Page 1 of 3 123 LastLast
Results 1 to 10 of 22

Thread: How do you make a 3 jaw chuck accurate?

  1. #1
    Join Date
    Jul 2007
    So. Cal.

    Default How do you make a 3 jaw chuck accurate?

    What is the best way to make a 3 jaw chuck accurate? I made a test bar using 7/8 drill rod about 4" long in my collet and it runs true. When I put the bar in my 3 jaw it has .028" runout. I have a burned chuck with a L00 taper on a Clausing 13" lathe. Is there a way to adjust the chuck, or do I need to cut the jaws with a boring bar? Thanks for any help.


  2. #2
    Join Date
    Jul 2005
    Chester, NH


    A burned chuck?

    3-jaws are not accurate, just by thier nature. That being said, .028 is real bad! If its not a "set-tru" chuck or similiar then there is no way to adjust it. If the runout is just due to screwed up jaws, then cutting them with a very stiff boring bar would help. However, if the chuck is old and the scroll is worn out or screwed up, you will cut the jaws true but as soon as you try to clamp something in it, you'll still have runout.

    What kind of condition is the chuck in? How old is it?

  3. #3
    Join Date
    Aug 2007


    The best way to recondition a beat up three jaw chuck is to take it off, and replace it with a four jaw or a collet when you need to. Wish I was kidding.

    If the scroll is beat up, truing the jaws will make it "right" only in one spot. Move it off, and the wear on the scroll will be different. Unfortunate, but there isn't really anything you can do to fix it up scroll. If it's the face of the jaws that's screwed, you can try to clean it up with a toolpost grinder (not the best for the lathe) or put a carbide boring bar in there and try to kiss the dings out.

    For a really FUBAR'ed chuck, I'd machine it to accept aluminum soft jaws. That way, it doesn't matter how screwed the chuck is (as long as it still holds the part, doesn't kick, et. al.). Whenever soft jaws are used, they need to be recut for the specific diameter to be held anyway, so they pretty much always reclamp under a thou. Better, if you have a three jaw in good condition. Look up the specs for soft jaws. It's something like two bolt holes at some spacing to hold the jaw, and 60 deg grooves on .060" spacing to keep it still. Since you'd be making your own jaws, you could mangle the previously-master jaws any way you want, and don't have to adhere to any particular soft jaw spec. This applies to single-piece chuck jaws; you might already have a two-piece jaw, in which case take the jaws off and use the pre-existing pattern.

    Soft jaws are really neat; you can hold all kinds of shapes at all kinds of offsets in a three jaw. Mount the chuck under a mill, and you can cut squares and hexes into them. You could also make some pie jaws, which open up a whole strange world of clamping options. Same with clamping on inner diameter features.

  4. #4
    Join Date
    Jan 2003
    Chilliwack, B.C.


    You can characterize the chuck, then see if it's worth trying to fix up. You'll need a few different diameters of round rod, some masking tape, a marker, and an indicator. The first thing to test for is whether the mounting of the chuck is repeatable. Indicate the body of the chuck and using tape, mark the point of furthest deviation and record the number on paper. Remove and remount it and check again. If it's not repeatable, fix this first. This would be the backplate/spindle interface area. Then if possible, rotate the chuck on the backplate and check the body for runout again. There might be one orientation of the chuck on the backplate which results in the least runout of the body, and/or minimum wobble.

    Now check to see if the chuck has a master key marked somehow. If it doesn't, then place some bits of tape and mark them as 1, 2, and 3. Chuck up one of your rods and do the following- tighten the chuck using the master key, or what you have marked as number 1. Indicate the rod fairly close to the chuck, and place a bit of tape at the point of furthest deflection on both the rod and the chuck. Mark each bit of tape as number 1. Rotate the chuck so you can tighten using number 2 key. Loosen the chuck, then tighten again with this key. Don't rotate the rod while you do this, but do hold the rod basically centered as you tighten the jaws. Indicate again, and place tape for the point of furthest deflection again, numbering these with a 2. Repeat again using the number 3 key. Record the runouts on paper. Repeat this sequence again to see if the runouts match when the jaws are tightened from each different key.

    So far, this procedure should tell you two things- one, whether there is a particular key to tighten with which will result in the least runout of the rod, and whether the nunouts are repeatable.

    If you find that the deviation is always towards the same position on the chuck, and minimal if you use a certain key to tighten with, you might be done. You can repeat this sequence for other sizes of rod to see if the pattern matches the previous one. If so, but the runout is still excessive, you can probably get away with internal grinding of the jaws, providing you do it properly. If the readings jump around seemingly at random, then the scroll is probably worn.

    Something else to check early on is whether the jaws can hold the rod without it wobbling. Try this- tighten firmly on the rod, but not too tight, then power up and look for wobble. If you can see it, stop the spindle and push sideways on the rod in the appropriate direction to make a correction. Power up again and check. You might want to go directly to using the indicator to see any wobble, if it's too slight to see. If you can push sideways on the rod and have it take a set in any number of positons, it's possible the jaws and the grooves they slide in are worn too much.

    If you can minimize the wobble by pushing on the rod, then tighten the jaws fully and check this again. Using the indicator, what you'd like to see is some small deflection as you push against the rod, but a return to the previous position once your force is removed. If the chuck can't control wobble, there's no point in trying to reduce runout.

    If you come to the conclusion that cutting or regrinding the jaws will work, then you need to clamp on something round at the very back of the gripping surface of the jaws to keep them tight in their bores for the cutting or grinding process. Remove the jaws and carefully grind away the last bit to the level of the freshly trimmed portion.

  5. #5
    Join Date
    May 2003
    Vancouver's Island


    If you come to the conclusion that cutting or regrinding the jaws will work, then you need to clamp on something round at the very back of the gripping surface of the jaws to keep them tight in their bores for the cutting or grinding process. Remove the jaws and carefully grind away the last bit to the level of the freshly trimmed portion.
    If the chuck and jaws are worn enough to require regrinding then clamping on something at the back of the jaws will likely cause them to be ground bell mouthed. The jaws will rock when clamped that way causing the front of the jaws to close more than the back.

    I have had excellent results by using a 1/2" carbide finishing burr and simply tightening the jaws slightly on it in back gear to true the jaws. I recently reconditioned a small three jaw this way to use on my 4th axis and it can now hold a bar true over 6 inches within a few thou. Previously it was more like 50 thou and very bell mouthed.
    Free software for calculating bolt circles and similar: Click Here

  6. #6
    Join Date
    Jun 2006
    Austin, Texas


    Quote Originally Posted by Fasttrack
    A burned chuck?
    Pratt Burned -- a very nice English chuck. The Clausing 5900, 6900's shipped with PB chucks.

    Like Toasty says, it sounds like your scroll is FUBAR. Darryl posted an excellent method of determining how bad the chuck is.

    You can grind the jaws with a toolpost grinder or a Dremel zip-tied to the toolpost, but that only corrects the runout for a specific diameter, and the best you'll be able to do is several thou.

    This is why you don't want to buy used chucks on Ebay.

  7. #7
    Join Date
    Dec 2003


    Or you can use a soft faced hammer to bump the work into place (using a DTI) then tighten.
    I saw the machinists do this often when working with very old worn equipment.
    I do it on my 3 jaw also. Sometimes you gotta do what you gotta do.
    I have tools I don't even know I own...

  8. #8
    Join Date
    Nov 2001
    Toledo, Ohio


    The Pratt Burned chuck is an excellent chuck, and the L00 mount is one of the best for repeatability.

    With any chuck runout problem, it is best to start at square one and make sure everything is as it should be before any fixes are attempted. It is always the best idea to first find the problem and then fix it. Grinding jaws is the last step.

    The first step is to remove the chuck and make sure the mounting surface is clean and free of dings and dents. Then check the lathe spindle and mount for runout. Indicate the lathe spindle and check for endplay or loose bearings.

    The second step is to dismantle and clean the chuck and inspect all parts. Take special care in cleaning and inspecting the scroll, and the fit of the scroll in the body of the chuck. If the bore is loose on the center hub of the chuck, it can float, and positioning will suffer. Make sure the scroll is clean and no chips are imbedded in the thread. Clean all the jaw threads and inspect for imbedded chips.

    If the chuck is mounted on a separate backplate, check the fit, that there is no float, or lost motion there.

    If it does have a separate backplate, install that without the chuck mounted, and indicate it to make sure it is parallel and has no runout. It might be necessary to take a skim cut to true the surfaces to the lathe spindle.

    If the spigot that locates the chuck has runout, it can be skim cut to remove the runout. If this is necessary, when remounting the chuck, it can be bumped to remove runout before tightening the mounting bolts. This method might be the ultimate solution.

    Once all of these steps have been taken, darryl's suggestions are excellent. When grinding the jaws, a preload is necessary to duplicate the loaded condition. There are several methods of accomplishing this.
    Jim H.

  9. #9
    Join Date
    Jan 2004


    Agree with the above, mostly.

    1) if the scroll is loose on its "pivot", very possibly a shim can be put in between. I have done this, and got an instant improvement.

    2) with 0.028 "off", it is either a simple problem, a re-grind issue or a big hassle....

    The simple one is if the jaws are just in the wrong slots, but still in the right order. There are three possibilities, and it takes only a short time to take out the jaws, clean them, and try them in all 3 possible slots

    Wrong order will cause a big problem, generally, and should be obvious, errors of several MM.

    The BIG hassle is if the jaws are actually "sprung", meaning they are loose and the chuck body is distorted. If you can wiggle them so that the gripping part will lean at different angles relative to the true axis of the spindle, it is paperweight time. Toss it in the scrap bin and don't look back.

    If the jaws are tight, but are worn "bell mouthed" (which is very common), then you can at least get them back parallel.

    Someone will now jump in and say "they will only be right in one place". They are correct, but it doesn't matter....... the chuck will hold things as well as any 3 jaw should be expected to.

    But the method of grinding is the key.

    This originated with Rich Carlstedt, and since he hasn't said anything, I will......

    1) drill a hole in the end of each jaw, parallel with the axis, in the top step. 1/8" or 3mm is a good size.

    2) put a pin of the drilled size in each hole, with the chuck in place.

    3) close the jaws on a slice of thin pipe, so that the pins hold the pipe, and it is pretty much out of the way of the jaw gripping surfaces. This pre-loads the jaws correctly. Jaws should be open enough for the next step, so choose the pipe diameter. Usually a couple inches is OK.

    4) clamp a die grinder, or Dremel tool, etc to the crosslide so that A suitable stone will enter the open jaws. Start up the grinder, spin the chuck by hand, adjusting cross-slide to just kiss the face of the highest jaw, and grind the full length of the jaw by moving the carriage and turning the chuck on the spindle. spin chuck gently to be sure no other jaw is now higher.....

    Adjust position of crosslide as the highest spots are ground down, until all jaws are being ground. Stop adjusting, and spin chuck slowly as you continue to feed the grinder with the carriage feed, until no jaws produce any sparks.

    Stop at the least amount of grinding that gives a good surface on all jaws.

    Yer done.

    The faces will be lightly concave, but you can now use the chuck again, so live with that.

  10. #10
    tattoomike68 Guest


    .028" is huge, you can hit some of the jaws with an air sander and do no harm.

    you can do this, chuck up a small ring at the back of the jaws, hose clamp a die grinder to a boring bar and grind the jaws up to the ring. once they are true at that point move the ring to the front of the jaws and finnish the tiny left over back side of the jaws.

    thats about all you can do with an old chuck, dont exspect much from a 3 jaw anyway, hell I use a hammer, it dont take me all day to dial stuff in.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts