I tried using the iOS app to check the tension on my Gates Carbon Drive belt (centre-track, CDX I assume as it's driving a Rohlof) and I could not get a consistent measurement over the length of the belt. It varied between 35Hz and ~70Hz.

  1. How is this possible?*

  2. Is the belt ruined?

  3. Is this what is causing the belt to make horrible noises and the spokes of my rear wheel to loosen, such that I've had to go around and tighten them three times within the space of a week?

[*] Might it be caused by "rolling the belt on" to the sprocket in direct contradiction to the Gates manual which says "don't do that" (pages 4, 16, 17)?

  • 1
    I assume that this app works by strumming the belt like a guitar string? If so, were you always plucking it at the same point between the chain ring and rear sprocket? Presumably you're supposed to pluck it at the centre; plucking it a third or a quarter of the way will give different notes. Jun 10, 2018 at 12:12
  • Can you post a video of the drivetrain running and include a link to it in your question? Jun 10, 2018 at 12:22
  • Turns out the root cause of all the problems I was having was the shop that built the bike had used the wrong (length?) bottom bracket—the default one that came with the Gates package, rather than the one Rohloff recommended to use with belts. When that got fixed, all my problems went away. Dec 15, 2020 at 12:21

2 Answers 2


Addressing question 1.

I understand that to measure the tension you pluck the belt and app measures the frequency. Anyone who plays guitar will know a string sounds differently when is plucked near the bridge compared to plucked closer to its center. Plucking near the bridge suppresses lower harmonics.

The app instructions should tell you where to pluck the belt, but halfway between the chainring and sprocket is probably the correct spot.

  • I'm pretty convinced I was doing it in the same place every time. And the low reading was in one part of the belt, then I wound it around to another part of the belt to get the high reading. It was pretty consistent. Jun 10, 2018 at 12:19
  • My guess would be that you have stretched a section of the belt when you rolled it on. Jun 10, 2018 at 12:24
  • Plucking at different positions changes the relative amplitude of fundamental and harmonics, but does not change the fundamental frequency itself. Another possibility is that the cogs are not perfectly centered.
    – ojs
    Jun 10, 2018 at 13:31
  • @ojs By which you mean there are more belt teeth on one "side" of the cogs than the other? That makes sense to me too. Jun 10, 2018 at 13:48
  • Not necessarily that bad, just that the mounting holes are not exactly centered on the cog is enough.
    – ojs
    Jun 10, 2018 at 15:13

To work out which component is responsible:

Find the tightest point, then mark the two sprockets and the belt (e.g chalk). Now turn the pedals looking for the next time you get the tightest point.

Is it happening every rotation of the front sprocket, or at a fixed point on the belt, or at the rear sprocket?

Make sure the crank is properly seated and tightened

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.