Product Craig's Chinese collet chuck cockeyed

[whydontu]

Don't forget to try the chuck in all three orientations of the spindle nose. On my machine there is a clear winner of which alignment gets the best results, and I've marked the spindle and all of my chucks to make sure they always go on the same way.

 

[Susquatch]

Don't forget to try the chuck in all three orientations of the spindle nose. On my machine there is a clear winner of which alignment gets the best results, and I've marked the spindle and all of my chucks to make sure they always go on the same way.

I was going to suggest the same thing. But then I got to thinking...... (I should never do that at my age). Doesn't he have a threaded spindle? Can they be indexed?

But he could certainly index the chuck on the backing plate. Maybe that's even what you meant.....

 

[whydontu]

 

[PeterT]

Initial results are really good: 1.5 to 2 thous of TIR on various dowel pins.

Maybe you have done this already but before you measure TIR using collet & dowel pins, get rid of the middle men & measure runout of the 5C collet face. A dial test indicator is preferred for this because you can put the probe inside. Example if the collet is out 0.001" & dowel is out 0.001" you could conclude 0.002" in one orientation or 0.000" in another orientation on a zero runout chuck. Now if your chuck is out 0.001".... hopefully you can see how stack up or cancellation can play tricks.

 

[Susquatch]

@whydontu - yes that is what you meant.

I was originally wondering about the backplate indexing.

My lathe has 6 dogs - any one of which can be indexed to the spindle. On top of that, I have (or soon will have) 6 screws to index the chuck to the backplate. Altogether there are 36 positions. That makes it a little more difficult to do, but not impossible.

Typically, my chucks have integral backplates so only six positions to try.

Since the 6 jaw is a scroll chuck, I will also have to try it with different work diameters and attempt to optimize it at the diameter(s) I use most.

 

[trlvn]

Maybe you have done this already but before you measure TIR using collet & dowel pins, get rid of the middle men & measure runout of the 5C collet face. A dial test indicator is preferred for this because you can put the probe inside. Example if the collet is out 0.001" & dowel is out 0.001" you could conclude 0.002" in one orientation or 0.000" in another orientation on a zero runout chuck. Now if your chuck is out 0.001".... hopefully you can see how stack up or cancellation can play tricks.

I started to do that but my DTI was in another setup so I went with the test that was easier to do at the moment. Maybe tomorrow I can get the DTI ready.

Honestly, I still think it is astounding that something so cheap can provide anywhere near this level of accuracy.

Craig

 

[PeterT]

Agree. Imagine buying just the materials locally.

 

[Susquatch]

Honestly, I still think it is astounding that something so cheap can provide anywhere near this level of accuracy.

Indeed! Yours has set a high bar. I'm at least a day away from testing mine.

It is particularly impressive when I consider the difficulty of cutting the scroll to that level of precision.

And yet here I am, like an old fool, thinking about ways to improve it.....

 

[trlvn]

I'd rather be an old fool than a doddering old fool! I like to think the mental exercise of this hobby may stave that off a day to two longer than otherwise.

😉

Craig
(Although there is no scroll in a collet chuck but I know you meant the ring gear and pinions.)

 

[Susquatch]

I'd rather be an old fool than a doddering old fool! I like to think the mental exercise of this hobby may stave that off a day to two longer than otherwise.

😉

Craig
(Although there is no scroll in a collet chuck but I know you meant the ring gear and pinions.)

Actually, I did mean scroll. But my leaky old mind mistakenly thought you were talking 6 jaw like mine. Sorry about that. Too many 6 jaws on the go right now.

For a collet Chuck, I don't think much matters beyond the collet body, taper, and of course the collets themselves.