View Full Version : Making a chuck plate "101"

02-04-2003, 10:05 PM
I read the previous responses regarding back plates but still need help. I would like to make a back plate for a 6" 3 jaw chuck for my 9" SB lathe. I have a 6" 4 jaw chuck on the lathe now. Can I use this chuck to make the plate? I saw a reference to "register" and what does this refer to? I was going to use a 6" x 5/8" bar bell weight - will this be adequate? What are the critical machining operations needed to get a functional plate?
I would appreciate any help or references you can give me. Thanks Paul

Mike Burdick
02-04-2003, 10:54 PM

The register is the area on the spindle nose that has no threads and includes the flange that the chuck's backplate will fit up next to. In my opinion the tighter the fit the better the chucks repeatability will be. The register is supposed to line everything up.

About using a barbell for the backplate: I tried that but found the material to be hard in some places and very soft in others, I also found holes in the casting. I would think that barbells are made from the cheapest of metal so I don't know if I would trust it. The possibility of a chuck flying across the room will not be something you would enjoy! Also, the 5/8" thickness is way to thin for a safe plate.

I too have a 9" SB lathe so here's how I made mine: (remember I'm not an expert)

First machine an exact duplicate of the spindle nose (threads and register) of the Southbend. This will be used for a "gage" for the new backplate.

Then get a piece of flat steel (4140 annealed would be nice) that is at least 6" X 6" and 1.25" thick, and put it in the 4 jaw chuck. Bore and thread this steel such that your spindle "gage" has a nice tight fit. Remember, the register will also be bored at this setup (you're making this piece backwards from fitting on the lathe). Also face the entire piece of metal. I repeat, do not remove or adjust this piece of material while it is in the chuck until the spindle fit is acceptable and material faced. This is needed to maintain accuracy (that's why the "gage" of your spindle was made).

Now remove the material from chuck and take a band saw and remove corners. Replace the 4 jaw chuck with this "new" backplate on the spindle and finish the side and face to suit your chuck.

I think using this method will ensure that you will have the maximum accuracy that your spindle will allow.

By the way, you don't have to use the square piece of steel- a round would be better- but I did not have a round that big in my supply.

This was really a fun project for me - I think you will enjoy it too.....Mike

[This message has been edited by Mike Burdick (edited 02-05-2003).]

02-05-2003, 12:32 AM

For a threaded spindle there are three key features -
a) the thread
b) the area immediately behind the thread - "register"
c) the radial flange or shoulder that the backplate runs up against

The register centers the backplate and is normally the same as the major diameter of the threads themselves. The shoulder has two functions - keeping the chuck perindicular to the spindle and driving the chuck.

Therefore for maximum accuracy the threads must be very close fit to the backplate. The inner diameter of the backplate and the rear surface can be carefully scraped for a precision fit. When this is done, the chuck will always return to the same relative position on the spindle giving maximum consistancy and accuracy. It is important that inside and outside corners (where appropriate) be chamfered to properly clear mating part corner radius' (or lack thereof). DO NOT SCRAPE THE SPINDLE - only the backplate! Sometimes the register and flange must be turned and re-faced to facilitate maximum accuracy, but it should only be done if absolutely no other choice exisits (last resort). Threaded holes should also be counterbored a couple threads to prevent pull out of the threads above the surface - this can cause accuracy problems if a bur is raised by tightening a bolt.

Paying attention to the details here will gve superior results. I also make withness marks to line up backplates with the chucks and number bolts to their respective holes incase it is disassembled by others.

[This message has been edited by Thrud (edited 02-05-2003).]

Herb Helbig
02-05-2003, 12:25 PM
Does it make sense to produce the dummy spindle nose that Mike Burdick describes by making a rubber mold and casting - like the dentist does?


02-05-2003, 01:24 PM
Mike Burdick:

"Then get a piece of flat steel (4140 annealed would be nice) that is at least 6" X 6" and 1.25" thick,..."

What do you pay for a piece of 4140 6" x 6" 1.25" thick? I puchased a piece of 1018 1 inch thick cut off of a 6.25" diameter round for about $15. Is the 4140 less expensive? Do you really need 4140 for a backplate?

Mike Burdick
02-05-2003, 03:27 PM

The 4140 annealed would not be required for a backplate and it is more expensive than 1018. I would imagine that any steel would work, but some are easier to work with than others. I used 4140 for mine because it's nice to work with and seems to be very stable (lack of internal stresses due to working) and more importantly, I had some!

I'm definitely not an expert, I just expressed how I made my backplates with my meager equipment.


[This message has been edited by Mike Burdick (edited 02-05-2003).]

02-06-2003, 12:47 AM

No, but it makes sense to make a metal copy and a female for making copies of your spindle for things like dividing heads. The rubber repro is ok for metrology, but not fitting and scraping backplates to your spindle.

When you make the copy of the spindle make sure you include the rear shoulder and the spindle register. It is easy to make errors on accessories that attach to the spindle if you do not - trust me, been there, done that.

02-06-2003, 12:35 PM
We are lazy down here in Texas because it is HOT for most of the year. I am too lazy to make a spindle adaptor, so I bought one on Ebay!!! Called a "GO gage" - it is machined much more accurately than I could ever do, and it fits the factory backplate of my 2-1/4" - 8 Sheldon like butter. Cost $15 + postage. MSC has them new for about $200. These are taken good care of usually (in someone's metrology dept), so it is easy to find one in good condition. It does not have the shoulder register, so you would have to add that if you need one for fitting purposes. I still need the plug gage for the bore, so this is an admission against my own interest!! Don't y'all go bidding on that GO bore gage when one comes up for bid, y'hear!! A.T.

02-07-2003, 04:46 PM
Here's my 2 cents worth. Go ahead and make an exact duplicate of your lathe's spindle nose. It's great practice, and will come in usefull in the future. Buy a raw cast iron blank and machine it to fit. Pay particular care to the register portion and scape in the mating surface to the spindle.

In searching the web I found widely varying prices for chuck backing plates. Victor Machinery (www.victornet.com) has a raw cast iron back plate that'll fit a SB 9" for $12.50. One already machined starting at $41.00. Not bad either way.


Herb Helbig
02-07-2003, 11:16 PM
Thanks, Thrud, for the advice about casting a copy of lathe spindle nose.

Spinrow asked a question I've been thinking of, too. And I didn't notice the answer. Could I use the body of the crummy 4-jaw chuck that JET supplied with my 9 x 20 lathe to make a back plate for a good Bison chuck?


02-08-2003, 12:52 AM
Using the mounting of another chuck might be a good shortcut, especially if the mounting elements (threads, register or taper, etc.) are in good condition and there is sufficient material so that you can machine it to fit.

I recently picked up some worn out L0 chucks for a few bucks each. I have separated the backs from them and they appear to have enough meat to them that I can face them and use them to mount flat back chucks.

Don Kinzer
Portland, OR

Rich Carlstedt
02-09-2003, 01:32 AM
Well, here goes !!!
There is no such thing as a register !
The counterbore relief area is just that , a relief area.
The "critical areas are the threads, ie pitch diameter and distance (TPI), AND the face at the rear of the back, that stops against the spindle. This surface MUST be square to the threads ( Read: made at same time as threads)

The reason is that a thread is a conical helix and registers itself, and the "stop" faec must be square to prevent sideload.

02-09-2003, 02:08 AM
That is one of the theories of chuck accuracy in relation to screw mounts - it has never been proven however. All three in unison produce more accurate and consistant mounting of the chuck. In theory it is posible to acheive the same accuracy with just the thread and shoulder but I have never seen it done.

You can if you like, but a high quality bison cast iron backplate is only about $40 - so why bother with the junk?

Herb Helbig
02-09-2003, 07:28 AM
Thrud -
I couldn't see such a thing in my MSC catalog. Where to buy?


02-09-2003, 09:07 AM
Since every lathe and chuck manufacturer that I am aware of provides a register finished to very close dimensions, I must believe there is such a thing.
There must be some play in threads to allow them to tighten up. The register prevents the back plate from squirming on the threads when final tightening is done.
All three items, threads, register and squareness of shoulder are important in proper fitting of chuck, and the closer attention is payed to each, the better the job will result.

02-09-2003, 11:33 AM
Hey, I'm with Rich.

The "register" is a darn good way of alining the chuck so the threads will hold it right.

Other than that, it has one function....staying out of the way. If the register is out of position, and close-fitted, you have a problem. The threads and shoulder can't fix the alinement.

Let's say it is made 0.0005 under the chuck dimension. Good alinement, but still a small eccentricity.

But, if it wears at all, the eccentricity gets bigger right away. Who in their right mind uses a straight sided fit for a precision alinement?

If it were TAPERED, now, you betcha. Then if it wears straight down, the alinement does not change as much.

You go ahead and loosen up the threads on yours, keeping the "register" tight, and we will see which one lasts 50 years.

I have seen loose registers on backplates and chucks aline just fine, and repeat just fine. I mean 0.050 oversize on the backplate, NO contact. If guided a bit, they repeat even better, although they will eventually tighten into a fixed known position either way.

I have seen somewhere a spindle that had a "register" at the nose AND back at the shoulder. Supposed to aline better as you can't cock the chuck as much putting it on. Still wasn't made to 0.0005 fit, though. Just an alinement assistant.

So, make it close, and you will have less trouble putting on the chuck. But don't obsess about it, spend that effort getting the alinement of threads and rear surface exactly right.

[This message has been edited by Oso (edited 02-09-2003).]

Rich Carlstedt
02-12-2003, 11:13 PM
My 6 Jaw Buck chuck holds to tenths when removed and replaced. It has a .025 clearance on the "register" dimension.
And I have had it for 28 years.
When I first got it I checked out my other threaded chucks and the clearance ranged from .010 to .030 so I looked at it closer as I thought it was important and would return the chuck to Buck if it was out. Well it doesn't change.

I want you guys to do this to check out what i am saying.
Take a 6-32 screw and put it into a 8-32 nut.
It will be very loose but it works.
Now square the nut to the thread and push it sideways, then pull on the thread and note, that no matter how much side load you put on the thread, it will stay centered as long as force is exerted by the screw. Thats because the helix is engaged and the thread angle is doing the centering.
Each time you put a chuck on , it finds its "Axel center" and that gives you repeatablity.
The only factor distorting this function is a non-perpendicular stop face, SO the face must be done at the same time as the threads when doing a back plate!

Have Fun

02-12-2003, 11:58 PM
You are both right and wrong. I disagree with you on the side of "not taking chances" - however, I will concede your point with a caveat. There is a problem with your theory. If the threads are worn, it cannot locate the chuck properly, but it can force it against the shoulder and "register". Both will insure accurate positioning of the chuck. Assuming that the cone section will force it into proper alignment is poor engineering. It will of course, with a in spec spindle and backplate. This is why it is so important to measure the thread, bore, and thread the backplate to the specific lathe. If you have no concern for accuracy, buy a prethreaded backplate or a chuck with integral mount.

A further step is to scrape the backplate flat where it contacts the shoulder of the spindle so it mates perfectly with the spindle. This will insure proper angular positioning of the chuck in relation to the spindle provided the threads of the spindle and backplate do not wear (this is an important relationship to note). The backplate can be checked for parallelism after the facing cut for the chuck is made - this will also tell you if your cross slide is out of whack (concave or convex surface rather than flat - check on granite plate).

If you are smart, you avoid all this and get a D, L, or A spindle. It is an imperfect world.

02-13-2003, 12:57 AM
I may as well add my too sense worth here. When I mount a chuck, of any type, I carry it up to the spindle, guide it onto the threads, and -while still carrying the same weight with my hand- turn the spindle to pull the chuck up to the register. Once it's snug, I give it a further hand tightening. This procedure takes any weight off the threads, which makes for lower wear on the threads and the registration surfaces, and allows for better repeatability. Of course, now you know my lathe is not a behemoth, I can lift and control the chuck with one hand. Some of you out there must need a forklift to do this. I wonder how big and heavy chucks get to be?

02-13-2003, 02:14 AM
I must agree with Rich. Just checked the spindle of my little Logan, it mikes 1.498". The ID of my 3 jaw Buck adapter plate is 1.530" . This chuck always reseats itself accurately. I've been using this same chuck and lathe combination for 20+ years.

Never heard of "Register" before, you fellows make that up?


02-13-2003, 02:57 AM
Today I received the two new (never used) 1-1/2x8 backplates I won on eBay for $15 each. I was holding out to buy unfinished backplates at about $40 each, but I figured I'd take a gamble on the threaded ones because of price. I cleaned off the cosmoline and mounted one of backplates on the lathe's spindle. I then marked the runout on the diameter and the face with a dial indicator and Sharpie. It was dead nuts on the face as far as my thousandths-reading indicator is concerned. It had .003" TIR on the diameter. I removed the backplate and then mounted it again. The face was dead nuts again, and the TIR was identical. I didn't bother measuring the bore of the backplate until after I concluded that it fit correctly. To my amazement, the bore is approximately 1.535", and the register of my spindle is approximately 1.499" So I conclude the thread is important, if not the only thing keeping the backplate straight in conjunction with the shoulder of the spindle.

02-14-2003, 01:02 AM
Anthing that forces an object into a particular orientation can be called a dowel (this would require two, of course), a key, a shoulder, guide, or a register (pilot). Register is more applicable to a shaft and sub assembly.

Perhaps calling it a Pilot or land would have caused less confusion - but that is not the way I think.

02-14-2003, 10:33 AM
Hi All, I am a self taught home shop machinist with limited experience, so please bear with me. Most of my knowledge has come from either reading or hands on combined with a little common sense. This thread (no pun intended) has been of particular interest to me as I have been planning some new backplates for my lathe chucks and was concerned about the fit of the "register" on my lathe. I've read the various opinions about centering due to either threads, register (or what ever your choice of terms is) or both and have come up with an "informed opinion" of my own based upon what I've read here and a couple of books; "How to Run A Lathe" by South Bend and "Advanced Machine Work" by Robert Smith. I now believe that the threads and shoulder are the two critical factors in repeatability. A case in point is the Headstock Spindle Chuck mentioned in How to Run A Lathe. Jacobs made many of these chucks to fit Logan, South Bend, Atlas and other lathes. They came ready to use, bored to the same internal diameter where they fit the "register" and were supposed to be accurate to within .002". Apparently it wasn't important that they be fitted to the spindle (register) of each individual lathe. The second case I point to is that in Advanced Machine Work there is a section on page 508 about finish tapping a back plate where it mentions "bore out about two threads... to permit screwing plate to shoulder on nose of spindle...". The book goes into detail about the accuracy of the threads, but glosses over the boring for the shoulder. I know these books are old, that what may have been considered close tolerance then might not be considered so today and that no book can replace the actual experiences of people actually doing the job, but these two examples seem to support the thread and shoulder theory.

[This message has been edited by firbikrhd1 (edited 02-14-2003).]

02-14-2003, 07:05 PM
The How to Run a Lathe book by Southbend gives spindle dimensions of 1-1/2"-8 threads for the spindle and diameter of 1.509" for the register.
Spindle nose length 7/8", thread length 11/16".
Dimensioning to three decimal points infers a degree of accuracy of +- 0.0005". Not giving a class fit for the threads infers lack of precision there.
I first learned of it being called register in an article by Frank Mc Clean 8 or 9 years ago, and have seen it referred to same several times after that. I don't have the issue handy, but it is also reprinted in Two Shop Masters.
Mc Clean uses the method described above.
Guy Lautard in TMBR 3 recommends making a plug gage the exact diameter of the register (his words also) to use in fitting a backplate. Also recommends use of a gage equal to the root diameter of the spindle threads and spindle nose copy. That's a total of three gages to fit backplate.
I feel that all three components are equally important in the fit of the backplate, and have seen enough other descriptions of the same method fitting a backplate to believe that that is the best approach to follow.

[This message has been edited by JCHannum (edited 02-14-2003).]

Mike Burdick
02-14-2003, 11:01 PM

Maybe the register has functions other than for repeatability. How about adding rigidity to the spindle-chuck connection?

Just throwing some fuel on the fire! http://bbs.homeshopmachinist.net//smile.gif


[This message has been edited by Mike Burdick (edited 02-14-2003).]

02-14-2003, 11:28 PM

A very good point. I never brought that up, because it goes hand in hand with accuracy when you consider a large mass spinning at high speed - possibly with a heavy part in its grip, and still expected to be consistant.

There is a very good reason modern precision lathes are NOT made with threaded spindle (unless by special order i.e. Hardinge) - because they they are too expensive to do properly. Hence D style Camlock, like VHS won out in the end. BTW the D series spindle is cheaper to make, it is not as accurate as the threaded, A or L spindles.

02-15-2003, 11:00 AM
I was directed to this board by Steve F., thanks.

The unthreaded portion of the spindle nose on a threaded spindle is NOT a register. If it were designed to locate the workholding device accurately, the fit would have to be so close that you would soon be unable to install or remove any devices.

As mentioned by someone else, the counterbore in the chuck back or face plate is just that, a counterbore for clearance. I does help to roughly center the device, so it should be fairly close, but with, say, 0.005" to 0.010" clearance.

The locating surfaces are the threads and the shoulder.

Some will probably call me a troll, since I have posted this, and will now fade into the background. Personally, I hate these web based message boards, preferring usenet and/or email based discussions. If anyone would like to contact me for further discussion, email me at ssl@loganact.com or visit the Logan Lathe Group at http://groups.yahoo.com/group/lathe-list.


Scott Logan
Logan Actuator Co.

02-15-2003, 09:30 PM
Hi again,
I hope no one minds, but with all this discussion and the varied opinions I thought it might be good to contact a lathe manufacturer, Scott Logan at Logan Actuator, to get his input on this issue. After all who might be more qualified to answer questions about a lathe than a manufacturer? I expected that he might contact me by e mail and I could post his reply here but he did one better. Thank you for your time and input Scott!

02-16-2003, 07:31 AM

I appreciate the input. I don't agree with it being brushed off in its importance to repeatablilty and accuracy. Until I can produce either theoretical or enginering evidence to the contrary I will concede the point.


02-16-2003, 02:26 PM
Sorry folks, but the Loco Agitator carrys not weight at all with me. He has no info to back up his statements and a poster who hits and hits and "Hates" etc. He may be right, but nothing has been given to change my opinion either way.

I started believing (for years) that the "register" guided the chuck to the shoulder and that the shoulder made the axis of the the chuck and axis of spindel PARALLEL, The the amount of off set of the two Axis was influenced by the difference between the chuck and "register" dimensions. If the register is 1.495 (most 1.5" registers I havemeasured are more like 1.498 or larger) and the chuck is 1.50", you still have .0025 inch max misalignment. Since i figure a 3 jaw with in .003 is good, I always thought the register was for alignment. Now I have doubts.

But- with a 60 angle on the threads the lifting force needed to cneter a chuck, just isn't there in my opinion. OSO mentions that a taper would have the functions of a "register". I agree, but the taper would have to be great enough to keep the chuck from being self locking. The 60 degree threads would center and not be self locking. So that adds to OSO contention. But I have seen spindle threads with grooves and razor sharp edges, where it looked like the chuck was held on by imagination.

I'd like to take a well worn chuck, spin it on the the threads just short of lock up, then measure and hold the chuck still and turn the spindel and see if the chuck rises. Maybe put a center drilled rod in the chuck and let the tail stock hold the fron of the chuck up then tighten and see if the ass end rises.

Maybe, this problem is like Daves (thrud) tool set "way low" and making a .000006 concave in X inches. a Very intersting thought problem but not often encounterd as a significant problem in the real world.

How about consideringthis theory? THe chuck hangs down at the front pivoting in(or on ) thespindel threads. The spindel/ chuck shoulders are misaligned (the bottom edge ofthe shoulders touch long before the the threads force the shoulders in alignment), the threads force the chuck node to rise, the rear end to drop, and that gives the "lift" that i said was not there!!!!. The "register" and chuck bores can be way diffenrt sizes and still get thesame result?

Now shoot that down, please, cause i just mounted a 4 jaw, the chuck sticks out from the headstock less than 3/8 inch, the threads are deep inside the chuck body. My intent was to keep the overhang to a minimum. I indicated the 4 jaw and it runs dead nuts on. but it shold cause i made it all on the same spindle. Might not carry from lathe to lathe too well. I may have been working with the wrong theory in mind !!!! http://bbs.homeshopmachinist.net//smile.gif

Any way, Its a good question, my mind has been changed- at least in the snese that I no longer know the purpose of the shoulder, "register" and threads and how they interact. Thenks for the question.

02-16-2003, 05:39 PM
I appreciate the reply by Scott Logan, and admire their machines. Glad some manufacturers are still in the USA.
It still is a puzzlement to me why manufacturers go to the trouble to provide a precisely dimensioned un-named area at the end of the threads, and a corresponding recess in the chuck. If it is unnecessary, why go to the trouble and expense to put it there? If it is merely relief for the threads, a chamfer would serve the same purpose at much less cost to produce, and a nice radius at the base of the threads would provide a stronger spindle nose.
In the event of wear or damage, it is easier to replace the un-named area in the chuck by shrinking in a sleeve and turning it to the proper size, salvaging the backplate, than to remanufacture threads.
As far as fit is concerned 0.0005" as Scott suggested is about what you should shoot for in fitting a backplate.
Got to change my vocabulary, had to remove surface grinder last week, register this week. Must remember.

Rich Carlstedt
02-16-2003, 11:12 PM
Well I will try to address some of the questions above.

First, Scott did not say .0005 ! but said .005 to .010 ....a big difference.

Second,regarding the criticisiam for the reason that "manufacturers" don't spell out certain sizes,is called..." Air Fits".
It happens all the time. Look at a 5 C collet, or a R8 and you will see the relief area at all kinds of dimensions. Why ?, because they are air fits and mean nothing. The same applies to the shoulders on spindles. they mean nothing.Air fits are usually a fractional dimension on a print.

Third, the reason for the shoulder is to maintain some rigidity to the spindle. A "groove for thread relief is only going to make the spindle shaft more flexible, an UNdesirable feature. Now for the Engineering reason for this arrangement.
But first lets talk physics..In this world , there are certain things that cannot be changed. For example 3 points make a plane if you are a math nut. In the law of "control", there can only be one Boss for control functions.
For example(in single plane conditions):
A vise has only ONE rigid jaw
A door has only a single hinged side
A compass has only One centering point...
Now before I tick you off, realise what I said ! Please think about it !
If you design Jigs and Fixtures (I have), you know exactly what I said is true.
The location of the fixed point(s) is extremely important to "CONTROL" the movement (or lack thereof). Also,the farther apart the points are in a fixture, the more rigid the control ( if you have a 24" long part, you do NOT put locating points 3"apart...Right?)
In jig /fixture work you need to look at all 3 planes as well but I don't want to get in that deep, so lets get back to the reason for the design of spindle shoulders.

Did you ever try to clamp a washer to a part and find the washer was "bell-ed", thats is it rocked....What did you do ?...chances are that you replaced it with a new one OR you flipped it over !!! Why did you flip it? ....so that it rested on the outside rim which gave it stability (read control!)
Next part...Say you have a pyramid and want to clamp it to the floor with one bolt. Do you choose a corner and bolt it down,or mark the center and put a screw through the center, or do you come down from the peak with a long screw and fasten it to the floor. the right answer is to do the last one, because it spreads the force evenly which gives a balanced load and "control"

So lets put it together.
The shoulder provides strength to the spindle structure.
The counterbored chuck clears the shoulder and keeps the threads from bottoming out (which would stop the chuck before being home)
The turned face at the back of the chuck acts as a totally flat surface (non-bell-ed) and is wider than the spindle allowing you to move the control points outward.
The threads are not at the face but away from the face and like the pyramid clamp apply the force over a larger area. You see, moving the clamping force away from the face, increases rigidity.
So that is the reason for the locations of the parts. I have not heard one guy here do what I have suggested earlier with the 6-32 screw and the 8-32 nut ...Just try it Please! Its the easiest way I can explain it
I'm sorry if I am belittling the point, but think about this....you Cannot have a shoulder locating parts with a screw clamping it at an angle...it does not work.You must realise that the helix of a thread is two axis control...no doubt. no screw in the world tightens on one side. If you think it does, then go try it.

My explanations above are from some of the things I discussed with our apprentices when we reviewed screw loads and stability.
Just as our DNA is a helix, so is a screw!

PS, ever wonder why some guys have stuck chucks on their lathes.. they should check for to tight a clearance.
Have fun

02-17-2003, 01:07 AM

Think about dimensioning and references. You want one reference for any critical dimension, not two.



So the two dimensions would never agree. While tolerances might fix that, not if the tolerance is at the edge of capability to start with.

So what does having a "register" do? It ought to establish a concentricity, right?

Sounds great. And if the chuck were held on with a drawbar, it would be fine (until it wore).
Max eccentricity would be set by the minimum clearance between chuck and register, and shoulder would hold the chuck against the drawbar pull.

NOW ADD A SCREW THREAD. The pull of the tightened thread is going to center the "screw" (spindle) relative to the "nut" (chuck). (Rich's 6-32 and 8-32 test).

If the chuck is off-center, one side of threads will be in close contact, the other will be looser. The pull tends to slide things until the chuck is centered.

Might help to think of just one thread circling the spindle, not the usual 6 or 8. That looks more like a recognizable cone surface, and we all know that mating cone surfaces will center up a part.

BUT, now the register is one reference for center and concentricity, the thread is the other. The thread is out ahead of the register, and it will be cocking the chuck to some degree if it does not agree perfectly with the register center. This may also cause a bind-up on the register surface if it is really tight.
If cocked, the shoulder will not be in full contact.

If the thing is tightened up enough that it bottoms anyway, the spindle will have a bending as well as a "bolt-type" stretching effect happening, and may not be precisely running on center.

So the reason not to have the register extremely precise is that you add a superfluous reference, whose demands cannot be perfectly agreeable with the thread. The result may be less precise than not having a close "register", but having good threads.

Of course, then how precise should the register be?
If it has 0.001 clearance, total (a half each side), it is almost a push-fit, but still can be 0.001 tir out, even if chuck and spindle are perfectly concentric as to the "register" surface. Better watch out if you measure them hot or cold, might shrink on if you don't look out.

If you only get it to 0.003 clearance (1.5 each side), you can get 0.003 tir, which is starting to be baddish for some purposes.

As another thought, what if you just had the shoulder? You would know where the chuck stopped, at least.

Now, imagine a tapered socket facing AWAY from the spindle, IN THE CHUCK. And imagine a matching conical plunger that would fit it.

If you could push that plunger into place, the chuck would be held down to the spindle, and the conical fit would establish perfect concentricity with the plunger center line(if made that way).

That is essentially the action of the threads. Why is it radically different in action if the "cone" is in the form of a thread? It isn't.
Still has a circular inclined surface, fulfills all the requirements.

The newer type noses use a tapered or cone fit, they just turn it around and use a solid cone.

[This message has been edited by Oso (edited 02-17-2003).]

02-17-2003, 01:16 AM
Ok. This is the premise of proper fixturing - prevention of movent in the x-y plane. Clamps are then used to prevent lifting in the z direction. Correct?

Therefore your premis is dead wrong about the shoulder and register.

Lets put the shoulder in the x plane. Now we pace a cylinder on the shoulder and clamp it. What is to stop it from being knocked out of alignment? Nothing.

Now lets mount a disk and secure it to the table. the clyinder is counterbored to the same diameter as the disk. The cylinder is now clamped and cannot be moved because it is restained in all three directions.

Why is this different from a lathe spindle? It isn't. Although the force required to displace the chuck sidways with the thread only forcing it against the shoulder increases with displacement, it does not men that it cannot be displaced.

Also, a 1/4" shorter spindle will be stiffer than the longer one with a "register". The backplate does not need to be counterbored if the thread hugs the shoulder - there is no need for this - it only increases overhang.

And if the register is so unimportant, why are the L and A spindles both tapered 7:24?

Location, location, location.

I guess all those old farts just did it for laughs, and no other good reason.

[This message has been edited by Thrud (edited 02-17-2003).]

Jim Hubbell
02-17-2003, 02:04 AM
When making a back-plate for my 6in Bison 3 jaw I used my 6in 4 jaw. When the rear was close finish I removed the chuck/back-plate combination and used the spindle nose for a test piece. When reinstalled it dialed in just fine. Bearing blue showed the back-plate met the shoulder all around. I guess whatever works works.

02-17-2003, 03:18 AM
I expect that the chuck will not ever touch the register portion of the spindle, unless it is forced to do so by something jamming, a jaw hanging up on the tool holder, for instance, and in order for this to happen, the threads would have to flex far enough to possibly be damaged. Thus a close fit for the chuck, but still a non contact fit. I like the explanation and I concur, that spreading the control points ( talking about the shoulder now) will result in the best stability, and that the threads are actually doing the centering of the chuck, and the shoulder keeping the axis of the chuck parallel to the spindle. The register part of the spindle is only there as a backup during extreme loads, and to keep the runout area for the threads away from the shoulder to ease machining of the spindle. An interesting excercise would be to mount a morse taper stub arbor to the spindle, also a chuck, and closing the jaws onto the stub end of the morse taper arbor. I'm assuming here that the arbor will have the best accuracy, and that closing the chuck jaws onto the stub lightly will allow a person to bring the chuck up to the shoulder already centered.( I'm also assuming that this operation is possible, I haven't tried it). If you then loosen the jaws, remove the stub arbor, and mount a rod equal in diameter to the stub, will it run true? I suspect not, and that the spindle threads will take over the job of centering the chuck, thus proving that the register is not doing any alignment work.
edit- upon further thinking, I don't think a stub arbor would have a long enough stub to be gripped by the chuck. Substitute a collet and mount and true if required, a piece of shafting.

[This message has been edited by darryl (edited 02-17-2003).]

02-17-2003, 03:44 AM
The most important characteristic of any chuck mount is that the chuck repeats its position when re-installed. For threaded mounts, the threads and shoulder allow this repeatability. The threads don’t have to be precision-made within .0001”, but they should have a proper 60 degree thread-form and be clean from burrs. This promotes maximum surface contact between the external and internal threads. The threads of my Logan 1-1/2x8 spindle are ground and don’t require a relief groove. D-, A-, and L-taper mounts locate off the taper itself; any threads employed are used to hold the chuck to the taper. This is different from a threaded mount. The threads of a threaded mount are the taper, in effect, they locate the chuck axially. One thing they do well is allow the chuck to repeat its position when re-installed. They are perfectly suitable for lighter demands. D-, A-, and L-taper mounts are produced easier and faster, are stronger and more user-friendly.

In my earlier reply, I mentioned that I just received two new backplates with the threads already cut. The counterbore in each of them is .035-.038” larger than the “register” of my spindle. This does not matter because the “register” does not position the chuck axially, the threads do. Remember, it works because the backplate repeats its position on the threads when re-installed. If the threads and the shoulder are clean when the backplate is initially mounted and turned down to create a boss for the chuck to be mounted, the backplate will repeat its position when re-installed, granted the threads and shoulder are clean. The threads cannot help but to position the backplate axially.

02-17-2003, 09:45 AM
I understand air fit, it is just a relief cut to clear anything that may be in the way at a non-essential point.
I added an extra 0, sorry. .005 clearance will give max runout of +-0.0025, better than most three jaw chucks can hope to be, especially after some use.
I still question why the register area is there on the spindle. This has not been answered. Why go to the trouble to put in something to add to the difficulty and expense of manufacture if it serves no purpose? It adds no strength, it makes threading more difficult, as a slight undercut would make threading the spindle much simpler.
I have read the analogy of the 8-32 nut on the 6-32 screw on another site. Put them together, tighten them up and smack the nut on the side and watch it move. Sorry, I don't buy it. Damage and wear to the threads will cause as many problems as the same will to the register, and improper cleaning before assembly is the most likely cause of stuck chuck.
The best way to fit a chuck is probably to turn down a small post hole, and insert it in the backplate. Save yourself some trouble that way.

Herb Helbig
02-17-2003, 11:35 AM
I hope the experts will confirm this comment (if it's correct) for the benefit of us new guys.

The only reason for all this fuss over mounting a 4-jaw chuck (with independent jaws) is so that a chucked work piece can be removed from the lathe for some other operation and then returned to the lathe without losing its concentricity with the lathe axis.

The original centering of the workpiece will always be dialed in anyway, and as Forrest so helpfully pointed out, it doesn't even matter if the jaw surfaces have a tendency to hold the work at a small angle with the lathe axis because you can hold the work near the outer ends of the jaws and "bang" it into concentricity.


02-17-2003, 06:36 PM
I have a theory as to why the "register" area might be made on the spindle for reason other than to center the chuck. Here goes: The most stressed area of the lathe spindle is where the backplate meets the shoulder. As side (lateral) pressure is applied while cutting, the shoulder acts as a fulcrum as the chuck, with attached faceplate, is forced away from the cutting tool. The minor diameter of the "V" thread is the weakest and most likely area for part (spindle) breakage or distortion (bending) not only becaues the spindle diameter is smallest at that point but because the sharp "V" bottom of the thread causes a stress riser. Were the spindle threaded all the way back to the shoulder, the area where the stress is greatest, would be weakened by the thread. That area would eventually become fatigued and break or bend at the shoulder. The "register" area moves the thread (including the sharp "V" minor diameter) away from the area of greatest stress. Granted the move isn't far, perhaps only an eighth inch or so, but the forces would be much lower even that small distance away from the "fulcrum" point.

Rich Carlstedt
02-17-2003, 06:59 PM
Dave....Your quote:
"Lets put the shoulder in the x plane. Now we pace a cylinder on the shoulder and clamp it. What is to stop it from being knocked out of alignment? Nothing.
Now lets mount a disk and secure it to the table. the clyinder is counterbored to the same diameter as the disk. The cylinder is now clamped and cannot be moved because it is restained in all three directions...... it does not mean that it cannot be displaced."

CORRECT, on both parts !
I completely agree about dowel type fits for location....been there ,done that
BUT,you are looking at a lathe spindle.
on a 9 inch (for example) SouthBend, you have a 1 1/2 X 8 " thread....this is horendous when you consider the loads placed on it!
(FYI a 1"-8 SH Cap screw can take 850 Foot pounds of torque !)
When you compare the thread size for clamping. If we took your first example and put a 1 1/2" thread boiler clamp, it would be really tough to move, but the part could slide, BUT not the thread !!!
On a spindle thread the engagement with the helix means that any side load TRANSFORMS into a tightening force(!) on the thread. As the thread tightens, it is even more centering in its motion

As for " why are the L and A spindles both tapered 7:24?" I can only say the most accurate spindle in my opinion is the "L" because of the taper..Its my favorite.

Herb..You said
"I have read the analogy of the 8-32 nut on the 6-32 screw on another site. Put them together, tighten them up and smack the nut on the side and watch it move. Sorry, I don't buy it."

Well, thats because you didn't try it and I didn't say to take a hammer to it, isaid to hold it in your hands. I was trying to show you the effect of force on a spiral helix and its self centering properties. Even with wear, a thread will always go to its' axial center because the loads and wear is even.

Have Fun

02-18-2003, 12:15 AM
I agree there is tremendous forces present in metal cutting - hundreds of thousand of pounds of force at the cutting tip itself. Otherwise it could not form a plane of cleavage to "cut" the metal. This force must be transfered to the spindle thread tightening the chuck. It would be interesting to actually measure the force applied to the shoulder by the back plate by the cutting forces. And in retrospect I wonder if the "register" is not only acting as a "dowel" of sorts, but also helps prevent tipping of the chuck under hign cutting forces - it certainly will prevent any side slippage, keeping the chuck parallel to the spindle axis.

I realize you do not agree with my analogy of the spindle/chuck to a fixture/workpiece, but if you think about it is a legitimate comparison even though we are talking about a very large "retaining bolt" - the 1-1/2"x8 spindle thread.

If threads are self centering as you have said, why bother with any precision threads on any equipment? (Fasteners, not leadscrews)

Seems to be a big waste of money on tight tolerance parts for all those jet engines, rocket motors, space shuttles, blah, blah, blah! (JUST STIRRING THE POT SOME MORE http://bbs.homeshopmachinist.net//biggrin.gif)

02-18-2003, 12:19 AM
I believe that the most stress in the spindle/chuck combination will be on the threaded portion itself. Consider the angle of the thread, and the stretching force exerted just by tightening the chuck. Now add sideways force to the chuck. The tension on the spindle threads will be greater than the sideways force exerted on the chuck due to the 60 degree angle of the threads. Add to this tension the initial ammount due to the tightening of the chuck. The portion of spindle thread between the register and the last thread in the chuck is unsupported either by the threads in the chuck or by the slightly extra meat provided by the register. I'm sure that if you broke the spindle off with the chuck attached, you'd find the register area still attached to the spindle, and the threaded part still poking out of the back of the chuck.

Herb Helbig
02-18-2003, 12:55 AM
Rich (et al.) -

Back on Page 1, please note that it was JCHannum who didn't buy your 6-32/8-32 model. My question remains unanswered. I'll repeat it here for convenience.
I hope the experts will confirm this comment (if it's correct) for the benefit of us new guys.

The only reason for all this fuss over mounting a 4-jaw chuck (with independent jaws) is so that a chucked work piece can be removed from the lathe for some other operation and then returned to the lathe without losing its concentricity with the lathe axis.

The original centering of the workpiece will always be dialed in anyway, and as Forrest so helpfully pointed out, it doesn't even matter if the jaw surfaces have a tendency to hold the work at a small angle with the lathe axis because you can hold the work near the outer ends of the jaws and "bang" it into concentricity.


02-18-2003, 11:27 AM
I've never been able to rechuck a piece in the 4-jaw and find perfect (or even nearly so) concentricity with its previous position. I'd say upon rechucking its about the same as if I were using the 3-jaw. Am I doing sump'n wrong?

Later...(02/20/03): My mistake Herb. Upon re-reading, I now see your intended meaning. It was there all along...I just missed it.

[This message has been edited by lynnl (edited 02-20-2003).]

Herb Helbig
02-18-2003, 12:17 PM
Maybe I wasn't clear. The idea is that you can remove the piece IN (WITH) THE CHUCK for additional work, say on the mill, and then screw the chuck back on the lathe in "perfect" alignment.


02-18-2003, 12:33 PM
Then I guess that the proper method to make a backplate for say an 1-1/2"-8 spindle is poke an 1-1/8" hole in it and shoot in 8 threads per inch right?
Bore the relief area to 1.6" or so, so it does not interfere with anything, but still fetches up on the shoulder.
No need to trial fit anything, just make sure everything is larger then the spindle thread, it will spin on OK, and center up perfectly every time.

02-18-2003, 01:10 PM
Dear Friends,

In medieval times the learned ones argued interminably on the question of----- how many angels could stand on the head of a pin?

Does anyone know if an answer was finally agreed upon?

Here's an easier one to ponder upon. I rely on cereals for my breakfast, a simple meal of oatmeal alternated with grits on some days. A couple of weeks ago I found a live worm in my grits. This happened once before. I assumed that an egg-laying moth had flown into the partially open box. To prevent a recurrence it has been my practice to empty the box of grits into a screw lidded (glass jar to keep out foreign creatures. Do I need lid with a better register?

Called the Albers grit company, received profuse and interminable apologies. To comfort me they sent me coupons good for 15 LBS. OF GRITS!

I can't even look at the 1# box of grits without imagining hordes of squirming worms.

Now what to do? Apparently worms are an inescapable part of the packaging process.


Rich Carlstedt
02-18-2003, 01:51 PM
To All
I have always taken a chuck or faceplate,with a part, out and returned it with out a problem..I think if you try it , you will understand the nature of the beast.
My 6 Jaw Buck Chuck, which many consider to be the finest chuck made for repetative work has not had a adjustment made to it in years, and I have removed it /replaced it many times because only very fine work or tubing goes into that chuck and i don't leave it in for "rough stuff"
As said earlier, it has over a .020 diff on the shoulder diameter.

Hey Herb, I appoligize for using your name in error ! Comes with age I guess, I can't even call my kids by their right name !

Dave, don't start me on thread specs. We have engineers out there who do not know the difference between a cut or rollformed thread. Fasteners are a whole ballgame in them selves. for an appetizer check out the diff from metric screws and US standards. Alot of metric is like an Afgan fleamarket

Have a great day guys....and make chips, not arguments !

02-18-2003, 11:51 PM
I will have you know the human body consists of approximately 1.6 trillion cells of which only .7 or so are human. We are a walking, talking biospheres - arguing about, of all things threads, chucks, and extra protein. We eat critters all the time. But you don't see them, so out-of-sight-out-of-mind. Even when we die we are still a viable eco system for quite some time.

I guess it will remain a mystery - I might call Hardinge and ask them. Just out of curiosity. It was a darn good arguement while it lasted though. http://bbs.homeshopmachinist.net//biggrin.gif

02-19-2003, 04:36 AM
All this talk about worms and grits and human biospheres makes me want to chuck my plate. I think I'll turn it down till it's only big enough for dessert. '101' indeed.

02-19-2003, 07:11 PM
Oscar: My wife is a "yankee" (her family fought for north in civil war- Mossouri)and she loves her real grits. Buys them from a guy who has a stone burr mill. Leave any corn product long enough and it gets "weavily". Throw em out and get some fresh stuff. You and I have been eatingthe eggs many years. they are well boiled. Adds protein to a pore mans diet!!!!!
Same with beans- which I dearly love. Never let your imagination over come your appitite. Me? i eat any thing with four legs other than the kitchen table!!!! Have gnaw marks to so prove.

02-21-2003, 10:26 PM

The trouble is: It was a brand new box. As soon as it was opened it went straight into that glass jar with the screw lid.

Damn------------- first time I encountered grits was in the Army. Thought I had been given a plate of plaster for breakfast. It took almost 50 years before I tried them again and found that "civilian" grits were delicious. No one told me it was the worms that flavored it.

Then there was my encounter with the army's Turnip Greens. War was hell indeed!


02-26-2003, 11:07 PM
Thanks to all who responded. I plan to get started on the chuck plate soon. Paul