Improving the Mini Lathe

My experience with these Chinese machines is that anything that is not visible has not been deburred. I would check the jaws and scroll for burrs before doing anything drastic. Also check that you see the same runout when chucking to a different diameter; it might be the scroll that is off.

If you have not done it yet, check the mating of the carriage V and flat to the bed. I found that filing and scraping this to fit better made a big difference. The lathe felt much more solid.

Does your 3 jaw that the lathe came with have a master tightening socket?, usually marked with a 0 in my experience. If so this is the socket that when the chuck is tightened with that socket will give the minimum runout, at least for quality chucks when new.

[ ... ]

Huh? I thought that this was an *independent* 4-jaw chuck -- a separate adjustment for each jaw. You tune out the runout by loosening the side opposite the high point, tightening the side which has the high point, and re-tightening the opposite side. If it is between two jaws, you will need to do this with both pairs of jaws. Tune it that way before jumping to the conclusion that it needs truing.

Using the grinder (as a toolpost grinder) on an independent

4-jaw chuck -- all you will be able to do is to make sure that the gripping surfaces are parallel to the bed axis.

If it is a universal 4-jaw chuck instead (one where a single key adjusts all four at once -- and those are rare) -- that is better used on square workpieces, not on round. But you might indeed be able to true it up a bit with the grinder.

I notice no warning about protecting the bed and all other precision surfaces from the grit produced by the grinder. Cover all surfaces -- either with newspaper wet with WD-40, or with aluminum foil, and be careful to not spill the collected grit on the precision surfaces. Protect *everything* below the toolpost. This includes the spindle bearings, too -- and leadscrews -- and tailstock ram.

Good Luck, DoN.

Grinding the jaws of a scroll chuck will ensure that they run true at that diameter only. Any other diameter workpiece - larger or smaller, may or may not be true.

I don't think they are all that rare. There are lots of them for sale, and this is what I have.

It does seem to grip aluminum pieces better with less force and less scaring.

Good advice.

After my original post I found another surface that was way out of true. One I had not expected it to be an engaging surface. Oops. Measure everything I guess. On the adaptor plate there is a flat where the bolt holes are and then a cylindrical protrusion in the middle. I expected it was there only for alignment. It turned out to be taller than the depression in the back of the chuck. I do not know if that should be the surface to engage the chuck or if I should face it off so that the outer flat engages the chuck.

Show us some photos. There may be other options.

jsw

Ok, I am just about out of ideas. One thing to note though is everything I have done has improved it some. I have concluded (several steps ago) that the majority of the problem is a wobble somewhere. I cut the cylinder on the adaptor plate down so it just sets in the chuck but does not contact on the face and I get .0023 apx at the tip of the jaws with and .0036 an inch away from the jaws. Sounds like wobble, not an out of center condition like a jaw being off would cause or mounting off center. Some mating surface or the chuck itself still isn't true. The only surface I have left that I could try to true up is the back of the chuck, and I haven't a clue how I would do that except in another lathe. One that is already good.

I gave up on the jaws a while ago by testing with some shim stock and opening and tightening over and over again to get a feel for where it starts to grip. They are not perfect, but they are not the main problem.

I do wish I didn't have to remove the adaptor plate from the spindle plate in order to remove the chuck. It always makes me nervous that I am getting a piece of dross somewhere during reassembly. I have marked the plate(s), chuck, bolts, and nuts so it gets reassembled the same every time.

Well, off to my day job. I'll play with this again this evening.

[ ... ]

Where did you find this -- and why did you get it? The usual advice of getting a 4-jaw chuck for minimizing runout is assuming that the 4-jaw is an independent 4-jaw, not a universal (scroll back) chuck.

An independent 4-jaw is a bit more work to tweak true when mounting a workpiece -- but you can get it as true as your patience holds for.

Yes -- I have a universal 4-jaw for my Compact-5/CNC lathe, but it was a mistake in the ordering, as I wanted an independent one. So, what I wound up doing was modifying a 4-jaw chuck for a Taig lathe (similar size) to properly mount on the Compact-5's spindle, and then ordered a replacement 4-jaw for the Taig from a (then) regular of this newsgroup.

If you want *that* -- go for a 6-jaw (or for things small enough and a lathe large enough, use a collet). The 6-jaw is particularly nice when working on relatively thin-walled tubing. Six points instead of three means less deflection of the walls for a given grip strength.

O.K. The register projection in the backplate should not bottom in the chuck. You probably turned off enough while truing the mount flange so the effective projection was increased beyond what was needed. Turn off enough of that projection so it does not bottom.. It *is* just for alignment.

Enjoy, DoN.

There is a way to do it, but you would wind up having to true the register cavity at the back of the chuck, which would mean that it is larger, and will need to machine a new backplate to match that.

The trick is to take a piece of steel which is the biggest which you can hold in your (other) chuck, clamp it, turn the diameter down a little where it extends from the chuck jaws to produce a surface which

*is* true -- no matter how bad the chuck may be (don't release it from the chuck until you are done, or you will have to re-true). Once you have this surface, clamp the chuck you are working on, backwards, and you now can true the outer rim which contacts the backplate to match the jaws. *But* -- *don't* do that. You have *now* gotten to the point where grinding the jaws to true them makes sense. The gripping surfaces are not square to the axis around which the chuck rotates. It doesn't matter whether the actual error comes from some other part. Once you have the jaws trued, it will neutralize everything else. I was objecting tot the truing of the jaws before when I was thinking that it was an independent 4-jaw chuck, not a universal.

BTW That 0.0023" runout at the exit of the jaws is probably close to the best you can expect to get from a universal chuck. After grinding the jaws, you will probably be better than that when gripping something the same diameter as the setting of the jaws when you ground -- but at other diameters you can have more error.

And that 2000 RPM to move the jaws out may or may not be enough, depending on how stiff the jaw motion in their ways happens to be. For a 3-jaw, you can put three identical width pieces of steel between the angled faces of the jaws to preload them all out. That *might* work with the 4-jaw universal which you have as well.

And running the chuck at high speeds with nothing gripped can result in the jaws working out and flying across the room, again depending on how stiff the motion actually is.

This is one of the reasons which make this kind of lathe more awkward to use than the various other ways of mounting the chuck to the spindle. The ones which I have (varying degrees of) experience with are:

1) The threaded spindle. The backplate (with chuck a part of it) simply screws onto the spindle nose. This is fine until you want to machine with the spindle in reverse for whatever reason at which point the chuck can wind up unscrewing. 2) The Long-Taper (L-00 is my example). A long taper with a key to prevent the chuck from rotating around the spindle, and a threaded collar which screws onto a male thread at the back of the chuck's backplate and draws it firmly onto the taper. 3) The Hardinge spindle nose. A long taper with a milled slot in it which looks like a very wide topped 'Y' (or a slightly angled 'T'. The chuck slide in place with the pin engaging the slot and is then rotated to let the angle of the top bar pull it on firmly. 4) The 'D' series -- a very short central taper and a number of pins which vary with the size protruding from the backplate, which go into holes in the spindle nose and are drawn in with cams in the spindle rotated by a key like a chuck key. (There is a cut in the side of each of the pins which engages the cams.) 5) The chuck slides on over the spindle nose (e.g. the chucks on my Compact-5/CNC) with a register (40mm in this case) to assure concentricity, and three or four Allen head cap screws go through the chuck into threaded holes in the face of the spindle. You can't remove or install the chuck while it is holding a large workpiece, because that blocks access to the screw heads -- but no need to disassemble the chuck from an adaptor to remove it. 6) Chuck on a collet body, so it can be used in a lathe set up for collets. This is the standard way for watchmaker's lathes. I've got a pair of these in 5C collet size (a 3-jaw and a 4-jaw independent) which are convenient when I don't need to put the larger chucks (6" 3-jaw and 10" 4-jaw) on the spindle, because those are *heavy*.

There are certainly other system which I don't have any experience with -- but the thing in common for *all* of these is that they don't require disassembling the chuck from an adaptor.

The nuts probably don't matter but the marking is in general a good idea.

Good Luck, DoN.

Thanks for all the help Don. Had some big service jobs come in the last couple days and have not had time to play with this, but now I am torn between taking it all apart and taking one more finish pass at the adaptor plate or just taking a swing at that jaws with a rotary tool. Maybe I'll decide by tomorrow when I have time to play again.