Page 1 of 2 12 LastLast
Results 1 to 15 of 21
  1. #1
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default Repairing a chuck/replacing a chuck/register frustrations!

    This is a bit of a saga.


    My sad old ex-school Hercus has a sad little Pratt chuck. Its jaws aren't parallel – splaying out toward the tailstock (from youth grabbing little things very tightly?), and its about 4thou out of true.


    Last week, I decided to try improving it.


    First, chuck off the backplate, and clock its front. It was wobbling by nearly 100 microns!
    This chuck and plate must have been swapped from another lathe spindle?

    Easy, I think. Re-machine the "spigot" to fit the chuck, and we should run true.

    A pleasureable 10 minutes to remove 2mm, remount the chuck, remeasure.

    Hmmm. Body & face about 60 microns out.



    Second, machine the jaws. Find a ring to expand the jaws out into, put pressure on, and bore a little off them with the sharpest tool I can find - an internal threading insert:
    IMG_1572.jpg

    5 or 6 passes, at a fraction of thou each time. About 1.5thou total D.O.C. until the front inside of the jaws got a little round machined in their middle.

    I even decided to re-cut the little deep grooves, for the next time I'm boring out washers.
    My 16IR was a little too pointy for that. Had to borrow a different internal thread tool:
    IMG_1573.jpg

    Now, vacuum out the swarf, and mount a 40mm precision round for testing, and...


    still nearly 100 microns, or 4 thou out

    I try a different round – a 16mm ejector pin – it comes in at nearly 2 thou.

    I try tapping the high side, and notice light under the front of the jaws.
    Not parallel? What the?


    Remove the round, clock the freshly machined groove along the middle of one jaw as best I can
    (it is hard to get good access in there).
    Inconsistent results.


    I fiddle around with the jaws a bit. They have always been loose in the scroll, but today I notice there is also slight wobble in the T slots. Maybe machining them while clamping outside won't work, because the jaws are rotating in the T slots?


    Scratch my head a bit. Decide to re-machine the front while grabbing something deep in the chuck. Find 20mm round slug, put it deep inside, past the front of the chuck body, and re-do the boring.

    Another 2 thou D.O.C., and I have jaws with a big step in them. Think about milling, Dremeling, hand filing, but decide to surface grind:
    IMG_1574.jpg

    Tricky to hold and shim for the level/taper, but end result looks something like:
    IMG_1575.jpg


    So, mount them back, expecting perfection, and... still nearly 2 thou out of round.
    Plus, there is still a slight step in the back of the jaws when grabbing small round stock.


    Another day, work out a way to machine that out in the chuck. I find a little steel sleeve, loosely grab it at the front of the jaws, bore extra clearance so it doesn't foul on my "boring" tool, and gently take the last of the step out:
    IMG_1576.jpg

    which seems to not line up with the previously machined front of the jaws, so I also grab my 20mm round inside the chuck, and finish the front of the jaws again.


    Clamp small precision round, and check:
    IMG_1577.jpg


    Yay! At last some improvement.



    Test with larger round stock:
    IMG_1578.jpg



    Aurrrrrrgh!!!!!!!!!!!! Worse than before.


    After two days of careful machining and grinding, I have a chuck which is better at some radii than others, and has a total of over 6 thou removed (so this chuck will never grab very small rod again unless I mill/grind the bevels on the jaws).


    The scroll is too badly damaged – time to cut my losses.


    Toolmaker takes pity on me, and gives me an unused – still in box – 125mm TOS 3 jaw. Onto part 2...
    Last edited by nigelpearson; 18th Jun 2021 at 09:03 AM. Reason: missing attachments

  2. #2
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default Part 2 of the saga - new backplate

    I have been keeping some old, side of the road throw out, barbell weights for this purpose (pipeclay's suggestion from a few years ago).


    5lb weight is just thick enough. Mount on sad 4.5" Pratt chuck, grabbing inside the barbell hole:
    IMG_1579.jpg

    and gently machine a little step:
    IMG_1580.jpg IMG_1581.jpg

    then grab that in the chuck, and use threaded rod to lock it onto the chuck. Machine the face with a positive clearance insert:
    IMG_1582.jpg

    (because the negative rake TNMG was rubbing and creating sparks). Lots of fine dust. Vacuum helps.

    Remove rod to finish off the centre:
    IMG_1583.jpg


    Now for the threading. Instead of a bigger lathe, use an old face plate, using the dodgy-est mounting system I can cobble up:
    IMG_1584.jpg


    Bore out to 1.33", with a little step to remind me about the register:
    IMG_1585.jpg


    Now, the change gears. Turn this into that:
    IMG_1586.jpg IMG_1588.jpg

    using two 32tooth gears. Note that the 80tooth is hard to remove - I had to use a bit of brass to lock a tooth:
    IMG_1587.jpg
    Note also that the Hercus chart says to use the 80tooth in the middle, but I don't have any axle/spacer hardware to do that, so the 72 had to go in there.



    Now for the tricky part. My job is mounted hard on a face plate – no gap. A face plate with an 8tpi 1.5" thread on it. So, I align the cutting tool exactly on the spindle thread.

    Tool in the valley (root) on the back of the spindle means it matches the peak (crest) on the front of the spindle, which lines up with the valleys in the faceplate thread:
    IMG_1589.jpg


    and seventeen careful passes later, here is the thread. If you look carefully, you can see the transition half way down where some metal has been pushed into the face plate's thread:
    IMG_1590.jpg
    Thread is a little loose, but as long as I do a good job on the register, it won't matter, right?




    Machine a spacer so I can reverse mount the new face plate:
    IMG_1591.jpg IMG_1592.jpg

    and carefully bore the register. It was hard to see in there without some focussed extra lighting:
    IMG_1593.jpg


    I decide here that I will keep the thickness/weight, and "history" of the barbell casting. So I machine a little of the outer face, but leave as much of the raised lettering as possible.




    Mount the correct way. Tight for the last half turn, rubbing a little at the very end of the spindle thread (will try to make a photo). Smooth off the outside edge:
    IMG_1594.jpg
    (which abraded one corner of a TNMG. Lucky it has 6!)


    Now for the chuck register ("spigot"). Take about 1/16" off, because I don't have a lot of thickness
    IMG_1595.jpg
    and finish up in there with a new TNMG's corner. Take my time, and I get a good friction fit – the faceplate can hold the weight of the chuck!


    Finally, the bolt holes. M8, so I use 19/64 drill to locate on the drill press:
    IMG_1601.jpg
    (using electrosteam's milling clamps in an odd way)

    then clamp and spot using an odd taper drill:
    IMG_1602.jpg

    and drill out at 1/8", then 8mm:
    IMG_1605.jpg


    Chuck came with bolts, but they won't clear lathe's belt tensioning handle, so I countersink for countersunk HT bolts (set screws?):
    IMG_1606.jpg


    Sadly, one of the holes is in an inconvenient place. Had to Dremel a flat before spot drilling:
    IMG_1608.jpg


    End result:
    IMG_1609.jpg IMG_1610.jpg


    But, mount it for testing, and 'tis not as true as I hoped:
    IMG_1611.jpg IMG_1612.jpg



    So, after 1 week, I'm at a loss.

    What is causing it? Spindle shoulder? Thrust bearing? (unlikely) Nigel's bad machining?



    Looks like some more careful dial indicating, and maybe some precision grinding, is in my near future

  3. #3
    Join Date
    Nov 2007
    Location
    melbourne australia
    Posts
    3,228

    Default

    I got a bit lost in all of that. I've never heard of anyone machining chuck jaws true- always grinding. Your variable runout is probably due to you measuring against stock of varying diameters. The runout of a 3 jaw varies with the position of the scroll. If you grind them true at a given diameter, they will only be dead-on at that diameter.

    Also, please decide on a unit of measurement and stick to it! Either thou or microns (pref the metric one).

    Sorry I can't be of more help.
    Chris

  4. #4
    Join Date
    Jun 2007
    Location
    sydney ( st marys )
    Age
    64
    Posts
    4,887

    Default

    How deep is the recess for the register?
    Is the back of the backing plate running out ?
    Did you cut the register at the same time you cut the thread?

  5. #5
    Join Date
    Jul 2016
    Location
    Melbourne
    Age
    35
    Posts
    1,522

    Default

    First thing to do is check your spindle for axial play, end float and run out. Have to trust each component and that means inspect each component.

    Sent from my SM-G973F using Tapatalk

  6. #6
    Join Date
    Sep 2012
    Location
    York, North Yorkshire UK
    Posts
    6,439

    Default

    Hi Nigel, Guys,

    If the original chuck is as bad as you seem to suggest, then do as I do and keep it for rough and rusty pieces to clean them up for subsequent machining.

    I think that you have already twigged that grinding, with a tool post grinder, is preferred to trying to turn the jaws by cutting them. You will probably find that there is some play in the spindle as well. Whilst the chuck end may not show any play the tail end could and this will allow the chuck to describe a wobbling motion.

    Also machining the new backplate with your bad chuck may cause the backplate to have a machined in wobble. I would have done the bore and threads first then machined the plate when its screwed onto the spindle, then you can guarantee that it will be square to the spindle.
    Best Regards:
    Baron J.

  7. #7
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default

    Quote Originally Posted by jack620 View Post
    I've never heard of anyone machining chuck jaws true- always grinding.

    If your jaws are hardened, grinding is probably necessary (unless you have exotic boring bars with CBN or PCD inserts).


    For soft jaws, or your average old steel like these, I guess it depends on how much metal you are removing. Best way would probably be to machine then grind?





    Your variable runout is probably due to you measuring against stock of varying diameters. The runout of a 3 jaw varies with the position of the scroll. If you grind them true at a given diameter, they will only be dead-on at that diameter.
    For worn out chucks with crappy scrolls, yes. For a new chuck, though?


    I have used good chucks that run out less then half a thou over their entire range.



    Also, please decide on a unit of measurement and stick to it! Either thou or microns (pref the metric one).

    When talking about the Hercus, with its imperial lead screw and thou-based slides, I usually talk thou.

    However, I bent my imperial plunge DI, so have had to use the metric Mitutoyo recently. Hence some metric crept in

  8. #8
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default

    Quote Originally Posted by pipeclay View Post
    How deep is the recess for the register?
    Nearly 1/4":

    IMG_1615.jpgIMG_1614.jpg



    Is the back of the backing plate running out ?
    Did you cut the register at the same time you cut the thread?
    Valid questions, Peter. The part of the back that I can easily measure does run out a lot:
    IMG_1613.jpg
    but that flat outer lip was machined at a different time to the middle.


    Thread was done much earlier than the register.



    The register was machined in several steps. Machined out to 1.46", then took off back plate and spacer, flipped over, tried on spindle, too tight, spacer back on, flip over and remount, machine a little larger, et c.


    Given more time, there are things I could have done much more accurately;
    1. machine the thread tighter
    2. machine the spacer more accurately, or surface grind it
    3. mark alignment of the spindle, spacer, and back plate for better consistency
    4. borrow an internal mic, or telescoping gauges, to do the register
    5. use a live centre against the spindle to reduce spindle float



    but, I thought the back plate was locating quite well on the spindle before I machined the chuck locating spigot/step/groove. I thought that was the only absolute requirement?

  9. #9
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default

    Quote Originally Posted by caskwarrior View Post
    First thing to do is check your spindle for axial play, end float and run out. Have to trust each component and that means inspect each component.

    There is some axial play (unmeasured). Pushing up and down on the chuck is possible. When I was building the headstock, I could decrease it by clamping the spindle harder, but then I couldn't turn the spindle by hand!
    For machining, light cuts should stay within the oil's float capacity. Heavier cuts might deflect enough for steel to rub on the iron headstock, but for the light machining of this back plate, I doubt it?


    Haven't measured end float. A DI on the stepped shoulder (that the back plate screws up to) shows maybe 5 or 6 microns when rotating the spindle. Basically, within the limits of surface finish and the size of the ball on my lever DI. To measure end float properly, I would need to push it in with the tailstock, and then back with a clamp?


    Run out on the flat part of the spindle was about 10 microns, so less than half a thou.
    (It is probably now worse, from some light scratches that the back plate has left there. It is a tight fit, even with the smear of machine oil)


    I did think of spindle float and run out, but decided that should average out in the final machining step?
    (the spigot that the chuck sits in)

  10. #10
    Join Date
    Jun 2007
    Location
    sydney ( st marys )
    Age
    64
    Posts
    4,887

    Default

    It would of been better to have finish machined the register and thread in the one setup, taking it off and on probably exaggerated the errors.

  11. #11
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default

    Quote Originally Posted by BaronJ View Post
    Hi Nigel, Guys,

    If the original chuck is as bad as you seem to suggest, then do as I do and keep it for rough and rusty pieces to clean them up for subsequent machining.

    Indeed. It was the not gripping (not being parallel) that really needed fixing. Having less run out would have just been nice.





    I think that you have already twigged that grinding, with a tool post grinder, is preferred to trying to turn the jaws by cutting them.
    I do have a little grinder that might have done the job, but I think turning them actually did OK.

    (basically, it is trading off a slightly inaccurate interrupted cut, against a nasty dust producing process that needs a flimsy little grinding wheel and shaft to fit inside the 20mm hole in a chuck?


    You will probably find that there is some play in the spindle as well. Whilst the chuck end may not show any play the tail end could and this will allow the chuck to describe a wobbling motion.

    Also machining the new backplate with your bad chuck may cause the backplate to have a machined in wobble. I would have done the bore and threads first then machined the plate when its screwed onto the spindle, then you can guarantee that it will be square to the spindle.
    That was pretty much my process, except that the flat centre that screws on to the spindle might not have been as true as it could have been. Still looking at correcting that.

  12. #12
    Join Date
    Nov 2007
    Location
    melbourne australia
    Posts
    3,228

    Default

    Quote Originally Posted by nigelpearson View Post
    For soft jaws, or your average old steel like these, I guess it depends on how much metal you are removing.
    Are you sure you don't have hardened jaws? Even my crappy old Chinese 3 jaw chuck has hardened jaws.


    Quote Originally Posted by nigelpearson View Post
    For worn out chucks with crappy scrolls, yes.
    Isn't that what we are talking about here? I mean, if you're grinding your chuck jaws, it's pretty much a given the chuck is old and worn.
    Chris

  13. #13
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default

    Quote Originally Posted by jack620 View Post
    Isn't that what we are talking about here? I mean, if you're grinding your chuck jaws, it's pretty much a given the chuck is old and worn.
    First post, yes, was cutting/grinding old Pratt's jaws.

    Second post, no. Newish TOS 125mm chuck on back plate machined in old chuck (and on an 8" faceplate).

  14. #14
    Join Date
    Aug 2019
    Location
    Revesby - Sydney Australia
    Age
    56
    Posts
    1,183

    Default Some improvements

    I took the new chuck off my back plate, and did some measurements;


    • Chuck body grabbed and centred in a larger lathe's chuck. TRO at both ends about 0.001".
      Rear lip (that fits over the spigot in the back plate) about 0.004" lateral wobble???
    • 2-3" micrometer, front face of chuck, to rear lip (that fits over the spigot in the back plate) are parallel.
    • Grabbed 16mm precision rod (ejector pin) in larger lathe's chuck,
      then clamped down on that with new smaller chuck.
      Inside and back of rear lip also about 0.004" run out.



    This seemed quite wrong for a new chuck. I ignored that for a while, and concentrated on my back plate.

    Thoroughly cleaned thread on spindle and in back plate, and oiled spindle thread. Unmounted and re-mounted back plate several times. To reduce any spindle float issues, I push the spindle into the thrust bearing with the tailstock, a live center, and a tube against the spindle.

    Back plate's chuck mounting face has about 0.001" wobble:
    IMG_1621.jpg


    so I spend an hour machining tiny fractions of a thou off that, chasing perfection:
    IMG_1622.jpg


    Sadly, only got it down to about 1/3 of a thou (10 microns)
    Started with a round, positive rake insert (RPMW) and got it to the point where it was just putting a little dust in the chip breaker of the insert, but then had to change to a pointier insert for the corner of the spigot. My only sharp inserts are TNMG160402.
    Somehow, a pass with that was cutting nearly a thou of wobble again. No idea how. Maybe their negative rake pushed the back plate further around on the spindle thread?


    Gently filed and stoned a few high spots in the back of the chuck:
    IMG_1623.jpg


    Even dilled the bolt holes out a little, from 8mm to 8.1mm, just in case the countersunk bolts are distorting something:
    IMG_1624.jpg


    Bolted chuck on, not super tight, and clocked the front face. Nearly a thou run out, but really tightening the bolt near the high side got it very accurate:
    IMG_1625.jpg



    Now, to test some round stock:
    IMG_1629.jpgIMG_1628.jpgIMG_1627.jpgIMG_1626.jpg


    Fairly mixed.

    6mm carbide rod was best, then 50mm machined 4140, then 12mm ground something, and 16mm ejector pin was worst at nearly 4 thou. Interestingly, the 6mm carbide was short - held in front third of jaws. Will have to inspect jaw surfaces more closely?


    So this might be a crappy little new chuck
    (that I can't get it any better than the crappy little old chuck),
    but that seems unlikely.

    Will strip and thoroughly clean it next. It might be full of that thick brown goo that manufacturers use for rust prevention

  15. #15
    Join Date
    Sep 2012
    Location
    York, North Yorkshire UK
    Posts
    6,439

    Default

    Hi Nigel, Guys,

    Two or three thou run out is not unusual for even a new chuck.

    I got a new expensive 125 mm three jaw chuck and a commercial Myford backplate. I faced it off so that it was as true as it could possibly be, then machined the register 2.5 mm deep exactly to the size given in the instruction sheet that came with the chuck. This made it a tight push fit into the chuck back. Spotted through for the mounting screws. Drilled and threaded M8.

    After bolting everything up I found that I had to machine about a thou and a half off the register because I couldn't get the chuck to sit flat on the backplate. Total runout with a 20 mm diameter test bar was around 3 thou and got worse as I used larger diameter test bars.

    I subsequently found that whilst trying to get the chuck flat on the backplate that I had raised a burr on the inside of the chuck register and it was that causing the chuck not to lay flat on the backplate. Trouble was it was almost impossible to see and I didn't realise it was there until I got a cut on my fingers !
    Best Regards:
    Baron J.

Page 1 of 2 12 LastLast

Similar Threads

  1. 4 jaw chuck
    By outoftown in forum THE HERCUS AREA
    Replies: 1
    Last Post: 12th Aug 2020, 11:40 PM
  2. Repairing Jacobs No. 34 chuck.
    By nigelpearson in forum METALWORK GENERAL
    Replies: 5
    Last Post: 17th Jun 2020, 05:13 PM
  3. Wax Chuck !
    By BaronJ in forum METALWORK GENERAL
    Replies: 10
    Last Post: 16th Sep 2018, 01:17 PM
  4. collets : to chuck or not to chuck
    By steamingbill in forum METALWORK GENERAL
    Replies: 4
    Last Post: 26th Jul 2014, 10:17 PM
  5. New 4 jaw chuck.
    By Alaks McGREGORE in forum METALWORK GENERAL
    Replies: 25
    Last Post: 8th Jun 2013, 03:28 PM

Posting Permissions

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