Collet nut tight enough? what quality collets? Chuck jaw faces could be tilting the block slightly away from chuck face

Edited By JasonB on 06/09/2015 10:08:33

Have you checked the roundness of the test bar,centre less ground bar can be ground to a lobe shape,check with a micrometer and it shows a constant diameter ,place the bar in a vee block set, up a dial gauge and then rotate the bar, if it is lobed the dial gauge will indicate the error,if it is ground true the dial gauge should be constant. It is caused by the centreless grinder not being set up correctly. You show clock readings to three decimal places ,have you read the gauge correctly,if so then you are expecting an awful lot of quality from a collet block. I have a hex er32 collet block ,have found it the ideal method to make whit thread studs for stationary engines,the er collets have fantastic grip and allow short studs with only half an inch of central plain steel (EN8) to be threaded each end without the bar slipping ,The block is held in the three jaw and the die in a sliding die holder, the er collet nut also allows the die to go right up to the collet face, I would not use my lathe collets in this way as I was taught that ordinary spring collets (5c and similar)should only be tightened just sufficiently to hold the work for turning. The ers can be really pulled tight and it does not matter if one does get damaged they are quite cheap, The collet block can also be held in the bench vice to grip round rod when it needs to be threaded using a die holder,far better than vice jaws and no marking.Never checked the accuracy of the block as I only use it for studs where concentricity dos not matter.Its solved the grip problem and has made stud making easy and quick,previously it was a pain in the backside,best tool I have bought for ages.

There have often been articles in ME about truing 3 jaw chuck jaws. Normally for concentricity but a side effect is parallelism. Same techniques should apply to 4 jaw.