PDA

View Full Version : Another chuck runout question



Buzzer John
01-22-2009, 08:15 PM
I just read the answer to the runout question for a 3 jaw chuck and it prompted me to ask that same question about a 4 jaw chuck. I have just finished making a backplate for a new 8" 4 jaw Chinese chuck I got from MSC. The spindle thread is close to 1.5 X 10, but is about .050 larger. I had my usual trouble doing the internal thread. At first it is too tight, and after the next small advance of the tool the thread is way loose. However, I have been told that the backplate registers on the spindle collar and that a loose thread is not a problem. Anyway, after getting everything turned, mounted ,etc. I find that when I dial in, close to the chuck, a 12" long 1" dia. piece of drill rod that when I check it 8" away the runout is +/- .020". It looks like the rod is sweeping a cone. This is after finding that the spindle sleeve bearings had about .005 too much clearance and correcting that. This runout is very disappointing. What might have gone wrong, and what might I try to improve
it? Thanks for any help. John

BillH
01-22-2009, 08:24 PM
I just read the answer to the runout question for a 3 jaw chuck and it prompted me to ask that same question about a 4 jaw chuck. I have just finished making a backplate for a new 8" 4 jaw Chinese chuck I got from MSC. The spindle thread is close to 1.5 X 10, but is about .050 larger. I had my usual trouble doing the internal thread. At first it is too tight, and after the next small advance of the tool the thread is way loose. However, I have been told that the backplate registers on the spindle collar and that a loose thread is not a problem. Anyway, after getting everything turned, mounted ,etc. I find that when I dial in, close to the chuck, a 12" long 1" dia. piece of drill rod that when I check it 8" away the runout is +/- .020". It looks like the rod is sweeping a cone. This is after finding that the spindle sleeve bearings had about .005 too much clearance and correcting that. This runout is very disappointing. What might have gone wrong, and what might I try to improve
it? Thanks for any help. John
Verify that your rod is straight first. Systematically work down where the error is coming from. Is the rear and front faces of the adapter parallel?
Theres many variables in play here, very small errors get amplified out on the arm.
Don't face the back of the chuck thinking it is the problem, it is not.

SGW
01-22-2009, 08:32 PM
Figure out what, in fact, the backplate seats against -- what it registers on. With my South Bend, the backplate seats against the face of the spindle, so the rear end of the backplate, the end that goes onto the thread first, has to be absolutely perpendicular to the thread axis or it will cock the chuck.

TGTool
01-22-2009, 08:35 PM
When you got the backplate threaded and mounted on the spindle (checking for a clean spindle without burrs or swarf) did you then face off the backplate before mounting the chuck?

Buzzer John
01-22-2009, 08:39 PM
Bill,Thanks for the fast response. The rod was given to me by an experienced machinist ( I am only experienced in making mistakes) and he said that it is dead straight. As near as I can check, he is correct. As for the parallel faces, I first faced the register side after threading and then took it out of the chuck and mounted it directly on the spindle and faced the chuck side. Checking the face of the backplate I do measure .001 runout. Is that enough to cause my problem? Again thanks, John

SGW
01-22-2009, 09:33 PM
It sounds as though you did things in the correct order...but if you did, I'm not sure why you have the 0.001" runout of the backplate. Why wasn't that eliminated when you faced it off?

BillH
01-22-2009, 10:06 PM
Bill,Thanks for the fast response. The rod was given to me by an experienced machinist ( I am only experienced in making mistakes) and he said that it is dead straight. As near as I can check, he is correct. As for the parallel faces, I first faced the register side after threading and then took it out of the chuck and mounted it directly on the spindle and faced the chuck side. Checking the face of the backplate I do measure .001 runout. Is that enough to cause my problem? Again thanks, John
That .001 runout is being amplified 8" out.
I had to go through the same ordeal on my new lathe which is not bolted down properly or leveled. Which triggered me to ask you if you had done the same? No matter how many light facing cuts I took, I kept winding up with .001 runout in the same area. I did not change speeds to see if it was harmonic. My chuck however is zero settable and I winded up with less than .0005 runout near the collet, so I left it be until a later date for when a project would require better run out. I am going to take a piece of rectangle tubing, fill it with concrete and bolt my lathe to the top of it and level it with feeler gauges and a good level when the time comes. Been playing with the mill too much lately.

Buzzer John
01-23-2009, 08:08 PM
Bill, My lathe is not bolted down, but it has been leveled carefully, and checked several times. It doesn't seem to shift. However, It does have a lot of vibration from a slightly out of balance drive pulley. I have never been able to get that pulley balanced well. I suspect my main problem is that the lathe is probably badly worn. It is a Senica Falls Star lathe, and I have been told it was converted from line shaft operation to an individual motor. I inherited it when my Grandfather died back in 1960. I have tried a facing cut several times to try to get rid of the run out, but the best I got was the .001. That is measured at the outboard rim of the chuck. That is 4" off the center line of the lathe. I would think I would get a linear error. So that at 8" I should see .002, not .020.
Again, thanks for any and all suggestions. John

bj139
01-28-2009, 08:18 AM
Are you measuring parallel to the lathe bed? I guess the term is axial to the spindle not radial. You may want to check both.
Bill

J Tiers
01-28-2009, 08:57 AM
Leveling will have exactly NOTHING to do with chuck runout. If it does, then your headstock is cracked right across.

No matter what crazy angle the spindle is pointing, it still should turn true on its axis.

You have a problem with an amplified error. If you show 0.001 on the backplate, then at 20X that diameter out on the rod you will see 0.020 from that cause.

If you checked the chuck at 3" diameter, and the rod at 8", the 0.001 is not sufficient to explain everything.

Indicate the front of the chuck as mounted. Any quality chuck will have fornt and back dead parallel, and so if you see a runout there, it should be reflected in whatever you hold in the chuck.

Then also, if there is a burr or whatever on a jaw, swarf, dust, anything, THAT will also be amplified as an error.

Take the rod out, put it back and measure again. Do this several times and see if it is consistent.

If it is, then remove the chuck and put it back a couple times, measuring in between.

It isn't hard to make the seating surface too far below the face of the "spigot", so that a burr or fin on the intersection of the surfaces, will hold the chuck askew.

You might skim the seating surface to get it dead-true, and also cut down the spigot length to about 60% or 75% of the depth of the recess in the chuck.