The main thing I found with a cub is that the plates sometimes 'dish' this then acts as a spring preventing clean release and also limits the friction area. You must ensure all the plates are completely flat.

I also found that cutting the clutch pushrod in half, polishing and case hardening the ends and putting a ball bearing in the middle made things smoother and lighter. Can't explain why, but it did. The idea was developed from somebody who did a top hat roller bearing for the pressure plate on BSA A50s /A65s, to reduce friction. SRM?

I also found that cutting the clutch pushrod in half, polishing and case hardening the ends and putting a ball bearing in the middle made things smoother and lighter. Can't explain why, but it did. The idea was developed from somebody who did a top hat roller bearing for the pressure plate on BSA A50s /A65s, to reduce friction. SRM?

It's because a long pushrod will tend to bow when under load, rubbing against the sides of the tube it is in and increasing friction. Two shorter rods are much stiffer so won't bow as much and won't rub as much, though the ball in the middle will have to rub it doesn't seem to generate as much friction.

The mathematical formula still used to determine bow in rods under compression was developed during the 18th century by Leonhard Euler. Don't think he had a BSA though.