PDA

View Full Version : Dividing head accuracy problem



gr8life
12-26-2010, 11:15 AM
I have a dividing head of unknown make. I have used it for years and it has worked for the jobs I have given it. Now I need to do some work that will require much more accuracy.
With a 5C chuck mounted and a 1/2" ground pin in the chuck I am getting .005 out of round at 4" away from the chuck, this is after I cut a new backplate for the chuck.
I am thinking of buying a Bison set thru and a new backplate but would like to have some input before I plunk down about $600.
I hope after you finish playing w/ your new Christmas toys you can offer some advice.
thanks
ed

DeereGuy
12-26-2010, 11:18 AM
Have you check the dividing head runout yet? I only ask because I ran across this with mine and I had to adjust the bearing nut.

philbur
12-26-2010, 11:26 AM
Depends on how you cut the backplate.

There are multiply possiblities for the source of the error, why do you assume a set true chuck will fix it.

What is the runout at the collet.

Pnil:)

PS: A half arse question will get you half arse answers - give more details of the dividing head and what you have checked.


I have a dividing head of unknown make. I have used it for years and it has worked for the jobs I have given it. Now I need to do some work that will require much more accuracy.
With a 5C chuck mounted and a 1/2" ground pin in the chuck I am getting .005 out of round at 4" away from the chuck, this is after I cut a new backplate for the chuck.
I am thinking of buying a Bison set thru and a new backplate but would like to have some input before I plunk down about $600.
I hope after you finish playing w/ your new Christmas toys you can offer some advice.
thanks
ed

J Tiers
12-26-2010, 11:31 AM
A 5C does not grip the part by a very long jaw.... it does not take much of a piece of swarf, dust, gnat hair, etc to throw it off that much. Plus, the short grip means not much stiffness against movement of the part sideways out at the end.

What would you hold in a 5C and expect to be solid out 4" from it?

That's not really what they are 'for'.

if you need to hold something 5C sized, which is no more than 1" diameter or so, solidly, it makes much more sense to use the dividing head's tailstock to assist in holding.

if the part is less than 1" diameter, that is even more important. While teh grip is proportionately longer relative to diameter, the part is more flexible also.

if you apply any repeating force to points around the diameter at the end of long part in the 5C, I'd expect to see it "walk out" of the grip somewhat. Smaller diameter parts not so much.

John Stevenson
12-26-2010, 11:47 AM
The normal method of making a backplate is to turn it on the spindle that it's going on.
With a dividing head unless you can spin it at say 300 rpm this isn't possible :p

So chances are the machine you made the backplate on isn't concentric to the DH spindle.

Reduce the register a couple of thou and then true the 5C chuck up to the setup bar you are using.

If you have no runout close to the collet but run out 4" out it's a problem with the face not being true and not centralisation.

gr8life
12-26-2010, 12:30 PM
Deere guy the bearing is adjusted
philbur I cut the backplate on the head on the mill, with no chuck the backplate reads 0 thru 360 deg. rotation.
J Tiers I have the tailstock on & centered
John S. see above.
I now think the answer is to farm the job out.
thanks
ed

Paul Alciatore
12-26-2010, 12:40 PM
The first thing to check is the runout next to the collet. If it is off there, then the backplate is probably at fault.


The normal method of making a backplate is to turn it on the spindle that it's going on.
With a dividing head unless you can spin it at say 300 rpm this isn't possible .....

I disagree. You can rough out the backplate on the lathe and then mount it on the dividing head and make the finishing cuts on the mill using a milling cutter while rotating the head. I wouldn't recommend making any heavy cuts this way unless you know the head (the worm and worn wheel) is up to it, but just a few thousanths should work OK. Do both the mating FACE as well as any locating shoulders this way.

I am assuming that it is a screw on design, like many of the older lathes. If it is fastened by some other means that would allow mounting it in several positions, then make witness marks to allow it to be assembled the same way each time.

oldtiffie
12-26-2010, 06:52 PM
I suspect that the OP has a "conical" eccentricity error which can be fixed by shimming the 5C adaptor to the rotary table.

The 5C adaptor should be able be clamped directly to the rotary table.

Here is my similar ER-32 adaptor (which I also use on my lathe).

http://i200.photobucket.com/albums/aa294/oldtiffie/HF-45%20Mill%20misc/HF45-4-1.jpg

philbur
12-26-2010, 07:18 PM
That setup doesn't look to rigid.

Phil:)



http://i200.photobucket.com/albums/aa294/oldtiffie/HF-45%20Mill%20misc/HF45-4-1.jpg

oldtiffie
12-26-2010, 08:34 PM
It works OK Phil.

'specially when I just stick the end mill up the MT taper sans adaptor.

lane
12-26-2010, 09:08 PM
IN all reality .004 out 4inchs out from the face of a collet is not any thing to worry about . Put your indicator out on the end of the part and tap it true with a soft hammer. Ain`t know body going to stick any thing in a collet with out tapping the end true.Its as simple as that. Next I have not seen very many tail stocks for dividing heads that would not pull a part off center after every thing is lined up. .004 is only .002 out center of rotation normally not enough to worry about.

J Tiers
12-26-2010, 09:49 PM
Next I have not seen very many tail stocks for dividing heads that would not pull a part off center after every thing is lined up. .004 is only .002 out center of rotation normally not enough to worry about.

Yabbut, he ought to be able to turn the part with the DH with end in T/S and NOT have it wabble 0.004, unless the center drilling is "out".

Seems like the error in the T/S ought to be a constant, just out of position, not wabbling. If it wabbles, seems like the drilling is "out" or the T/S is not all the way into the drilled center, so it isn't guiding.

If it was center drilled in a scroll chuck, and then loaded into the chuck later, well maybe that would account for the center being out.