lathe chuck

[John D Smith]

I have a Jet 9n20 Jet metal lathe that has a chuck that reads about 15 to 20 thousands off of center. Took out the jaws and swapped them around. No luck. I am thinking I may have to grind some material off of one of the jaws. Have any one got a better solution?

 

[Bob]

A few years ago, I bought a VERY slightly used Horror Fraught lathe that was declared as surplus to his needs by the guy's new wife to be.

I had the same issue with the 3-jaw "scroll" chuck, and FINALLY found the right combination of jaws and slots and scroll position when inserting the jaws that it's now pretty darned true. (Remember that if any of those factors are changed from when it was machined at the factory it will be inaccurate.

Then, I marked them, for future reference.

Jet stuff is pretty good, and I'd bet it was better than that when it left the factory, and unless someone swapped a jaw or two with another chuck, or it was somehow damaged, I'd think it should be possible to figure out the puzzle.

As you have probably figured out, if you have to grind the jaws "true" they have to be secured OUTWARD against the scroll if it's an "external" chuck, or inward against the scroll if an "internal" chuck while grinding.

 

[leroy in ne]

Check if this is a plain back chuck that the body is movable on a mounting plate attached to the lathe spindle. Some chucks were of this type and have to be indicated in on the machine to repeat. My description is not good, I hope you can figure out what I mean. Leroy

 

[Butch(OH)]

As 3 jaw chick will only be right when the jaws are in the same slots they were in when it was trued up at the factory after assembly. Assuming you have tried every combination of jaws and slots the only way to get it right is with a truing ring and a tool post grinder. It would take a very good tool and die maker a very long time and lots of trial and error to true one by hand. Another thing that most dont realize is which key hole you use affects the accuracy of the chuck. For no more than a new chuck that size costs you might be ahead to go that route. Hindsight, always match mark jaw locations prior to dissasembly and experiment with key location that gives the least runout and mark that hole. Repeatable .005 should be obtainable with the cheapest of cheap chucks. .001 with a good one and thats over a range of diameters.

 

[Butch(OH)]

Good grief I did proof read that post before I hit submit,, and I havnt been drinking,, sorry

 

[flembo]

Most lathe chucks have numbers on the slots and numbers on the jaws and they need to be installed correctly. Also quite often a chuck is mounted on a backing plate as Leroy mentioned, the better ones have 4 or more adjusting screws so you can center them. It sounds like you already messed with jaw location so I would look close for any numbers on the jaws and on the chuck body, if you find them then start with number 1 in slot 1 until it catches then #2 then #3. Good luck, BTW grinding jaws would be my last choice but it can be done as Bob mentioned. I have done it with a tool post grinder which I am sure you don't have but maybe a dremel tool could be securely fastened to the tool post. Again good luck.

 

[Janicholson]

One of my lab Clausings has a (Non Clausing) chuck that was .013 out. as best assembled. I put an accurately ground ring into the back of jaws to tension them outward. Then used a vise grip 2 prong welding clamp to secure a die grinder to the tool post using a flat bar in the post to clamp to. It took about an hour to grind the jaws as deep as I could go (1 inch cylindrical stone). I then removed the jaws and removed the high spots remaining where the ring was holding the jaws outward. It works very well. I ran the lathe in back gear forward rotation, and the grinder on the operator sode of the opening. Jim

 

[RedRoundhead]

How does the OD of the chuck body run? Always good to check that too. Could be a mounting error between the chuck and the plate or the spindle taper/mount.

I would check the face run out of the chuck as well. Can help find issues.

If those are good, you have to follow the other advice - swap jaws around, etc.

BTW, is this a 3 jaw scroll chuck, or a 4-jaw independent chuck?

How are you checking run out, btw?

 

[PJH]

I have a collet chuck that has just a little bit of runout. I'm not a machinist, but I'm pretty sure the backplate thread needs just a little bit of material removed. The thread appears to bottom out before the backplate seats against the lathe spindle. I'm trying to work my skill ability up to the task. Not ready yet.

 

[JF in MI]

Did same on my Clausing. Came out to just under .003".
Like this?

 

[Russ from MN]

When I went through my machinist training the 3-jaw was adjustable, it could be dialed in. I made a nice pulley for the back of our JD M, used the 3-jaw for everything.

 

[flembo]

LOL. I had to reread your post to see the error. I have heard rumors of chicks with 3 breasts but not 3 jaws, you couldn't get a word in edgewise lol.

 

[John D Smith]

Thanks guys, I will go back and redo the jaw swapping and clean the front face to see about marks. I would grind the jaws as a last resort.

 

[John D Smith]

I apologize for not getting the info on the chuck. I measured the chuck hosing and I couldn't believe my eyes, I runs just under .001!!! Now the jaws, I cleaned them up with 0000 steel wool and found inside of the slot that slides into the chuck body the factory numbers. I have a round piece of steel that can off a lawn tractor. it was the rear axles. Before the jaws were in the right slots the best reading I could get was about 15 thousands. now I am down to about 7 thousands. Still not good enough. With the chuck body that close to perfect I should be happy if I can get it down to about 3 thousands. So that's where I am at. One other thing I measured the run out of the axle rood one inch out from the jaws. I am trying to one change per setup so I can keep the mistakes to down.

 

[Bob - MI]

When you are measuring the runout I am guessing that you are measuring the free-state condition with nothing in the jaws?

Ideally the runout should be measured in the clamped state to be done correctly. I realize this sounds stupid or questionable but as the jaws close down the variation in the scroll as it engages the grooves in the back of each jaw the alignment can shift a bit. All said it's not what you get in the free state it's what you have when tightened up that matters.

Precision shops will many times use soft jaws that are bored at each setup to assure the best case control of run out. Guessing that your jaws are hardened they have to be ground which gets tricky but it can be done by mounting a die grinder on the cross slide and slowly working in and out while the chuck spins.

My experience comes from younger days running a large turret lathe and an ID grinder in a Gardner Denver plant that made air tools. The ID grinder was the one that tolerances were really tight on and the jaws had to be accurately sized and all ground at the same time. Concentricity of the ID to the OD was often in the 0.0001 range so there was no room for error. The lathes had a lot more forgiveness but we still looked for runout in the 0.002 range.

If you need help figuring out how to put tension on them post back and I will try to help you.

 

[1948CaseVAI]

Once you start grinding a check it is done. Learn how to adjust it. You tube is your friend. Watch Adam Booth (abom79). He has dialed in several chucks over the years.

 

[Oliver 500 Wa]

Most lathe chucks have numbers on the slots and numbers on the jaws and they need to be installed correctly. ......

One of my cheapo Chinese 3 jaw chucks does not have numbers on the body or jaws, but just has small punch marks to show which jaw goes in which slot. Had to look it over pretty closely to spot them when I first got it.