Mini-lathe chucks repair
When I bought the mini-lathe, found two chucks inside the box. The standard 3"-3jaws self-centering and a uncommon 3"-4jaws self-centering lathe chuck. Since then I've been using almost only the 4jaws chuck, as the runout on the other was too bad. Now I decided to look at it and try to get the best I can... What a surprise!
I decided to bore the 16mm center hole to 17mm. After disassembling the chuck and found some problems, the chuck was centered on the spindle only with the rear cover assembled, as it also need to be bored.
After boring the center hole, it's time to start fixing chuck problems. It wasn't the jaws or the scroll. They were fine and outside the chuck they could be good meshed. When assembled in the chuck the scroll was too far from jaws, about 0.6mm more than optimal, causing a poor scroll-to-jaws mesh. Also, the scroll have about 0.2mm axial play on the chuck body. With scroll play and poor mesh, runout was up to 0.08mm with variations up to 0.04mm when re-chucking a piece. Also gripping force was always poor on this chuck. No radial play was measured on scroll, and it has to be disassembled and reassembled with a press.
To remove those 0.6mm, the chuck was mounted backwards on the faceplate and centered. It was also checked for any axial slip.
As it was needed to remove the chuck from the faceplate during the facing process to test the mesh, three more clamps were mounted on the faceplate as indexes to save time when re-mounting the chuck on the faceplate. You can see on the photo the shiny faced surface inside the chuck body.
After removing those 0.6mm (it was the result of the previous measurements comparing measures outside and inside the chuck) the mesh was verified. It was mounted again on the faceplate with the scroll inserted. The rear, at center (where the rear cover mounts) was faced until the rear cover touch the back of the scroll. Then it was faced 0.1mm more for final adjustment on the cover.
The rear cover was also faced inside, for a good fit with the chuck body.
Now the time consuming process, removing 0.0125mm at a time, testing the cover on the chuck to check if it still touch the scroll. Many times cutting and testing until it finally don't touch (for less than 0.0125mm).
Here is the result. This can be seen with only the first jaw inserted on the chuck. Now the scroll is fine meshed with the jaws. There is now 0.8mm less play, as the back cover went forward 0.8mm, pushing the scroll to the jaws.
Chucks should run balanced to avoid vibrations, and the optimal balanced mount is done centering it by the external body, so I started this way. This is an uncommon way to center the chuck, I'm aware, but the chuck was repaired centered and I wanted to make a first test this way. I wasn't expecting that using this centering method on a cheap Chinese chuck would gave good runout on the chucked piece, but it did with this one. After centering the chuck, the first TIR test after this work. With internal jaws chucking a 8mm piece - 0.015mm runout, with a 13mm piece - 0.02mm and with a 32mm piece - 0.02mm runout. On repeatable mounts no more than 0.005mm variations (as the dial indicator reads 0.01mm intervals, this 0.005mm measure was estimate). With external jaws chucking a 32mm piece - 0.03mm runout and with a 70mm piece - 0.03mm runout. On repeatable mounts the same 0.005mm variations. Not bad!
As the pinions have to move more inside the chuck to get a good mesh with the scroll gear, new holes were drilled and tapped for the pinions holding setscrews. New socket head setscrews are now on the previous holes to prevent swarf to get in.
For fine adjustment when mounting the chuck, the chuck has now 4mm setscrews. I decided to do it instead of a set-thru backplate as it would reduce some of the working bed dimension. Also because the closer the chuck is from bearings, the stable the spindle is.
Centering the chuck body again, but now with the good help of the setscrews. Of course for a optimal centering the dial indicator must be measuring on the chucked stock. The previous test was very good, maybe too good indeed for a Chinese "no brand" cheap chuck. Now I want to test it all, for sure. I will qualify three stock diameters and test it all again.
To be continued...