View Full Version : Continuing chuck woes

01-05-2005, 12:35 AM
So my wacked chuck http://bbs.homeshopmachinist.net//smile.gif I put away and am now using the Bison which is much better but I'm trying to figure out if it's bad too or there is something else wrong. Understand that I am a novice so it could be something I'm doing that might be causing it. It's a 5" threaded chuck with 1.5" x 8 threads and screws on the spindle. The spindle runout at the shoulder measures .0005" as does the spindle itself. The chuck is a loose fit until it snugs up to the shoulder. When indicating with a DI using the outside of the chuck, pointing the DI down I get .005 and if I chuck some drill rod in the chuck the rod is out .007 (more the further out I go). Any ideas? When looking at the chuck spin it actually looks like it is spinning ecentrically around the spindle axis. I'm wondering if I take the chuck apart and re machine the backplate to my lathe that it might take care of the problem. Looking for a list of things to check. Spindle bearings seem OK to me; nothing moves when chucked and I feel no play. I do feel radial play, as in bearing backlash but not back to front or up and down play. Logan 10"


[This message has been edited by kenrinc (edited 01-05-2005).]

[This message has been edited by kenrinc (edited 01-05-2005).]

01-05-2005, 06:07 AM
Is the Bison chuck new? By loose, do you mean easily threaded onto the spindle, or a sloppy fit? Is the .0005" R.O. on the spindle with or without the chuck mounted? Wether or not the chuck is mounted, that R.O. may vary. Give everything a good look before you mount your chuck. Make sure there are no burrs, chips, etc. on any surfaces. Check the actual thread O.D. (major dia.) on the spindle. Compare that to the actual I.D., (mininum I.D.) on your chuck. If possible, check the actual thread depth of each also. "Loose until it snugs up to the shoulder" and "spinning eccentrically" to me, indicates too much thread clearance. If that is the case, I would start with a new, blank backplate and fit that to better, and more closely, fit the spindle/chuck.

J Tiers
01-05-2005, 07:56 AM
OK, its a threaded chuck, right? So no backplate?

I assume you have an indicator. Find and mark the "high spot".

Take off the chuck, noting the spindle rotational position. Give the spindle a part of a turn, note new position, then put the chuck back on.

Find and mark the NEW high spot.

Repeat twice more, being sure to have the spindle in a new position each time as you put the chuck on. That is so displacement due to chuck weight is different each time.

If the high spot is random, you have a looseness problem. You cannot be sure what the high spot will be. There is no fix, because the chuck itself is threaded.

If the high spots are concentrated in the same quadrant of the chuck, you may have a thread, nose "register" (unthreaded area), spindle, crud, or chuck problem.

I don't believe much in the "register" theory of alinement, but it IS possible for the thread and "register" to be non-concentric. That will cock it off line if the fit is snug.

01-05-2005, 08:20 AM
One of the "Shop Wisdom" books available on this site, I think the one by Frank Maclean(spell), has a great section on chucks, how to test and correct and how to avoid them in the future. The article is titled A FIRE SIDE CHAT ABOUT CHUCKS. In fact I recomend this book highly. It is filled with great info and projects that start out where books like South Bends HOW TO RUN A LATHE end. You might want to get it.

When I fit a new threaded mount chuck to a lathe I try to get a very close fit on the threads and then check the register contact. This book will explain the whole process.

Good luck,

Forrest Addy
01-05-2005, 09:22 AM
Loose threads will not cause chuck run-out by itself. Threaded chuch have two registration surfaces: the shoulder and the load flank of the thread. When the chuck "shoulders" it automatically rides up the thread flanks and centers itself. It's not the best chuck spindle mount but it's damn accurate when clean.

If there's a little piece of adherant dirt or a chip on the thread flank it can throw the chuck off center. I suggest you clean out the chuck thread with a toothbrush (a machinist's best friend around the lathe) and a little solvent. If that doesn't work go over the thread flanks with a hooked scribe and a dental mirror. Pick off anything that looks doubtful. Clean the threads to bare metal, oil lightly, and try again.

If that doesn't solve your problem we'll go to Plan "B".

01-05-2005, 10:09 AM
Don't indicate anything off the body of the chuck - get a gage pin (spend $20) or buy a ground and polished carbide rod and check it on a granite plate to make sure it is straight FIRST. Then you can gently chuck it in the jaws and indicate off the carbide rod or gage pin CLOSE TO THE JAWS. Follow Forrest's instruction on cleaning the threads - you should always do this prior to installing a chuck or a taper (clean both and check for burrs).

If the chuck is mounted on a backplate I am assuming that you just bolted it on and started using it - you can do that, but you are not going to get accurate results - the backplate must be carefully fitted to the chuck and the lathe spindle for maximum accuracy!

01-05-2005, 10:30 AM
Thanks all! I'll check this stuff tonight when I get back. The chuck is brand new, 5" 3-jaw, and came threaded for a 1-1/2 x 8 spindle. I just took it out of the box, cleaned the threads and installed it. So far I seem to be getting the .005" runout whenever I put it on the lathe. I have figured out that whatever I make in the chuck does turnout OK but it seems that for $200 I should expect a bit more. Or is this not the case? I realize that the problem is compounded by lack of knowledge.

I'm assuming that a threaded chuck means that it's just a regular "plain back" chuck mated to a backplate at the factory and if this is the case should I expect this amount of runout? What is ironic is that I "wrestled" with the decision on if I should buy a plain back chuck and a backplate blank and machine it to fit my lathe or just buy the pre-threaded chuck. I choose the latter. I have no problem taking the new chuck apart and re-machining the backplate as I am eager to start a "real" project but I want to make sure it's a solution.

Be back soon....


J Tiers
01-05-2005, 11:39 AM
<font face="Verdana, Arial" size="2">Originally posted by Forrest Addy:
Loose threads will not cause chuck run-out by itself. Threaded chuch have two registration surfaces: the shoulder and the load flank of the thread. When the chuck "shoulders" it automatically rides up the thread flanks and centers itself. </font>

Forrest.... I agree, except for a detail....which could be part of Ken's problem.

If the threads are loose, it is possible for the chuck to initially seat jammed in an off center position. It likely will not STAY there, under vibration and cutting forces, but it can START there, because the chuck "hangs" off-center on the threads before it seats. The off-center amount depends on how loose the threads are.

Since Ken is not applying much force, beyond the normal seating force, his chuck may be in that "starting" position.

That will be fairly random in amount and direction, depending on the position of the spindle when the chuck is mounted, seating force, etc.
If there is un-seen crud, etc, it will likely be in the same place repeatedly, regardless of spindle position.

That is the reason for checking several times. (of course the crud, if any, might get knocked off, too, which is just as good a result....)

If the unthreaded area is tight but not concentric, that can also cause a problem, solved by cutting it back....

[This message has been edited by J Tiers (edited 01-05-2005).]

01-05-2005, 12:20 PM
sounds to me like a bent piece of drill rod.
chuck up a known straight and then check your TIR. I wouldn't go machining my backplate just yet.

01-05-2005, 11:17 PM
OK. So if I do as J Tiers says, I spin the chuck on and made a mark on both the spindle bearing shield and the back of the chuck that conincided with the high spot. Almost .006" out. If I lock the spindle with the back gears, remove the chuck, un lock the spindle, move foward a 1/4 turn or so, lock the spindle and put the chuck back on the .006" is still there in the same place I marked, both on the spindle face and the chuck. Did this 3 times with the same result measuring botht the outside of the chuck and the backplate:


Now I did find a little burr on the very end of the screw threads of the spindle. At least I think it's a burr. I can't actually feel anything but here is a closer look. I think I can scrape it off with a 3 sided scraper if I'm careful:



[This message has been edited by kenrinc (edited 01-06-2005).]

Rich Carlstedt
01-05-2005, 11:21 PM
You said
"The spindle runout at the shoulder measures .0005" as does the spindle itself. "

You should have "0" runout on both surfaces.
sounds like the spindle was bent when you crashed...
any chuck mounted will reflect the angularity of the error as a larger error.

01-05-2005, 11:27 PM
put the chuck backplate onto the spindle withouth the chuck screwed to it, face the mating surface of the backplate where the chuck mounts to. final few facing cuts, take off less than .001.

Now, if your spindle is like my South bends, make sure the collar on the rear of the spindle is properly set, so the spindle does not move in or out of the headstock or you will get waves in your facing cuts. I know this from expirience http://bbs.homeshopmachinist.net//smile.gif
Then bolt the chuck back onto the adapter and test the runout near the chuck, then away from the chuck. My bison 4" always got between .002 to .003 runout.

The More I use the 3 jaw, the more I like the 4 jaw.

J Tiers
01-06-2005, 12:12 AM
If there is a burr there, it conceivably could cause a problem.

I admit, I couldn't quite figure out how a bent spindle could cause radial runout right at the chuck back on a screw-on chuck.

But that chuck has a goodly amount of "stick-out" due to the threads being in a 'spigot" extension at the back.

Also...you have called that a "screw-on" threaded chuck.

But are you SURE that REALLY IS a screw-on chuck? Because the split of a chuck usually is through the tightening pinion holes.....but there is a visible ring that looks like a split back towards the threaded part, NOT at the pinion area. That looks a lot like a backplate on a regular plain-back chuck. The split in the chuck body is often invisble in a new chuck.....I bet there is a back plate...the screws in the back say it may be one....

if so, there are lots more things you can do. I would start by getting that chuck loosened up, and checking the part of the backplate that should fit into the chuck recess to hold it concentric.

Dollars to brass chips that the parts that the chuck seats on, and the part that fits into the chuck recess are not true or concentric, due to being turned in a bad setup and not with the plate mounted by its threads.

You should be able to correct that fairly easily.

But I hate mounts that stick out like that...they look wrong, and add potential for error.

BTW, I would be surprised if your spindle measured dead-on perfect....it's a commercial product, not originally made to Monarch type specs, and it isn't new either. The errors may be in teh bearings, also. Those are also commercial products.....not the highest precision originally. The lathe wasn't that expensive back in the day.

But it works fine, at least mine does (now).

Forrest Addy
01-06-2005, 12:22 AM
My personal favorite proof cylinder is a nearly new automotive wrist pin. They are generally within 50 millionths of a perfect cylinder.

Drill rod is usually centerless ground to a close tolerence but it's seldom very straight.

01-06-2005, 01:05 AM
For what it's worth, I always keep the weight of the chuck in my hand, while turning the spindle into it with the other hand. I am always careful to let the chuck adjust itself to position while it becomes tight to the spindle, before I take my lifting force away. Then I tighten it to the spindle the appropriate amount. That way the threads don't have to bear the weight of the chuck as they center it.
I don't think that's your problem, though, Ken.

I agree, there should be virtually no runout measureable on the shoulder or register surfaces. If there's any axial bearing play (lengthwise of the spindle) this could be temporarily removed by pressing from the tailstock using a live center agaonst a plug inserted into the spindle bore. It might help to leave the tailstock clamp a bit loose for this. That way it has less chance of throwing the spindle around a bit. Recheck the runout after making some preload like that. If nothing changes, maybe try a morse taper in the spindle bore. I'm assuming that's what would be there, in this case of a threaded nose. Indicate on the extended end of the tapered adapter you have mounted. As you insert the morse taper, feel for any rocking motion that might indicate a warped spindle. The adapter should go in and seat firmly, without play, and without requiring force. It will be good enough to indicate from without using a drawbar. Add to the test then by using the drawbar. Note any changes. If the taper needs a considerable force to stop it rocking, then the spindle could well be bent or warped. If that does seem to be the case, and replacing the spindle isn't an option, then go the new backplate route. It will likely reseat repeatably, and if machined well, it will continue to run true. Make sure that the new backplate threads on properly and fully without becoming too tight before the mating surfaces come into contact. That's been a problem with some threaded adapters I've bought.

Do remove the backplate from your chuck and check the runout on it mounted alone. If possible, have a second backplate on hand and check it on the spindle. If the runout relates directly to spindle rotational position with either of the backplates or the shoulder and register surfaces of the spindle itself, then something's bent.

01-06-2005, 01:53 AM
I marked the spindle where the chuck was out the most. It turns out I have a faceplate that I use and it too is .005" at the same place as the 5" 3 jaw. The 4 jaw is similarly out but not as far but at the same place. Interestingly enough I have a collet holder that threads on the spindle nose and indicating on that shows .0005"; same as spindle. Maybe I'm not indicating correctly? To me .0005" spindle runout seams OK for a 60 year old lathe. Am I high? What do you guys get on yours?

I just got the lathe, so I didn't put any tools into the chucks (yet :-) ) Meaning I didn't run the toolpost into the headstock. But a former owner certainly could have. But like I said the bearing cover was thrashed but was tapped down and is not causing any issues. I feel no play at all in the spindle other than play from rocking the spindle back and forth in it's bearings (rotational). No up and down, no front to back motion. Dead solid.

J Tiers: Yeah the chuck looks exactly like a typical plain back with a backplate just that they did the work instead of me ;-)

Going to take apart tomorrow night and see what else I can find.


[This message has been edited by kenrinc (edited 01-06-2005).]

[This message has been edited by kenrinc (edited 01-06-2005).]

J Tiers
01-06-2005, 08:10 AM
If you have a consistent 5 thou on mounted dtuff on OUTSIDE, but 1/10 that (half thou) on inside, then there is some sort of interference or problem with the OD.

I thought you said the OD was at the half thou also? Sounds like anything mounted is out the 5 thou in the same place.

Remember, a 5 thou movement (5 thou TIR) is 2.5 thou of problem, giving a + on one side, and - on the other.

If it is that consistent, there must be a ding or problem somewhere on the threads that sets your stuff 2.5 thou off of concentric.

I don't recall if you said you had a "cam" issue with the nose....if you put indicator on flat part of shoulder of spindle nose, does it give a high spot?

Did you measure your half thou on the unthreaded portion of the nose, AND on the inside of the spindle? If so, and if the collet is true, then it seems difficult to deduce a bent spindle. Those spindles are pretty stout, it isn't a "109".

I think you need to find the "ding" or crud on teh nose of the spindle.....

01-06-2005, 08:35 AM
What darryl said ... take the weight off the chuck while tightening it.

Another trick you can try is to wrap teflon pipe tape around the threads until the threading is more snug. This won't be perfect but will help you to find out if the thread slop is the problem by temporarily removing it. Den

01-06-2005, 10:57 AM
Yes, I measured the unthreaded portion, the shoulder, the front nose perpendicular to the spindle and the inside of the spindle taper and all are .0005" or less. I don't think it's the spindle either. If I look down through the center of the spindle from the gear end, while it's spinning, I see no movement, almost as if it's stopped. But with chucks I can easily see that they are not spinning concentric. I will continue on the thread path. With the 4 jaw I can easily adjust the runout out so it's moot and that's how I've been working up to this point but I've got a lot of other tools for the lathe that I'd like to use.

I always support my chucks with my left hand when spinning them on because their so damn heavy I'm always afraid they'll damage a thread and god forbid drop on the ways! FWIW I also tried pushing up on the chuck as I threaded it in order take up the slack and it always reseats itself to the same place regardless. The thread tape sounds like an interesting try. The burr may have been there but I can't snag anything on it and I have chased all the threads on all the chucks and the spindle multiple times with various tools starting with a dental pick and cleaned with solvent.

I need a magnifying class or some of those mag glasses.