 2’‘x1’’ tubing. Even using (what I’d consider to be) sufficiently large pulleys, FRC drives are usually skirting (or outright violating) Gates’ stated torque ratings for 9mm belts.

Would this be on 9mm or 15mm belts?

Sorry, is the 24T small enough to be used inside .062" 2x1 or 2x2 tubing? I can’t remember.

We used 42T pulleys on 8" wheels both years. The test drivetrain used 42T pulleys with 6" wheels.

Only if you don’t rivet or bolt to the top or bottom. I’d recommend 3x1.5 if you add a counterbore to your pulley for the bearing clearance or 3x2 if you do not. This gives you plenty of room for hardware so you never have to worry about it. You can probably do 2.5" tubing if you can find it?

Just because I’m curious, why the switch from gear drives? I still remember seeing your 2013 bot at the Smoky Mountains Regional and being in absolute awe. I also remember yours being the best drivetrain at the Regional.

In answer to your question, yes, 9mm wide (5mm pitch) HTD belts from VexPro can be used on an FRC drivetrain. This will save you space in one dimension (width), but cost you space since your pulleys will need to be larger. I would be uncomfortable using anything smaller than 30T 9mm pulleys with 4" wheels and a reasonably geared speed. If you’re using bigger wheels or active belt tensioning or voltage/current ramping or anything else that affects how your robot drives and accelerates, you’ll need to change your numbers. Even the center distances make a small difference, though that’s more a factor of alignment and tensioning than actual torque. But in my opinion and for my team, I would suggest 15mm just for peace of mind and smaller pulleys. YMMV.

While the GT2/GT3 profile is superior to HTD in almost every way, including service life, backlash, vibration, and load-carrying capacity, HTD belts respond slightly better to reversing shock loads. GT2/GT3 are much better suited for most industrial uses, but reversing shock loads are so common in FRC, what’s best for industry isn’t the same as what’s best for us. The people at AndyMark and Vex knew this and made the informed decision to sell HTD pulleys and belts instead of GT2/GT3 after discussing it with people from Gates and other industry vets.

Just to add on to what Eli is saying about Gates’ stated torque ratings not always being applicable to FRC purposes, remember that all of those values are rounded averages of many tests. Really, those values show a range, not an exact value. A few months ago, I tried to copy all of Gates’ relevant data in to an Excel sheet, convert it to ft-lbs and rpms, and make a calculator that gave you a minimum pulley size for a given gear ratio and wheel size and motor combination. I was curious about just how precise all those values from Gates were, so I emailed them about it. They were pretty helpful, but the end message was that, since I was really trying to read a lot into a very small portion of their tested range, the precision over the range I was looking at wasn’t tight enough to give accurate distinctions between pulley sizes. So, basically, if Gates said a certain pulley size could handle .5 horsepower, the fudge factor in that .5 was the difference between failing and not failing on a specific torque ratio.

So yeah, Gates’ user and design manuals are awesome, but the stated load ratings aren’t super applicable.



If you aren’t direct driving a 6wd center wheel with a transmission, I would not recommended using 9mm wide belts. That is the only drive configuration that would make them manageable. If there are small pulleys driving large high traction wheels, the 9mm wide belts will still give you problems in the 6wd direct drive center wheel.

This is also a function of pulley diameter. I am not familiar with 987’s setup, but from what I remember they use pulleys at least 30 tooth, not tiny like 24 tooth or smaller.

I (and many other teams) have had some problems with 24T pulleys, 9mm belts, direct driving 6wd from the center etc etc. It’s right on the margin of can or can’t work, and it depends on how careful the user is. I don’t like my drives to ride the safety margin that closely.

We’ve used 9mm vexpro belts with 30 or 36 tooth pulleys for 2013-2015 (Rhino last year) with no issues. Never broken a belt.

On the sourcing belt front, you have to remember that the GT2 and GT3 profiles are still patented by Gates, whereas the patent on HTD ran out so there’s many, many more suppliers of HTD.

On load ratings on the design manual, those also might be based on fatigue and not instantaneous loading.

Though it’s really more the name that’s been patented at this point, and not the actual profile, right? Because there are several suppliers, WCP included, that sell a “GT2 compatible” profile. I don’t know where the lines are drawn on how close other profiles can get before being in violation of the patents, but I’d say the “GT2 compatible” profiles are pretty close.

And yeah, none of Gates’ load ratings are off of instantaneous loading or reversing loads.

I believe patent on gt2 belts are up but gt3 belts are still protected (source: gates rep)

My biggest question when it comes to belts on drive teains is what am I gaining buy using two 15mm belt vs any chain. Weight savings is all I can find. I worry that the larger space taken up might not be worth it. That said belts are my go to choice for everything else, 3D printed pulleys anyone? :stuck_out_tongue:

Unfortunately, when you sum up the extra width of 15mm belts and the weight of pulleys, the theoretical weight savings compared to #25 chain are nulled. :\

Thanks for the compliment.

We might not. Really going on a simplicity kick and gear drives are not simple. As usual everything depends on the game.

I know in the most recent instance I’ve looked at replacing a chain with a belt I was looking to remove slop from the system and would be using GT2 profile belts and pulleys to do so.

Belt offers more advantages over chain than just weight savings. They are slightly more efficient, they are cleaner, they do not stretch, and they do not require maintenance. We have used 24 tooth pulleys with 15mm HDT belt on both 4" and 6" wheels with 3 CIM single speed gearboxes for 4 seasons. We have never broken a belt. If you size your pulley correctly for the wheel and use exact center to center, you should only have to touch your belts one time, I can not say the same for chain.

We ran 9mm wide, 5mm pitch belts in 2014, GT3 with a 3.5" diameter wheel and a 24 tooth pulley. The center wheel was direct driven, with belts to the front and back wheels. We carefully measured the tension, and worked hard to keep them properly tensioned to avoid ratcheting or over-tensioning. They all broke right after as the season ended.

We are going to go to 15 mm wide belts next year unless the game is something weird again. It’s just not worth pushing the limits. If you read the spec sheets, we are well outside the rated range for the belts.

Has any team attempted 24t HTD pulleys with 9mm wide belts this season? I’m playing around with the idea of doing an offseason WCD live axle setup but with pulleys instead of the the typical 16t #25 chain sprockets.