Lathe chuck not parallel

[N4NV]

I have a Buck 3662 6 jaw chuck. I have set the chuck so it is centered within .0005". The problem is when I clamp a precision ground bar in the chuck, it is not parallel to the z axis. If I measure the bar 6" from the face of the chuck, it is .006" off center as it rotates. I called Buck and they said it was time for another $1200 chuck. I would like to know it there was a way to regrind the jaws so they were parallel. I thought I could grip a ring in the back of the jaws, then put a round grinding bit (the type for a Dremel tool) in the tool holder, run the chuck and very carefully run the grinding bit back and forth with the carriage. Any suggestions?

Thanks

Vince

[cam1]

Hi:
If the chuck jaws are worn this will work, if the scroll is pooched then you will not be able to correct the problem. It's the same as a worn lead screw on a machine tool.

regards

[Oldmanandhistoy]

Any chance its the head stock out of alinement?

John

[Geof]

Have you taken the chuck off and thoroughly cleaned the spindle nose, including any adapter plate the chuck might have by taking it off. Make sure there are no chips or anything in there and make sure when everything is re-assembled things go together nice and straight.

Then when you true the chuck do not true the body; put your precision ground bar in the jaws and true up to that. You should be able to get it true both near the chuck and also a few inches away.

It is only after doing this that you can safely conclude that the jaws are not accurate and if this is what you do determine then you can consider grinding them.

However, putting a spacer at the back of the jaws is not really suitable. You need to load the jaws in the same manner they are loaded in use which means gripping something mnear the front. This can be tricky if not impossible but if these are two piece reversible jaws you can make a ring that is gripped by some small bits of round bar in the counterbore for the jaw bolts to load the jaws near the front.

[Konrad]

Sounds like a worn chuck.
Yes you can grind, as you said...clamp a ring in the very back of jaws then grind or machine with ceramic.
If you clamp pieces larger or smaller then ring used and you still have run-out problem, this indicates a worn scroll.

Konrad

[N4NV]

Originally Posted by Oldmanandhistoy Any chance its the head stock out of alinement? John

Anything is possible, but all my collets point parallel to the bed, i.e. if I put the same precision ground bar in a collet, I do not get the run out. Yes I have taken the chuck apart. I put just the adapter on my lathe and made a skim cut. I cleaned the entire chuck and reassembled it, put it back on the adapter (still on the lathe) put a precision bar in it and set it true to the bar next to the jaws. 6" from the chuck the bar runs +-.009" out of round. If I go ahead and grind it I will use the method here:

http://www.homemetalshopclub.org/new...04.html#truing

The nice part of that is the pencil grinder will fit in my boring bar holder.

Vince

[N4NV]

I would like to thank JST for the information regarding grinding my chuck. I followed the instructions about mounting a pencil die grinder to do the grinding. I cut 6 identical pieces of Delrin and put them between the jaws and tightened them. I did not need a ring. I ordered a cheap pencil die grinder (56,000 rpm) from eBay $25 plus shipping. The die grinder was 5/8" OD which fit perfectly into my turret tool holder on my CHNC. I ran the grinder back and forth about 20 times an removed about .0015" total. I installed the test bar again and get +-.00015" at the chuck face and +-.00015" 6" from the face with two different test bars. See the attached picture for the set up. Now that I have seen how well this works, I will not hesitate to do it again if I ever get another chuck with this problem.

Vince

[cam1]

Yes, but try installing a larger diameter test bar to validate the accuracy.

regards

[N4NV]

Originally Posted by cam1 Yes, but try installing a larger diameter test bar to validate the accuracy. regards

I installed two different test bars of different diameters and get the same results.

Vince