View Full Version : Actual value for timingbelt tesion.....

22-12-2018, 11:31 PM
I'm familiar with checking the tension by prodding or twisting the belt, but does anyone have a proper figure in Newtons as to what it should typically be please? (for 10mm or 16mm AT5 belt on 24 ish tooth pulleys). I need to calculate springs for an auto-tensioner, and also want to make sure I don't exceed radial load limits on the bearings. Thanks

22-12-2018, 11:32 PM
Manufacturer's data sheets should give this data.

23-12-2018, 12:17 AM
The most accurate method for timing belt tension is resonance, followed by deflection.

I'm pretty sure the Gates belt calc gives tension figures, but I've not used it for a couple of years.

23-12-2018, 08:35 PM
So I went and googled, and eventually found a data sheet with some maths in it and the recommended figures for pre-tension are suprisingly high, something like 1/3 to 1/2 of the full running tension for a 2 pulley system, more if you have multiple pulleys. This means that for a 4Nm Nema24 stepper and 27 tooth AT5 pulleys (which are quite big) you're looking in the region of 60...90N, which is more of a radial force than some steppers of that size will be happy with: and with smaller pulleys things would be worse.

Robin Hewitt
24-12-2018, 02:37 AM
With CNC precision I don't bother with tension, I simply put the wheels to the correct separation and it just works. If using a belt for linear positioning then you should know what tension you require and make sure you have enough teeth engaged to carry it.