4 Steps to Avoid Motorcycle Final Drive Belt Drama: Secrets from Experience

5 Steps to Avoid Motorcycle Drive Belt Drama: Secrets from Experience

Drive belts do have less maintenance and advantages for motorcycles over a drive chain, but they still can cause drama. What secrets can old belts share?

A while back, the final drive belt on my 1999 Harley-Davidson XLH 883 Sportster failed in my garage. If a bike’s drive belt is going to break, that’s the place to have it happen. We had a look at that breakage event back in May 2018.

I have since learned that the kind of failure that belt had—total failure in tension (tensile failure mode), snapping right across its width—is actually not that common.

Rather, a more common failure mode is where the drive cogs around the inside face of the belt begin to peel off in chunks known as tooth shear.

The belt from the 1999 XLH (R) shows shiny glazed surfaces on both sides of the belt that is not present on the belt from the 2007 bike.
The belt from the 1999 XLH (R) shows shiny glazed surfaces on both sides of the belt that are not present on the belt from the 2007 bike.

The benefit of having that kind of belt failure is that by tightening the drive belt, you may be able to avoid being left on the side of the road. I learned that from an old hand at handling drive-belt equipped Harleys on a group ride a few years back.

Because he was slowing down consistently over a few miles, I figured there was trouble. The rider then pulled off the road. Sure enough, the drive belt had lost some of its drive cogs and only the belt backing was left on those spots. After he tightened up the belt, we were able to slowly make it to the end of the ride.

I kept that failed belt from my ’99 Sporty in hopes of learning something about how a belt should be maintained over the long haul and also be able to perhaps to more detailed drill down on how modern drive belts fail. The belt that failed was, to the best I can determine, the original equipment drive belt. That means the belt lasted 35,366 miles over 19 years.

Fast forward to the fall of 2019 and I’m having a new rear tire and rear brake pads installed on my 2007 H-D 883R Sportster. Because the drive belt on that bike was also original equipment and had reached 12 years of age and 26,145 miles, I decided to have the drive belt changed at the same time and keep that old belt to see what tales it might tell.

Frankly, I didn’t expect to see anything notable at all. As usual, I was wrong. In comparing the remains of the belt from the 1999 XLH and the old belt from the 2007 883R, I noticed something I hadn’t before.

The clearance between the new drive belt and sides of the pulley is visible in this view of the 2007 Sportster.
The clearance between the new drive belt and sides of the pulley is visible in this view of the 2007 Sportster.

The belt from the 1999 Sporty was glazed to a bright shine on both sides of the vertical surfaces of the belt. The vertical surfaces of the belt from the 2007 model had no such glazing. The glazing seemed to be the kind of thing I have seen over the years on snowmobile drive belts or old-fashioned automotive V-belts that have worn to the point of doing a lot of slipping and developing a lot of heat as a result.

The thing is, in the type of internally toothed (Gilmer-style) drive belt used in a motorcycle final drive, slipping in the pulleys isn’t possible unless there are teeth missing or the belt is so loose, it can jump the drive cogs in the pulleys—neither of which is the case in this belt. And, the belt isn’t intended to ride against the sides of the pulley, unlike a V-belt.

That leaves two other possible causes—either the original belt was actually too wide for the specification width of the drive and driven pulleys or there was a misalignment problem. While the first scenario is possible, I regard it as remote, so misalignment is a probable cause, unless a belt change did happen in the first 6000 or so miles of the bike’s life, which I think is unlikely, but possible.

I took my trusty Vernier caliper to check some dimensions.

Model Year Old belt width (in.) New belt width (in.} Driven pulley width (in.)
1999 1.122 1.093 1.291
2007 1.000 0.950 1.099


Not knowing what the original equipment belt width was on the ’99 model before any side surface wear, it is nevertheless clear that the new replacement belt for the 1999 model is narrower than the OEM belt was when new.

OK, I know 0.029 in. is not a big deal at first blush, but think about it this way: it is about as wide as the spark plug electrode gap in your bike.

This simple device is available from Harley-Davidson for testing drive belt tension.
This simple device is available from Harley-Davidson for testing drive belt tension.

On the other hand, the apparent difference between the inside width of the driven pulley and the original belt is 0.169 in. That would seem to rule out the belt having been too wide from day one. So, again, misalignment would seem to be the cause of the glazing.

According to Acorn Industrial Services, the most common cause of Gilmer belt failure is a misalignment of the drive and driven pulleys—Acorn says: “Excessive or uneven tooth wear, belt tracking, and tensile failure can all be attributed to misalignment.”

Excessive load and shock loading were next among the most common causes of drive belt failure. In the case of a motorcycle final drive, this correlates to a bike loaded to the max with cargo and passengers and/or lots of throttle delivered with a dump of the clutch.

An under-tensioned belt—simply too loose—can cause the teeth to skip in the pulleys resulting in rapid or uneven belt wear, drive-line noise and ultimately, premature belt failure.

Other leading causes include weakness in the drive structure—such as worn bearings on drive or driven ends—damaged or worn pulleys or debris in the pulley surfaces or in/on the belt itself.

Another great source of intel on how Gilmer-type drive belts can fail and how to prevent it is available on the Gates website.

Take any of these or any combination of these, throw in plenty of miles and you have the perfect combination of drive belt failure factors.

All things considered, given the glazing on the sides of the 1999 belt showing evidence of apparent friction and heat from possible misalignment, it’s pretty amazing the belt lasted as many years as it did. But since some authorities cite an average service mileage life of about 50,000 miles for this type of belt in normal use, with tensile failure occurring at 35,366 miles, it may be that there was a price paid for that problem.

In my estimation, the misalignment of the pulleys right from the factory is very unlikely. Since the ’99 Harley has had several rear tire changes over the life of the belt, where the belt was not changed, misalignment could have happened at any of them. So exactly where some degree of misalignment may have happened we’ll never know. Work on that part of the bike has been done by official Harley-Davidson dealers as well as experienced professional independent cycle shops.

The belt on the ’99 Sportster now has 4,747 miles on it and the just-changed belt on the 2007 883R has only 70 miles on it and probably won’t get many miles between now and the spring thaw. As I write this, neither of the new drive belts has shown any of the glazings on the side surfaces that are present on the failed belt off the ’99 model.

The take-away from all this are some steps that can help avoid drive belt drama.

5 Steps to Avoid Motorcycle Drive Belt Drama: Secrets from Experience

Prevent Motorcycle Drive Belt Replacement: 4 Steps

1. Keep the belt and pulleys clean

2. Keep the belt tension set per your owner’s manual specs (Harley-Davidson has a tension testing device available)

3. I plan to feel the drive belt with my bare hand after riding thirty or forty miles. Even on a hot day, when I’ve felt the belt after riding it can be pretty warm, but If it is almost too hot to touch, I’m going to get it into a shop to have the pulley alignment checked.

4. Also, I’m going to monitor the condition of the side faces of the belt for glazing—if there is any, I’m going to have the alignment checked. These steps are no guarantee of 50,000-mile service life from any belt, but they may help.