View Full Version : Newbie question

02-29-2004, 11:47 PM
I just tried my new 4 jaw independent on my SB9". Problem is the boss end of the chuck is too thick. It only threads onto the spindle for about 2.5 threads before running into the 'register?' at the end of the spindle. Is it typical to need to cut these down? How would you cut this down?


J Tiers
03-01-2004, 01:50 AM
Is it a threaded-back chuck, or a separate backplate?

I can't be sure from what you said what your issue really is....

1) Maybe there isn't any relief inside the threads for the unthreaded register?

2) Maybe the threads aren't cut right and it jams there?

3) Maybe there isn't enough depth of thread in the chuck or backplate and it can't screw down far enough?

03-01-2004, 11:35 AM
FWIW, the threaded back I got with a 3-jaw I bought was apparently made to fit every conceivable 1 1/2-8 threaded spindle ever made, and as a result it fit mine pretty badly. I ended up boring it out, Loctiting in a sleeve, and re-threading it to a decent fit on my lathe.

So...having you say the backplate you got doesn't fit your lathe very well doesn't strike me as particularly surprising. Figure out what's wrong with the backplate and see if you can fix it. (You probably wouldn't, anyway, but DON'T modify the spindle! Mess with the cheapest part, which is the backplate.)

03-01-2004, 12:07 PM
Frank has same problem as I had. Frank, you can turn a spacer that covers the unthreaded part of the spindle, and thread the backplate of the chuck on backwards to turn down the SHoulder.
THIS is IMPORTANT!! That spacer needs to be perfectly flat, mine was not, Im still trying to solve the resulting hell I caused. Need a surface grinder to get it perfectly flat. That or turn it betweencenters which I might do next.
My 4 jaw chuck is doing much better than my 3 jaw, I can get .001 runout at the chuck, or less if I take more time to zero it. about 4" out, its around .003 runout, still not perfect. My 3 jaw... thats another story.

03-01-2004, 12:41 PM
J Tiers: This is a separate back plate that bolts to the chuck. There are enough threads and the threads are good, it's just that the boss bottoms out to the spindle too soon. In other words, if there should be 1.5" of thread and then 1/4" of boss, I've got 1" of boss. Maybe boss is the wrong term...

BillH, I think your right. I Think I need to turn it around and then turn the boss down. I didn't know that I need a spacer though... Right now, I'm not at home to try it & see why.


[This message has been edited by frankc64 (edited 03-01-2004).]

J Tiers
03-01-2004, 03:14 PM
OK if I have you correctly, the unthreaded portion is so long that it prevents very many threads from engaging, although it otherwise is OK.

Then the solution is to turn down the portion that is over-long until it will screw down and use a sensible number of threads.

I think, if the backplate is thickish, I would NOT try to make a spacer, though, unless you have no faceplate.

The right way to do it is to center it bolted to a faceplate, and face the contact surface down to where you need it carefully and very smoothly. Then you won't be far off, as you will have it referenced to the faceplate surface. You may need some spacers due to the center "spigot", but if they are close to similar in thickness (a few thous) you should be OK.

You might, before you do that, check to see if the spigot and the OD are concentric. If so you can use the OD as the reference. Otherwise you may have to use the inside relief (unthreaded area) for centering.

Then, mount the backplate normally, without the chuck, brushing out the threads first.

Carefully face the chuck mounting surface OF THE BACKPLATE true, removing as little as possible, and making sure you don't leave a large radius where the surface meets the "spigot".

Indicate the "spigot" OD to be sure you are still centered, it's possible for that to get fouled up if something didn't go right, in which case you will have to face down enough to re-produce it in a concentric position.

Before you do any work to the spigot, remove it from spindle(all the way), and re-seat it a couple times, indicating the face and spigot each time to be sure its repeatable. Brush out the threads at least once as a triple check.

Even if you don't need to work on the spigot, and for sure after you do any work on it, I'd do the reseat test anyway.

Now you should be good to go if it passes.

The reason this is OK is that you reference the previously true (we assume) chuck mount surface to the faceplate, thus automatically getting the threads square to the faceplate.

Normally, you would machine the contact surface, the threads, and the inside relief in one setup, so they'd have to be correct. But here, you have to pick up something that you figure is a reference for the contact surface plane.

If you want to be really fussy, (and why not?) indicate the contact surface before you face it down. if it runs out, shim the mounting to bring that in as well as possible.
I don't know if the presence of other faults shows that surface to be unreliable, but it and the chuck mount surface are all you have to go on as a ref.

Now you see why its better to get a (cheap, $13) unmachined plate and make your own.

03-01-2004, 09:06 PM
I have sent you a detailed email with pictures. let me know if this is your problem.


03-02-2004, 01:12 AM
I think you can see the photo at the attached link.


The register section of the back plate is about 1/2" too long. The plate hits the spindle shoulder too soon.

J Tiers: Maybe I don't understand something, but if you indicate the contact surface prior to turning and then shim to bring it in - wouldn't you loose your concentricity? I can understand indicating after you've cut it down... Other than this, what your saying makes sense to me.

Thrud: I am lacking the skill set & the tool set to scrape this. As a matter of fact, although I've had my equipment for several months, I've just cut my first thread yesterday! I'm determined, but slow. Are you in aggreement with J Tiers for a workable solution?

Thank you gentlemen.

J Tiers
03-02-2004, 01:30 AM
<font face="Verdana, Arial" size="2">Originally posted by frankc64:
I think you can see the photo at the attached link.


J Tiers: Maybe I don't understand something, but if you indicate the contact surface prior to turning and then shim to bring it in - wouldn't you loose your concentricity? I can understand indicating after you've cut it down... Other than this, what your saying makes sense to me.
Are you in aggreement with J Tiers for a workable solution?

Thank you gentlemen.</font>

I only suggested that you might want to assume that of all the surfaces, the spindle contact surface is most likely to be right, as it should have been machined with the threads at one setup.

If you set on faceplate according to the CHUCK CONTACT SURFACE and the SPINDLE CONTACT SURFACE is then out-of true, I would assume the spindle contact surface is probably right, and shim to bring it in.

You can always skim the chuck contact surface to bring it in, which is part of the sequence anyway.

That way you should have the best chance of preserving the thread axis and spindle contact surface at right angles as they should be.

You don't want the spindle contact and the threads to "fight" each other, as it may likely not repaet well that way.

That's all.

03-02-2004, 03:32 AM
I might be missing something here, but I would be tempted to mount a chuck that fits, chuck a piece of rod, turn a section of it to clean and true it up, then carefully mount the four jaw on this stub, using the four jaws. Carefully center the backplate threads, using the four jaws to get this right. This should also coincide with the chuck body running true. Once this happy state is achieved, trim the excess backplate to suit. If needed, bore a section towards the threads to insert a shim ring, which is then bored to a close fit over the unthreaded portion of the lathe spindle. This might be a good idea if there's already too much play between these surfaces. Obviously some measurements have to be done first to determine the order in which things are done. If the shim ring is needed, it's outer diameter will have to be turned before the chuck is chucked, so that boring the backplate will result in a tight press fit for this ring, which is then installed without removing the chuck from the lathe.
I have not actually done this procedure, but I think it would work.
I guess the only things I would add are that it's important that the four-jaw chuck is clean and assembled properly, including the backplate, and that any play adjustments for the jaws be snugged up before it's mounted on the stub.
A danger would be if the thread bore is wobbling to some extent, and it can't be centered over the length of the threads. This unhappy situation would put an uneven force on the spindle threads, and the risk would be snapping the threaded portion of the spindle off.
I'm sure someone will correct me if this procedure has serious flaws.

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

03-02-2004, 04:15 AM
I think Darryl has it right. That is what I would do. If the chuck is any good at all the OD of the chuck body will be concentric with the mount. This is assuming that it was factory mounted. Maybe not in China...

The other possibility is to remove the backplate from the chuck and screw it on the spindle backwards with a carefully made spacer. Indicate to make sure it is true. Then turn it down to the required thickness.

03-02-2004, 04:37 PM
I think I would do it on an arbor between centers, better support for the spinning chuck and more likely to hold accuracy of the set up.

J Tiers
03-02-2004, 07:44 PM
Darryl et al sparked another idea, which does not require any other equipment except the ability to turn between centers. You would NOT put the chuck on it, and you would have positive alinement.

That way is to turn up an arbor between centers that has the backplate thread on it, a close fitting thread. You would put the backplate on it from the non-interfering direction.

You could put a stop ring or pin on your arbor to hold the plate against spinning on the threads. But you might be better off to use a little releasable locktite, so that the threads are the only alinement. That way no other forces get into the picture to throw it off if something isn't faced flat.

You could rough turn the contact area of the backplate down to close to where you need it, on a different setup, and spare the setup rough handling. Or do it all right from the arbor, depending.

Face the back contact surface to where you need it. Since you are referenceing the backplate threads to other threads which were turned between centers,you know its right with them and so right on axis.

Now you can screw it on the spindle and finish the job on the chuck contact area as it should be, and you will know nothing is out of kilter.

03-02-2004, 11:28 PM
Ok, I get it. Maybe this weekend I'll get going on this.

Thank you very much!


03-03-2004, 04:33 AM
I sent you details on how to fix it - you do not have to scrape it in - that is an option for us anal retentive types! http://bbs.homeshopmachinist.net//biggrin.gif