PDA

View Full Version : Desperate for help with Chuck



BillH
02-25-2004, 10:12 PM
I've been working on that damn chuck all day, I only stopped when I noticed my blood smeared all over it. Funny how you dont notice your injured until you see blood sometimes.
I just don't get it, the backplate on the spindle, with dial indicator, reads less than .001 runout, probably .0003. That is after many facing cuts.
My only thought right now is that the other face, in the middle is too high and hitting the back of the chuck, as in one area, I cant get the chuck to sit firmly. Im going to face that face one more time, but each time I do that, Im losing precious shoulder for registering the chuck. If that does not fix it, then the back of the chuck must be off, I even faced that part. Now, if this does not work, Im lost, I probably will need a new chuck back plate.

I cant get rid of the runout, this runout is wobble in the chuck along the y axis, the runout forms a cone farther away from the chuck.
Another thought is that perhaps my spindle has too much play and I need to adjust the shims on it, but is not noticible with just the backplate on, but with the weight of the chuck on the end, it causes problems.

Mike Burdick
02-25-2004, 10:29 PM
Bill,

Could you machine the plate (or collar - not sure what your have) while it's on the spindle and avoid the chuck altogether? This way it would be the most accurate for your lathe.

If so, just face it within 0.050" (or less if possible) of the spindle threads then take it off and machine the remaining material in a chuck - undercutting slightly so it will not play a part in the fit.

Just a thought.....

Mike


Oh, when facing make sure your carriage is locked to the ways and the compound's gibs are fairly tight. If you can, set the compound at an angle so that the pressure is pushing into the dovetail rather than along it. This will make for a more rigid setup and help avoid the "dish" you mentioned.



[This message has been edited by Mike Burdick (edited 02-25-2004).]

franco
02-26-2004, 03:22 AM
Bill

Another thing which will cause increasing runout as the distance from the chuck increases is one or more chuck jaws which are not parallel to the lathe axis. This can be due to wear or a manufacturing fault. I have a (nameless) 4 jaw chuck which I bought new, which would never hold work true, although the chuck circumference and front face were reasonably true when mounted on the spindle. After weeks of trying everything, it turned out that the jaw guide slots were not parallel to the face of the chuck! However, opposite pairs were parallel to each other, but in different planes. I ground the jaws so they were square with the chuck face and it now holds work reasonably true. Still not much of a chuck though.

franco

Carl
02-26-2004, 03:56 AM
I agree with franco, I'd suspect the jaws may be the problem.

[This message has been edited by Carl (edited 02-26-2004).]

BillH
02-26-2004, 10:40 AM
Ok, I will look into it then give Enco a call, this is a Bison Chuck,I wouldn't expect this kind of stuff from them.

RD51
02-26-2004, 11:24 AM
What type of material are you using for the back plate?If using anything other than cast iron that could be the problem depending of course on the diameter of the back plate.

Del
02-26-2004, 11:31 AM
Bill,
A friend of mine bought a Bison Chuck from Enco and it had a runout problem as well. It was parallell with the lathe bed but wasn't centered. The runout on the back plate was fine but the chuck was offset. He ended up undercutting the backplate and moving the chuck over until the runout was good. Turned out really good but I would have sent it back. This shouldn't happen with a Bison.
Del

Carl
02-26-2004, 12:24 PM
One way you might consider is making an arbor between centers that will pass through the chuck and mount the chuck back side facing out using the jaws to clamp the chuck to the arbor. Then you can face the mounting surface perpendicular to the holding axis of the jaws.

franco
02-26-2004, 01:34 PM
Carl,
I tried that in desperation when trying to get my faulty 4 jaw chuck to hold work true. Not a good idea! When the chuck was remounted on the back plate, the face and circumference of the chuck were seriously out of true, and work held in the jaws was, if anything, further out than before. It only took a few seconds to convince me to remove the chuck from the backplate and true up the back of the chuck with the face and circumference again. DON'T go that way!

Del,
If the chuck jaws are square with the lathe spindle but offset from the chuck center, am I right in thinking that the runout should remain constant no matter how far out from the chuck? Bill says the runout gets worse as you move away from the chuck.

franco



[This message has been edited by franco (edited 02-26-2004).]

franco
02-26-2004, 03:45 PM
Bill,

Have you tried the second set of jaws in your three jaw chuck? If so, is the runout similar, and does it increase as the distance from the chuck increases in the same way it does with the set you are using? It is most unlikely that two sets of jaws would be faulty; if the runout does not increase with the second set as you move out from the chuck, one or more of the jaws in the first set is probably faulty; if the runout increases by a similar amount with both sets as you move away from the chuck, the problem is unlikely to be with the jaws.

franco

Carl
02-26-2004, 04:20 PM
BillH, are you sure you're not expecting too much from that chuck, I went back and read you're original post where you stated the run out was .002 near the jaws and .005 2 or 3 inches out from the jaws. As long as the chuck will hold work firmly, and you can cut all diameters on the work without removing the work from the chuck the work will be accurate. If you can't turn all diameters in one chucking, or you need to indicate a piece true, you should be using a four jaw or turning between centers.

BillH
02-26-2004, 04:51 PM
This Bison chuck is one that came with its own Threaded backplate, and 2 piece hardened jaws. Problem with it originally was that the collar was too long, and the short spindle of my South Bend 9c, not too many teeth could make contact with the spindle. I made a spacer which I thought was was very flat, took off the backplate, and faced the rear of the backplate until all the threads could be used on my spindle. Reasambly, and poof, out of whack. The chuck and backplate have tick marks on them, so I made sure that they were ligned up when I redid it.
IT went poof because my spacer collar was NOT flat, but had .008 wobble on it, since I made it with the bison chuck before I decided to face down the back plate. Hmm, I guess that tells me that the CHuck was out of whack since I first got it!

[This message has been edited by BillH (edited 02-26-2004).]

BillH
02-26-2004, 06:06 PM
One other thing I did, for kicks, was that I used the 3 jack screw bolts for removing the backplate, as adjusters to Get RID of the runout. However, I noticed, it only got rid of the runout at the specific point where the dial indicator was near the chuck, It did really funky stuff towards the end of the reamer, run out there was crazy, almost like it was not linear compared to where the dial indicator was. Is this a sign that goes to suggest the theory that the chuck itself is defective?

Allan Waterfall
02-26-2004, 06:13 PM
BillH,
Just a thought and I might be totally wrong,but have you put a piece of 1" bar about 6" or so long in the chuck and turned it into a slight dumbell shape i.e. waisted along the middle,leaving a collar on each end.
Take a light cleaning up cut off both collars and then check their dimensions against each other. If they are the same I would think all is well.If not I think you either have a problem or your lathe bed is twisted and might need a shim under the front or back legs at the tailstock end.
I hope this is of some use as it's all I could think of.It will need checking like that anyway even if it is the chuck that proves to be at fault

Regards Allan

BillH
02-26-2004, 06:20 PM
Allan, I just stick one of my reamers into the chuck and turn on the lathe, the thing is wobbling very badly. Human eye can detect movement as little as .030? This thing is definetly exceeding that.

Allan Waterfall
02-26-2004, 06:29 PM
BillH,

Have you tried it with a piece of silver steel or drill rod as I think you call it ?

Allan

Carl
02-26-2004, 07:17 PM
The test Allan is referring to is to test the alignment between the spindle and the ways, it will tell you nothing about the chuck, as soon as your cut cleans up the test bar it will be turning true to the spindle axis of rotation. Any taper will indicate misalignment between the spindle axis and ways, but it will tell you nothing about the chuck.

franco
02-26-2004, 07:27 PM
Hi Bill,

It certainly sounds as though the chuck is not square with the spindle from what you say above about using the jacking screws. Check this by mounting a dial indicator in the toolpost and run it along the side of the stationary chuck in several different positions. There should be no movement of the needle in the width of the chuck.

Am I right in thinking that the spacer you made is between the backplate and the chuck, and is somewhere near the same diameter as the chuck? If so, is it possible to attach it permanently to the backplate using say three short countersunk screws from the back of the backplate? You could then screw the backplate onto the spindle and using light cuts, skim it flat. This would ensure it is square with the spindle. Then bolt the chuck onto the backplate, preferably without removing the backplate/spacer assembly from the spindle, and without using the jack screws for levelling. Hopefully the chuck will then run true, and the dial indicator needle will show negligible movement either along the side of the chuck, around the circumference of the chuck, or across the face of the chuck. In the unlikely event that the chuck is still not true, then it would be reasonable to suspect a fault in the manufacture of the chuck itself. While you have the chuck off the backplate, check that the back and front faces of the chuck are both square with the circumference "just in case".

Regards, franco

Carl
02-26-2004, 07:40 PM
If I'm not mistaken Bill said he has faced the back of the chuck body to. If you have even slight run out in all the different surfaces you have faced, they are probably all adding up to one big run out.

wierdscience
02-26-2004, 09:13 PM
You could try taking the chuck all by itself and gently chucking it onto the spindle itself,then turn the spindle by hand and see what the register and the back plate look like.

BillH
02-26-2004, 09:44 PM
I kinda did that, I left the backplate on the spindle, and Clamped the chuck to the backplate, on the back of the chuck, it read a .005 runout. I then faced the back of the chuck

wierdscience
02-26-2004, 10:23 PM
I hate to ask,but did you check out the spindle?

BillH
02-26-2004, 11:08 PM
Oh wierd science, in class today, I was pondering that very thought all day, it was driving me crazy... Oh no what if my spindle is bent?...
I checked it, its not http://bbs.homeshopmachinist.net//smile.gif

wierdscience
02-26-2004, 11:17 PM
That's good!Now do you have a chamfer at the front of the backing plate bore so that the backing plate bottoms out on the register?

BillH
02-27-2004, 12:03 AM
Yes, I do have that, I had a feeling that are was giving problems, but I chamfered that in such a way that its no longer a problem, The last thing Im going to do still is face back that plateau on the back plate, I think the back of the chuck is hitting somewhere, but using calipers depth function, it says no.

wierdscience
02-27-2004, 12:27 AM
Just for giggles and grins,maybe try numbering chuck jaws and slots and switching them around.

J Tiers
02-27-2004, 12:50 AM
With stuff like this, I like to have a tie-breaker or "sanity check".

Most chucks are quite true in themselves, that is if you get the back surface turning true and centered, the front will be true also, and the OD will be concentric.

Have you checked the chuck OD and faces to see if they are rotating true? If not, then for sure the mounting would be suspect.

If they are perfectly true, but straight and true rod in the jaws wobbles, then the jaws are suspect, or crud might be on one of them.

Failing a fault with the chuck or jaws, as indicated by the sanity check, I would be sure of the following on teh backplate:

1) That the threads and back contact surface where backplate hits spindle were machined at one time and without moving anything. That way you know the thread axis is perpendicular to the contact surface

2)that the threads are relieved enough that the backplate actually hits the spindle shoulder, and the threads don't hang up on the "register" (sorry, Rich C). (try a 0.0015 feeler all around).

3) That the "register" relief isn't so tight that it is "fighting" with the threads and cocking the chuck off-axis. Again, try a feeler all around the shoulder.

4) That the mounting face of the backplate is flat and rotating true without "wabble", and that is was cut smooth and has no swarf, burrs, or roughnesses on mating surfaces.

5) that the spigot register on the backplate to aline the chuck is rotating true.

6) that all mating surfaces are clean and don't have swarf or burrs etc.

If you methodically work thru starting with the spindle and moving on to the backplate, and then chuck, there is no way for you to be fooled, you WILL find the problem.

If you scatter-gun your checks, you will have no fixed points of reference, and you may never find the problem.

And, DO NO HARM. Don't modify anything you can't be sure you can aline or turn true again.

I sincerely hope you have not messed with the actual back surface of the chuck. Sounded like you did from what you said. If you have, you probably will never get it right no matter what you try.

Carl
02-27-2004, 01:44 AM
Before giving up on this chuck, I still think I would try truing it up on an arbor between centers. I know franco didn't have any luck with this technique but he did say his chuck jaw ways in the chuck were faulty. Make sure your centers are aligned using a test bar and dial indicator, then face and center drill both ends of a piece of bar stock. Take a truing cut to clean up one end of the bar long enough to accomodate the length of the chuck jaws. On the other end turn a shoulder and threads to simulate the end of your spindle. Clamp the chuck on the bar using the chuck jaws and face off the back, then remove the chuck and use the threaded end of the arbor to face both sides of the back plate.

BillH
02-27-2004, 03:03 PM
well the backplate is running true, that is a constant since I indicated it, bolt the chuck to it, now getting .010 runout near the chuck, and .015 near the end of the reamer. I called up Enco, they are issuing me a Return number.

canonicalman
02-27-2004, 03:58 PM
One last thing to check related to facing the backplate is whether the cross slide on your lathe angles in toward the spindle slightly as it is fed in towards the bed. All lathes are set up in this way as to machine a plate slightly concave. The usual amount is about 0.0005" in 6".

The reason for this slight concavity of cut is to ensure that two mated surfaces will seat well. If the surfaces are convex(bowed out at the center) then the two surfaces will never seat well. No matter how you tighten the two parts together they will always be off.

Perhaps when the backplate is machined it gets just such a convex surface?

BillH
02-27-2004, 04:04 PM
Canon, I dont think that would create the kind of run out im having. Its old news anyhow.
Enco is going to take it back, probably wont refund me, and I'll probably of had made my first 200$ mistake.
Let this be a lesson to all of you..
DONT MODIFY ANYTHING if you think its defective, Return it or risk screwing yourself.

BillH
02-27-2004, 04:13 PM
Oh, and if they don't accept it, I will make a new back plate, and True up the chuck between centers like others have said.
If that still does not work, I will try grinding the jaws.

Carl
02-27-2004, 04:30 PM
If you do true it up on an arbor between centers, make sure the arbor isn't any longer than necessary so you don't get any sag in the arbor from the weight of the chuck.

J Tiers
02-28-2004, 12:02 AM
<font face="Verdana, Arial" size="2">Originally posted by BillH:
IT went poof because my spacer collar was NOT flat, but had .008 wobble on it, since I made it with the bison chuck before I decided to face down the back plate. Hmm, I guess that tells me that the CHuck was out of whack since I first got it!

[This message has been edited by BillH (edited 02-26-2004).]</font>

Not necessarily, because it is notoriously hard to machine a thinnish spacer flat when holding in a chuck. That's what surface grinders are for.

That can be a problem even when held with the inside jaws on one of the steps, you have to be sure to tap it down, and then may run afoul of the clearance at the step of the jaw.

Another thought, which applies less, as I see that you bought the backplate with the chuck.......but still might be true, because you faced it again.

Does the backplate have a little undercut at the inside corner where the "spigot" that alines the chuck meets the faced
portion of the backplate?

If the chuck is riding on a radius there, it will never sit straight. It will be off one way or another, maybe a very small amount. You should be able to detect that with feelers or sighting against the light.

Most chucks do have a beveled off area where that inside corner is, but it all depends on the relative radii of the chuck and that inside corner.

That could have been a problem from the first, as you suggested.

franco
02-28-2004, 11:09 AM
BillH,

I am not looking to argue with Carl, who I suspect has more machining experience than I will ever have. I completely agree that the chuck should be salvageable using his proposed method of mounting it on an arbor between centers, but with three important provisos. When the chuck is mounted on the arbor:
1. There should be no movement in the needle of a dial indicator mounted in the toolpost when it is moved across the circumference of the chuck in several different positions around the chuck.
2. There should be little runout around the circumference of the chuck - probably up to 0.004 in. is OK
3. There should be no runout on the front face of the chuck. No runout here shows that the jaws are square with the chuck face, which should be the case with a new chuck.

If there is runout on the front face when the chuck is mounted on the arbor, this method will not give a satisfactory result, because the axis through the chuck jaws is not the same as the axis through the chuck. If the back of the chuck is machined in this condition, the front face of the chuck will wobble after it is mounted on the backplate, (ask me how I know!) and it won't be possible to hold work accurately in the steps on the jaws, even if the jaws have been ground to match the axis of the chuck.

If the front face is not true when on the arbor, the chuck can still be salvaged. Remove the jaws, clamp it onto your faceplate (after taking a light cut across the faceplate) and machine the back true. Another alternative is to find a lathe with a big enough 4 jaw chuck to hold the body of yours and remachine the back after making sure it is mounted true in the 4 jaw chuck.

After the back has been trued, follow J Tiers list of checks when remounting the chuck, then if necessary, grind the jaws. Note that if it is necessary to grind the jaws, (and it should not be), they should be ground while under load. The chuck should then be quite satisfactory.

Good luck with Enco, but if the worst happens, you should certainly be able to salvage the old chuck.

franco



[This message has been edited by franco (edited 02-28-2004).]

Carl
02-28-2004, 01:49 PM
Franco is right, I'm assuming that since this was a new chuck, and assuming it wasn't defective, the axis of the chuck body and the clamping axis of the jaws should be very close to the same. If you get it on an arbor between properly trued centers and the circumference or the front face run out, then the chuck was defective to begin with. Then the only option is as franco said, to mount it on a face plate or in a four jaw chuck and true the rear face to the front face. I suggested the arbor method because I assume Bill doesn't have access to a lathe large enough to do the work on a faceplate or in a four jaw. The arbor method made the work possible on a smaller lathe assuming that the chuck body wasn't defective, and that only the rear face was surfaced incorrectly.

Thrud
03-02-2004, 06:24 AM
Bill

Am I correct in saying that you Actually machined the chuck itself?

If so, in God's name what would posses you to do this? The ONLY part you ever need to modify is the backplate.

If the counterbore was too deep the proper procedure is to remove the backplate and use a hardened ground parallel bearing spacer thicker than your spindle "register",

place it on the spindle,

install the backplate on the spindle backwards and insure that it is snug angainst the rear shoulder of the spindle.

once this is done the rear of the backplate can be carefully faced so that the counter bore is slightly deeper than required for use on your spindle.

The opening of the counter bore must be properly chamfered to prevent interference when it is screwed onto the spindle.

remove the spindle,

mark the rear of the backplate with prussian blue and spin the backplate on snugly

remove and check for even flat fit - at this point you should decide if you want to scrape the backplate in to your spindle (40 bearing points/inch) otherwise look for even contact.

clean off the prussian blue and remount the backplate on the spindle.

Now you have to re-true the backplate so it will run the chuck on your lathe accurately.
take a light facing cut and check for fit after stopping the lathe. (note most backplates will have a step feature that sets into the back of the chuck for added support - this should be a snug fit - hand pressure required to assemble) REMEMBER TO CHAMFER ALL OUTSIDE CORNERS OF THE BACKPLATE SURFACE THAT MATES TO THE CHUCKS - THESE CAN BE INTERFERENCE POINTS AS WELL!

to finalize assemply with a scroll chuck it is wise to check run out with the chuck bolted in each combination of bolt patterns (a chuck attached with six bolts could have one of 6 more accurate positions) once you find the best fit mark it with a witness mark for future refference and asemble and torque the bolts evenly with a torque wrench.

Independent chucks can just be assempbled and torqued down per spec.

A nice final touch to your chuck is to machine a UHMW PE tube to take up the hollow space between the chuck body and the backplate. I find this to give a cleaner chuck as chips never accumulate in there the tube forms an extension from the spindle bore to the chuck body. If done right it fits tightly and no dirt get in.



[This message has been edited by Thrud (edited 03-02-2004).]