PDA

View Full Version : Some observations on three jaw self-centering chuck accuracy.



gmm22
12-19-2006, 11:04 PM
I am new to lathes, but I have experience in trying to hold high tolerances in making things. I have made some interesting notes on my chucks for my new mini-lathe. No doubt most if not all of these are well known to experienced persons, but I have questions regarding the topic of accuracy.

1. There are nine permutations of mounting and tightening sequences with a three bolt mounting. One mounting orientation and one tightening sequence emerged as most accurate with much variation (.001" to .007" deviation).

2. The choice of which of the three chuck receptacles to use to tighten also had great effect on accuracy with as much as .006" difference from best to worst.

3. The best I have achieved in a consistently repeatable way is .0003" run out on 3/8" diameter stock.

My questions are:

(a) Is .0003" run out considered very good for three jaw independent chucks?

(b) What will most machinists accept as a deviation when using a four jaw independent chuck?

(c) I have set up work in the four jaw to get the same .0003" and I find the time taken to achieve that is probably the same or more than what it takes me to get it by tweaking the three jaw self-centering chuck, therefore, is there some secondary reason why I should be using a four jaw for very accurate work that I am not aware of?

I must confess a preference for the self-centering chuck. The tweaking is just as lengthy, but it is of a different nature than the four jaw independent chuck.

Any thoughts would be appreciated.

nheng
12-19-2006, 11:13 PM
I prefer the 3 jaw chuck, self centering especially, as it brings to the table most of what the 4 jaw offers but IMHO, with more precision for finer work without the hammering, shimming, etc.

Of course, for odd shaped or eccentric work, I reach for the 4 jaw in a heartbeat.

One factoid about the 3 jaw that always fades into the archive ... the 3 jaw chuck has ONE marked key hole which should be tightened for best accuracy. This from the horse's mouth at Rohm. I've seen this marking now on numerous good quality 3 jaw chucks.

The scroll will ALWAYS have some clearance to the center hub of the chuck body. When the marked keyhole is used, the scroll will be forced toward one side of the chuck. Presumably, this is the same position used at the factory when the jaws were given a final grind.

On a Bison 3 jaw, non set-tru, using the marked keyhole give around 0.001" runout while using one of the others seems to add an average of another thousandth or so. It is pretty repeateble.

I've heard old timers advise tightening all 3 holes on keyed DRILL chucks and I haven't explored their construction enough to know if there is a good reason for this ... but I trust them.

Den

Scishopguy
12-19-2006, 11:25 PM
Hi guys;


One thing I became aware of early on was that round stock, even stuff that looked like it had a good mill finish, is not always that concentric. I found that if I were going to have to rechuck a part it had a better chance of going back into the chuck and running true if I marked a line on the part and the #1 chuck jaw when I first put the blank stock in for the first time. Sometimes, if the stock is out of true, I can tap it lightly with a soft face hammer while the spindle is rotating. That will fix any misalignment from a small bit of swarf or grime.

wierdscience
12-19-2006, 11:38 PM
.0003"?That's getting into spindle runout range,so ya that's pretty good.

Like said one pinion is used at the factory to lash out the scroll and jaws prior to final grind,it's called the master pinion.

C - ROSS
12-19-2006, 11:39 PM
A repeatable runout of 3/10's is good if fact super good. I will agree the three jaw is much easier and faster to use but when chucking odd shaped items the 4 jaw is the only way to go.

As Nheng stated rounds are not always round.

If you need repeatable accuracy in the 10's you need to use Collet's.

My advise is only worth what it cost you.

Ross

gmm22
12-19-2006, 11:47 PM
Just for the record, I erred by saying "stock". I have been using brand new high quality center drills for testing as I assumed they would be of higher tolerance than typical round stock.

CCWKen
12-19-2006, 11:51 PM
I just got through truing up an old 6 1/4" Bison 3-jaw for a 1 1/2"-8 threaded spindle. I'm happy with the ~.0002 runout at about 2" from the chuck. I don't do work for NASA so that's dang close enough for me. I did notice a difference if only one pinion was used to tighten the jaws. The runout jumped to .003. I'm already in the habbit of using all three pinions in the tightening sequence so I'm satisfied. (I do it with drill chucks too.)

Keep in mind that a speck of dust can change your runout. Think what a metal chip can do.

Forrest Addy
12-19-2006, 11:57 PM
0.0003" is very good indeed for any chuck or collet. Usually those adjustable three jaw chucks hold concentricity well allowing repeat work and good contricity. They do not suffer overload and minor wrecks well so be careful when parting gummy materials.

I don't suppoes anyone wants to hear my 4 jaw rant - again - so I'll stop here.

J Tiers
12-20-2006, 12:07 AM
I prefer the 3 jaw chuck, self centering especially, as it brings to the table most of what the 4 jaw offers but IMHO, with more precision for finer work without the hammering, shimming, etc.



What would the "hammering and shimming" be?

Hammering is NOT S.O.P on any work I do with any chuck......I save that for bedding parts in the vise on the mill or shaper...... and I don't hit hard.....

The ONLY thing the 3 jaw offers that can beat a 4 jaw is quickness........and holding things with a number of facets divisible by 3......"more precision" is not what I associate with it..... Unless you mean a 3 jaw WIRE chuck *

It won't hold as tight, It can't get as accurately centered, it won't hold anything with an even number of facets well.

Sheesh... I think I am starting to sound like the other 4 jaw bigots...... well, too bad...... cuz they are right ! :D

I did have a 4 jaw that needed to be shimmed for anything precise.... but I replaced it... it was also worn so it wouldn't grab anything under about 0.375 diameter... had a mostly stripped jaw thread.....etc, etc, etc....

* wire chuck = collet

Mcgyver
12-20-2006, 12:11 AM
most of the inaccuracy in a 3 jaw is in the scroll. noting the tir at one spot isn't necessary indicative of the accuracy of the chuck, ie .375 might be spot on but when the scroll is turned to say .395, its a different story. if on the other hand, you've tuned your chuck to produce .0003 tir across its range, that is truly impressive and i hope to read the article soon :D

As has be discussed many times, it really only matters though if you have to keep the section held by the jaws concentric with the part being turned. so long as you are turning without removing the work, the accuracy of the scroll doesn't matter. if you do disturb it, and the work is anything but crude (ie better than 2-4 thou tolerance), use the 4 jaw or collets to do the next op. Den i disagree re the 4 jaw, it is as accurate as your indicator, far more than the three jaw and is regularly used to keep things concentric to a few tenths.

the question isn't what a machinist thinks is tolerable, its what the design and job at hand requires, a hitch pin for a farm wagon is different than engine parts for example. there is no standard answer, but most wouldn't expect better than a thou or two from a high quality 3 jaw, but it doesn't much matter imo, whether your three jaw is out 5 thou or 2 as most of the time 2 is just as unacceptable as 5 - neither is acceptable for accurate work. When you need to keep and OD or ID concentric to something already turned on the part, you must either use a collet or indicate with the 4 jaw. I use a tenths indicator because it lets you get easily better than 1/2 a thou, with care to 2 or 3 tenths.

I have never thought of the 3 jaw as being something to fiddle with beyond the inherent accuracy of the scroll, ie which pin is tightened etc. Other than experimentation, it would be a waste of time to fiddle with the 3 jaw accuracy as it lacks the direct control a 4 jaw has, ie indicator says i'm out 1 thou in this direction, tighten this jaw to correct sort of thing. the 3 jaw is a convenience we like to use frequently, but is the poor cousin the the 4 jaw both in accuracy and versatility. Use when concentric accuracy doesn't matter or more commonly when you can turn everything without disturbing the work, ie turn, face, bore, part off, is the 3 jaw's domain.

gmm22
12-20-2006, 01:35 AM
"Other than experimentation, it would be a waste of time to fiddle with the 3 jaw accuracy as it lacks the direct control a 4 jaw has, ie indicator says i'm out 1 thou in this direction, tighten this jaw to correct sort of thing."

-----------------------------------------------------------

Oddly, this is at the heart of my inquiry. If I am able to develop a repeatable work mounting protocol that can net me .0003" accuracy at any dimension, does this not qualify as direct control?

gmm22
12-20-2006, 01:39 AM
[QUOTE=Forrest Addy]

They do not suffer overload and minor wrecks well so be careful when parting gummy materials.

-------------------------------------------

What does this mean? Can you elaborate?

A.K. Boomer
12-20-2006, 02:29 AM
I've heard old timers advise tightening all 3 holes on keyed DRILL chucks and I haven't explored their construction enough to know if there is a good reason for this ... but I trust them.

Den

Nheng, this is good advise, at the very least it give a better grip without any added stress and it generally improves accuracy...

I mentioned this before on this site but you can make eccentrics on a three jaw and quite a variety, just put the jaws in where they dont belong or leave certain ones out one tooth and record your results, --- dont turn anything heavy or long or to fast though:cool:

BadDog
12-20-2006, 03:06 AM
[QUOTE=Forrest Addy]What does this mean? Can you elaborate?
By their nature, they are less stable than a traditional scroll 3 jaw. You have an intentional gap that the adjust-true screws use to skew the chuck on the back plate. That means instead of having a full circumference register to hold/support the chuck on extreme side loads, you now have a point load at the end of that screw. The crew can deform, the point blunt, or the chuck itself dented by the point; all to the tiniest amounts, but cumulatively you’re now out of true again. Furthermore, such “crashes” can leave it so that the chuck and spindle/backing plate are no longer even axially parallel for similar reasons. And, you still have the jaws and scroll which sustain wear and deformations as well just like a normal chuck.

John Stevenson
12-20-2006, 04:52 AM
What work are you doing that requires a 0.0003" tolerance and is it repeatable every time ?
What happens if you wind on 20 thou as a last cut, do you get the same results as winding on a 5 thou finishing cut ?

There is more to this than just quoting numbers, speeds, feeds and more importantly tool marks come into play at these stated tolerances.

I spend a lot of time doing bearing diameters, external and internal and I always aim for the best finish at a thou off what I want and then polish to size, That gets me the finished size I require and a good finish.

Chucks are cheaper now then they have ever been, even good ones. I'm amazed when I read adverts for someone wanting a set of jaws for a chuck that went out of production 40 years ago.
Chances are if they find a set they will be worn or the chuck isn't even worth it and the cost will be 75% of a new chuck with both sets of jaws.

If absolute accuracy is needed get a set of soft jaws and you won't look back and with soft jaws those weird jobs like disks and stepped flanges are a thing of the past.

I know you still have this chuck that you are loath to get rid off as it's only missing a set of jaws but believe me it isn't worth it.
If you are that much of a pack rat then weld it [ shudder ] to a plate and bolt it to the bench as a vise for holding round work.

Last place I worked we welded a big 15" 4 jaw direct to the bench and believe me that chuck could hold any shaped material safely, pipe flanges, elbows, castings.................


.

Your Old Dog
12-20-2006, 05:26 AM
Mcgyver, well put and easy to understand. I believe you are right on all counts.

My SB9's 6"chuck has runout of .009 on drill rod at 3" from chuck. I always have to take a truing cut and then I index the number one jaw with a center punch mark so that if I have to remove the piece I can get it back to close to where it was.

Chasing accuracy too far can only cause heartache! Pure accuracy or "perfection" is for the most part just a goal and not a real destination for most folks. If you can keep from getting carried away with the chase, it does lead to better workmanship. But don't get carried away! :D

After hearing one of Forrest's rants for the 4Jaw I tried it using his methode of tuning it in and it only takes me a minute longer to set it up.

Great thread. Makes me want to melt down my 3jaw for fishing weights !!

JCHannum
12-20-2006, 06:56 AM
[QUOTE=Forrest Addy]

They do not suffer overload and minor wrecks well so be careful when parting gummy materials.

-------------------------------------------

What does this mean? Can you elaborate?

The previous comments on three jaw versus four jaw chucks are correct.

If you are fortunate enough to tweak a three jaw chuck to the accuracy described, it is likely that that number is only achievable at a given diameter, and will change at other diameters for no other reason than manufacturing tolerances in the scroll and jaws. Normal use, wear and dirt will add to the inaccuracies.

The scroll is not robust, particularly in smaller chucks and it is subject to damage and wear, which will directly affect accuracy. A crash can cause the scroll to be deformed at that point. Cutting off is one of the most common causes of wrecks, and gummy material the most common culprit. It can only take a second to spring the scroll and destroy all the accuracy you have so carefully developed in your chuck.

0.0003" is very good accuracy, but maintaining that level in day to day operation is not likely. Buck guarantees their Adjust Tru chucks to 0.0005" TIR, but only on duplicate parts, recognizing the variable nature inherent in the chuck's components.

If you want speed and reasonable accuracy, the three jaw chuck is a convenient chuck to use, but for the best in accuracy, holding power and versatility, the four jaw is the better choice.

Mcgyver
12-20-2006, 09:10 AM
Oddly, this is at the heart of my inquiry. If I am able to develop a repeatable work mounting protocol that can net me .0003" accuracy at any dimension, does this not qualify as direct control?

I guessed since it was a mini lathe and that your adjustments seemed to be via which pin was tightened etc, that you are not working with a set-true or similar where you actually can fine tune the jaws.

Like I said, if you've managed to really do that, over its range I'm all ears. I suspect you will have to go through the same gyrations for a different dia, which led to my conclusion that other experimentation its not a practical way to line things up compared to a 4 jaw because its hit and miss, you can't assert a simple action/reaction like move this jaw, and the work moves this way etc.

I think its good that you question and experiment, who knows you may come with something new, so don't take the responses as negative it's just this stuff is old, well understood technology that many here are well versed in. consider as part of the experiments cutting down the cycle time setting work true in a 4 jaw; with a bit of practice its done very quickly.

YOD, don't melt that puppy, 3 jaw sits on my lathe too more than the others, unless you've one of those super soft jaw models, the weights will rust

gmm22
12-20-2006, 10:11 AM
Thanks for all the interesting replies. I better understand some of the limitations of the three jaw.

What are the soft jaws that John Stevenson refers to?

Another curious note. I get the best numbers by using only one specific tightening pin. If I use all three sequentially, the runout is greater.

japcas
12-20-2006, 10:37 AM
Soft jaws are jaws that are usually free machining steel or aluminum that are bolted on the chuck in place of the hard jaws. Then they are bored to hold the piece you want to chuck. Repeatability is excellent since they have been machined on the machine they will be use on. You must chuck on something back in the jaws while machining them or other wise the slack in the whole chuck will not let you bore an accurate set of jaws.

It is nice to have a real accurate 3 jaw but as others have said it probably won't be that accurate forever or over it's entire range for long. There is no substitute for being able to setup a four jaw chuck accurately. It may seem like a bit of a chore but with a little experience one can set one up in no time at all. And the holding power of the 4 jaw versus the 3 jaw is so much better also.

pcarpenter
12-20-2006, 02:34 PM
Whether a job requires .0003 runout limits or not, it is always nice to know that your equipment is contributing as little as possible to the tolerances you do end up with. I find that making a part without screwing something up can be trying enough sometimes. Add in extra hurdles to keep in consideration while working and you just make things worse.

I posted in a similar post some time back, results that were very similar to what Ken suggested. My 3-jaw scroll chuck on my fairly new Chinese lathe would produce about .0002-.0003 on a ground and polished dowel pin about 1.25" diameter and maybe an inch or so from the jaws. It would do this consistently if you tightened at all three locations, no matter which one was done first. If you tightened only one, runout was about ten times that. I changed the dowel pin orientation to take out any possibility that it was the pin tolerances that were cancelling the chuck tolerances. It will be interesting to see how long this lasts as the jaws become more loose in their "ways" as the chuck wears. The chuck body is softer than the jaws and it is sure to wear over time.

If you are spending as much time diddling with a scroll chuck to get it on center as a 4-jaw, then something is amiss...and the value of the scroll chuck is diminshed. I often want to be able to interchange parts or flip from one end to the other quickly and so keep the three jaw mounted most of the time.

I would love to have a magic 4-jaw scroll chuck that would scroll the jaws but also allow them to be disconnected easily from the scroll so it could be used as an independent chuck when needed. This would save swapping out chucks dependent on the work.

Paul

japcas
12-20-2006, 04:00 PM
Paul[/QUOTE] I would love to have a magic 4-jaw scroll chuck that would scroll the jaws but also allow them to be disconnected easily from the scroll so it could be used as an independent chuck when needed. This would save swapping out chucks dependent on the work. Paul[/QUOTE]

Believe it or not we have one at work. It is independently adjustable on each jaw but has a scroll that moves all the jaws at once. Great way to dial in a job and then quickly take parts in and out of the chuck. It had a D1-3 mount and we used it on one of the Monarch 10EE's. I haven't seen it in a while though so I believe someone must have destroyed it or gave it a new home some where else. It sure was handy to use though.

David S Newman
12-20-2006, 04:04 PM
This is my first post on this forum, hope it meets with some approval.
I have been tinkering with lathes etc for many many years as a hobby, what I did as I have a couple of three jaw chucks with seperate back plates, is on one of them I turned the register undersize and drilled the three holes in it oversize, when I need to get something running spot on I grip it in the chuck slightly loosen (just about gripping) the three screws and tap the chuck with a rubber hammer until the piece to be turned is running true against a TDI. I then tighten the three screws. It's always worked for me and far quicker than a four jaw independent. David

pcarpenter
12-20-2006, 04:20 PM
David-- I think Kap Pullen posted something similar here about having made a sort of "poor man's adjust-tru chuck" that same way. Not a bad way to go and it might make sense for me once I find that I cannot hold great tolerances with the one I have. For me, dial inaccuracies seem to be more at fault when I screw something up than would ever be accounted for due to chuck runout:(

I thought I had heard of a 4-jaw that would both scroll and not--and if I recall correctly, they were *very* expensive, perhaps prohibitively for us home shop types. The trouble is I want a *magic* one:D I want one that will allow you to use it independently and then go back to scrolling without having to be re-trued to center....which I assume you would have to do with the commercial type that allows you to unlock the jaws from the scroll. Oh...and it has to be cheap too....that really would take magic;)

Paul

Scishopguy
12-20-2006, 04:29 PM
I would love to have a magic 4-jaw scroll chuck that would scroll the jaws but also allow them to be disconnected easily from the scroll so it could be used as an independent chuck when needed. This would save swapping out chucks dependent on the work.

Paul

Hi Paul,

We had a six jaw, 10" chuck that fit the Monarch. It had six independently adjustable jaws that were on a scroll. You just indicated the part in and could take it in and out at will with just one adjustment.

japcas
12-20-2006, 04:39 PM
I thought I had heard of a 4-jaw that would both scroll and not--and if I recall correctly, they were *very* expensive, perhaps prohibitively for us home shop types. The trouble is I want a *magic* one:D I want one that will allow you to use it independently and then go back to scrolling without having to be re-trued to center....which I assume you would have to do with the commercial type that allows you to unlock the jaws from the scroll. Oh...and it has to be cheap too....that really would take magic;) Paul

Paul, the one I talked about earlier worked that way. The scroll moved the adjustable part of the jaw in and out. Then it had a socket like a regular four jaw that you turned to dial the jaw in or out to true the part. Once trued, you only moved the scroll adjustment. You don't have to mess with each jaw after that unless changing to a new size or part.

pcarpenter
12-20-2006, 04:50 PM
Scishopguy...it sounds like you are just describing an "adjust-tru" type chuck (I think that is Buck's trade name for theirs but there are others). Those typically only give you a small amount of independent adjustment to make up for the normal inherent inaccuracy of a scroll chuck or for small variations in the roundness of round stock.

...unless I am mis-reading the re-description.

I was asking for something that may not exist.... A chuck that operates as a scroll chuck *and* is also capable of operating as an independent-jaw chuck with independent adjustment of as much as the entire jaw opening range. This would allow you to work on odd shaped stuff (rectangular etc) as with a normal 4-jaw independent chuck, but without changing chucks, go back to scrolling for working with round or square stock. Like I said, that sounds like magic as you would have to be able to detach the jaws from the scroll and still adjust them across several inches and then be able to have them re-mesh with a scroll, all in correct relation to one another.

I don't want much :)

Paul

John Stevenson
12-20-2006, 05:02 PM
Got one here on a high end dividing head.
3 jaw scroll chuck as standard but there is a moving jaw on each jaw with a screw like a 4 jaw has.
You can dial in dead nuts on and use the scroll to swop parts or even setup for an eccentric and repeat every time.

.

JCHannum
12-20-2006, 05:59 PM
The topic of adjustable scroll chucks was thrashed over on the PM site a couple of weeks ago;

http://www.practicalmachinist.com/ubb/ultimatebb.php/topic/30/2089.html#000003

daryl bane
12-20-2006, 10:47 PM
Here's the beast. Just got it, will see how it is.
http://img.photobucket.com/albums/v417/brufsupbane/3cbb_12_sb.jpg

WLW-19958
12-21-2006, 01:04 AM
Hi There,

Daryl,

That looks almost exactly like the one I have.
What make is it?
What is its diameter?

Good Lick!
-Blue Chips-
Webb

pcarpenter
12-21-2006, 01:30 PM
Thanks for the pic....I am a sort of visual learner;)

The chucks that are being described are centered around turning a bit off center or correcting for runout in round stock.

There are two qualifiers that I am not explaining well.

1. it would need to be a 4 jaw so that you could turn rectangular stock as well as round etc. You cannot safely turn an odd rectangular piece in a three jaw chuck.

2. It would need a way to have the individually adjustable jaws return easily to zero. If you had to re-zero it back to "centered" after you turned a rectangular or odd shaped piece in order to use it as a true scroll chuck for round stock (without spending several minutes with an indicator) then the value of being able to save the time of mounting and remounting chucks would be a moot point.

If you can just run the adjustable top jaws back to a stop and have say a thou or so runout, then it would solve the chuck swapping problem.

I don't want much, do I:D

paul

daryl bane
12-21-2006, 02:06 PM
It a 6 1/4 " Rohm, I'm looking forward to using it. I thought is was a conventional set-tru, but found out differently.

Scishopguy
12-23-2006, 12:52 PM
Scishopguy...it sounds like you are just describing an "adjust-tru" type chuck (I think that is Buck's trade name for theirs but there are others). Those typically only give you a small amount of independent adjustment to make up for the normal inherent inaccuracy of a scroll chuck or for small variations in the roundness of round stock.

...unless I am mis-reading the re-description.

I don't want much :)

Paul

Paul,

No, this was not an ajust tru. I had one of those on another lathe. Each jaw on this chuck had an adjustment, just like on a four jaw, and it had the adjustment on the side of the chuck, just like any three jaw. It's not just wishful thinking, they do exist!

japcas
12-31-2006, 08:06 PM
I was looking through the January issue of the Home Shop Machinist mag when I stumbled across the add from New England Brass and Tool. They are now selling combination chucks like pcarpenter was wanting made by Bison. They look very good and are available in 5,6, and 8 inch sizes I believe. They are a bit pricey though. They are 4 jaw on a scroll with independent adjustments on each jaw.

These chucks can be trued up pretty quick by realigning all of the jaw to one of the lines on the chuck face and then indicating the first part in. They are used when multiple parts need to be chucked quickly but accuracy is a concern. Once set they give 3 jaw quickness with 4 jaw accuracy although with a scoll in them probably wouldn't hold as tight as a standard 4 jaw.