View Full Version : Correcting misalignment

Mike Amick
12-09-2012, 04:01 AM
To get more use out of my RT, I mounted a 8" cushman 3 jaw chuck with a
2-1/4" thru hole in it. Mounting was mechanically easy, it had 3 countersunk
bolts to mount it with. I had previously mounted an aluminum plate with a
bunch of threaded holes in it ... so I just bolted it to this.

When I had the chuck aligned .. I am afraid that while tapping the transfer
punch in the first hole I may have moved the chuck just a hair. First clue was
the bolt heads fighting me to go into their countersunk holes.

But the disappointing test came when I chucked up a piece of round aluminum and
nudged it next to an endmill ... and rotated the RT to just skim some off of the
aluminum rod. It took more off of one side than the other. Not a lot ... but ..
enough to see it. So I guess it wasn't concentric .. or .. the runoff was
bad ... or what ever you call it. Obviously it's not sitting in the middle of the RT

So .. how do I nudge the chuck back into the middle. I could ream out the holes
and make the heads slightly smaller do that I actually could bump it a little. I don't
think that sending a cutter down into the jaws would work. I sure would hate
to start from scratch again.


12-09-2012, 04:44 AM
You need some radial slack between the hold-down bolts and the chuck body. So you either drill the chuck body holes and/or counter-sink oversize or reduce the hold-down bolt and/or bolt head diameter. You need to do some measuements/investigation to see where the constraint against radial adjustment actual is.


12-09-2012, 07:18 AM
Once that is done, fit up 3 or 4 blocks with jacking screws and get out your test indicator. Three jaw chucks are not terribly accurate so if you indicate it with a 1/2" dowel it might be off with a 1" dowel. I can't remember the last time I had a 3 jaw chuck on my lathes as it takes so little time to indicate a part in a 4 jaw chuck.


12-09-2012, 09:05 AM
So you have a milling machine and a rotary table and you used transfer punches? Sort of like surgery by the Braille method .

12-09-2012, 09:18 AM
Your rotary table should have a morse taper hole in the center. Use that as an alignment pin for the chuck.

I don't like those pivioting jaws on that chuck !!!
That could cuse some serious error.


Lew Hartswick
12-09-2012, 10:01 AM
So you have a milling machine and a rotary table and you used transfer punches? Sort of like surgery by the Braille method .
:-) Very good. :-) I was about to post a similar idea before I read that one.
My first thought was " Transfer punches for a precision location????" :-)
But as has been pointed out 3 jaw runout would negate most of the error

12-09-2012, 11:24 AM

Do you have a digital readout on your mill? If so, just indicate the center of the rotary table and then use the digital to put the tapped holes in a new location ie: rotate the RT and drill new holes. Also, the holes in the chuck should be clearance holes and therefore you should be able to snug up the bolts and then tap the chuck into place before final tightening.

Also, the term you are looking for is "RUNOUT" not runoff. Think running out of round.

Good Luck,

edit: forgot to mention that it looks like you have your end mill sticking out of the collet too far for your setup. Use as much of the shank as you can for added rigidity.

12-09-2012, 12:01 PM
:-) I was about to post a similar idea before I read that one.
My first thought was " Transfer punches for a precision location????" :-)

I disagree , transfer punches provide more than enough accuracy to locate holes for bolts.

Mike the problem you have is that the bolts and their holes are not an accurate way to locate things. period. Whenever you need to mate two parts accurately you need to use a register, shoulder, pin etc. ie make your adapter plate with a spigot that fits the centre bore (mines a bore, some are tapered) on one side and a shoulder on the other that fits the inside of the back of the chuck. That is what positions the chuck, the bolts just hold in down. Or you can just indicate the thing in each time

as noted 3 jaws will have some eccentricity, a function of the scroll. How much runout are you getting?

Jaakko Fagerlund
12-09-2012, 12:28 PM
Do NOT use bolts to locate stuff and especially countersink bolts, as they try to force the threaded object to move to where the countersink is. Use SHCS's and oversized holes to get it where you need it.

12-09-2012, 01:18 PM
This is how I locate my rotary table to the mills spindle dead center. I have 1 1/8" dia. x about 2" long piece of ground shafting that I drilled and reamed to .500. I put a .500 dowel pin in the drill chuck or collet depending on what I'm mounting in the spindle and just position the chuck under the pin and move the x & y until the pin slips into the hole. If I'm off a half a thou. I'll feel it bump the side of the hole. I've found this to be a very acurate way to zero in the chuck, quick and simple. The other end of ground shaft you see in the chuck is drilled and reamed to .250 in case I'm using a drill chuck that won't accept 1/2".



12-09-2012, 01:53 PM
this is why I wanted a 4 jaw for my rotary table. Sure, I'll have to dial it in every time, but at least it will be accurate every time, and dialing it in does not involve loosing several mounting bolts and hiting the rotary table with a hammer in hopes to make a 80lb rotary table jump 0.001"

That, And my rotary table has 4 slots and 4 (oversized) bolt holes through a 3 jaw chuck looks funny...

The Artful Bodger
12-09-2012, 02:02 PM
I made an arbor that fits the taper in the rotary table and drop the chuck over that then clamp the chuck down after closing it (not tight) on the arbor. Presumably the chuck is then centred with moderate accuracy especially at the diameter of the arbor.

Bob Fisher
12-09-2012, 05:05 PM
Do you know that your 3 jaw chuck is that accurate to start with? Especially at that dia and with a piece of aluminum? Hardly as good as a dowel pin or other precision rod. Just how bad was it? A dial indicator would tell the story. Did you expect to get a zero runout under these conditions? Bob.

12-09-2012, 05:15 PM
There is a simple method.

Chuck up a piece of stock in the mill spindle, say a 1/2" piece of drill rod. Then, tighten your chuck onto the protruding part of the stock so that it's dangling in mid air.

Lower the quill gently till the chuck contacts the rotary table, which you already centered using a dial indicator on it's bore. Now bolt the chuck down.


The Artful Bodger
12-09-2012, 05:32 PM
Um.... that might get the chuck centred on the mill spindle but it wont do much to align it to the rotary table.

Mike Amick
12-09-2012, 08:15 PM
hmm ... you guys that went to school .. or apprenticeshiped .. I envy you. Trying to teach
yourself this stuff without even a big brother is a bitch. Got it .. bolts are not good enough all
by themselves for pin point accuracy. I can actually use the practice as far as dialing stuff in so
I think I will give myself some wiggle room ... and just dial it in every time. Thanks so much for
all the great ideas.


12-10-2012, 01:35 AM
I think I'd start by turning up a test piece on the lathe. It would fit into the center hole in the rt, and then you would indicate against the part sticking out. If that's good, then set the chuck over it with something round already chucked in the jaws. Then you're going to find out where it's going wrong. If all is good at this point, then it's the bolts pulling the chuck to one side.

If you're still good with two of the bolts in place, then possibly you could enlarge the third bolt hole only- just enough to allow the bolt to thread in easily. You could enlarge the hole for only about 2/3 of its length, from the back of the chuck, leaving the bolt head seating surface and the first part of the hole intact. That could work, depending on how far off it is.

I agree- you should have some kind of locating pins installed, or some means of accurately locating the chuck as the bolts are tightened.

Maybe once you get the chuck well centered, you could make a pair of spacers that would bolt to the aluminum and nest against the side of the chuck body. Once located, these would have pin holes drilled through and into the aluminum plate. These are then bolted into place temporarily while you mount the chuck, then are either left there or removed.

Come to think of it, why not make three locating spacers to center the chuck, and leave them in place. I don't see where they'd be in the way of anything, and they would help to ensure that the body of the chuck doesn't move sideways. Each would be marked as to where they go, and with two pins each they are going to be spot-on every time they're mounted.

12-10-2012, 03:59 AM
Looking back over this thread I see several problems. One, a 3 jaw chuck is for round , triangular, or hexagonal materials. Your chuck with it's swiveling jaws belongs in the Museum of dumb ideas for fixing problems that don't exist.
You refer to countersunk holes. None are visible nor should they be on a chuck' There is a vast difference between countersunk and counterbored holes and you should learn this before any further screw ups.
Threaded fasteners are for restraining a part, not for locating it. Properly applied, they should not affect location.
Transfer punches are slightly above blacksmith grade work. You have a mill and a rotary table ( I assume that is what "RT" means. Pick up the center, move off in one axis the bolt circle radius, and drill a clearance hole every 120 degrees. That's how you do it. To locate you need a pilot or rabbet fit. This is what determines location, not screws or bolts.
There is a difference between a screw and a bolt. Learn that too.
If you take off that aluminum piece you will gain it's thickness in your Z axis and you might find the holes you need.

12-10-2012, 09:53 PM
several things:
What does the back of the chuck have? The fixture plate is not really going to be your friend but, if you machine the plate to the back of the chuck and bore a hole and pin it to the table you might be reasonably close. That assumes the table rotates concentric, have you checked that?

The chuck looks to be old and the jaws... well, questionable is the best I can say. You might rework them and make inserts that you could bore out each time you wanted to use the chuck. But I tell you now that is a very very bad idea.

Keep in mind you have the concentricity of the rotary table and the repeat ability of the chuck all working against you.

It would be possible to use undersized bolts to hold the chuck and then "dial" the part in the chuck in. But again you would have to dial the part in, rotate the chuck and check it again.