PDA

View Full Version : Expected runout for a 3 jaw chuck?



PeteF
03-05-2011, 07:27 AM
I just replaced my old 3 jaw chuck with a TOS 3 jaw, and it seems to be a nice chuck. It's a plain backplate version. Fiddling around with the backplate, changing the orientation etc I can get the runout down to around 0.01 mm on a piece of ground 12 mm rod measured just out from the jaws. Normally for really precise work I'd use collets, and indeed do mostly use them, so I'm not sure just how accurate I should be able to get a brand new chuck like this. Is it worth continuing to try to refine it more or does that sound about what I should expect?

Pete

philbur
03-05-2011, 07:38 AM
0.01mm is way better than you have a right to expect from a low cost 3 jaw. However before you get to excited recheck the run-out with a 1" rod without readjusting the chuck position on the backplate.

Phil:)


I just replaced my old 3 jaw chuck with a TOS 3 jaw, and it seems to be a nice chuck. It's a plain backplate version. Fiddling around with the backplate, changing the orientation etc I can get the runout down to around 0.01 mm on a piece of ground 12 mm rod measured just out from the jaws. Normally for really precise work I'd use collets, and indeed do mostly use them, so I'm not sure just how accurate I should be able to get a brand new chuck like this. Is it worth continuing to try to refine it more or does that sound about what I should expect?

Pete

PeteF
03-05-2011, 08:01 AM
Thanks Phil, I wouldn't have any precision ground rod that size here that I can recall. I just measured that before I stopped working for the night, but tomorrow I'll also do the usual tricks for more accurately locating the rod in the chuck etc. However I just don't want to be spending hours fiddling around with something if at the end of the day it's already about as good as I should expect. Having endured clapped out chucks for so long now, it's a bit of a treat having some brand new chucks!

.RC.
03-05-2011, 08:03 AM
Probably within 0.05mm would be good.... The TOS 3 jaw I bought years ago (little 100mm one) had horrendous runout, but was probably within spec...

philbur
03-05-2011, 08:11 AM
Concentricity on a 3 jaw self centering chuck is not a high priority. If you need concentricity use a 4 jaw independant or a collet or centers. A 3 jaw is about convenience, once you skim the workpiece it will be concentric to your spindle precision anyway.

Phil:)

dian
03-05-2011, 08:28 AM
i believe a straight rod is not needed. if you make multiple measuments rotating the rod in the chuck and find the largest runnout, then rotate the rod 90 and measure again, you are measuring the runnout of the chuck (assuming the rod is round). is this correct?

PeteF
03-05-2011, 03:08 PM
Concentricity on a 3 jaw self centering chuck is not a high priority. If you need concentricity use a 4 jaw independant or a collet or centers. A 3 jaw is about convenience, once you skim the workpiece it will be concentric to your spindle precision anyway.

Phil:)

With due respect Phil, it certainly is when you mount it! Otherwise one would just buy the cheapest Chinese POS chuck available, not bother to even turn a register, and simply bolt it to a piece of rough cast slag beaten into submission! Instead, clearly when mounting a 3 jaw, one tries instead to get it as concentric as is possible, so the only concentric errors when it comes time to actually use it, are those that occur due to the very nature of the design ie the scroll and the errors that design naturally introduces at different diameters, as you implied earlier.

PaulT
03-05-2011, 03:28 PM
Concentricity on a 3 jaw self centering chuck is not a high priority. If you need concentricity use a 4 jaw independant or a collet or centers. A 3 jaw is about convenience, once you skim the workpiece it will be concentric to your spindle precision anyway.

Phil:)

Actually many job shops use 3 jaw chucks even for work that requires concentricity. Most of them do this by using "soft jaws", typically aluminum jaws that you bolt in in place of the normal jaws (on 2 part chuck jaws) and then bore to size right at the size of your workpiece.

The other way to acheive this in a home shop is to use a "Tru-Adjust" style chuck that has set screws that let you dial in the concentricity at your workpiece diameter.

But if even you don't have a Tru-Adjust chuck there's a "poor man's" way to make yours into one. Take the chuck off the back plate, and cafefully machine the shoulder on the plate so instead of being a close slip fit its about 0.005" less in diameter than the mounting recess on the chuck.

Now when you need high accuracy, loosen the mounting bolts on the chuck, put an indicator on your workpiece and use a rubber mallet to tap the chuck in for zero runout and then tighten.

Ok, here comes the truth. I made my first home made tru-adjust chuck like this by accident when I blew it and over cut the shoulder on the mounting plate for a new chuck by about 0.005". But I had some work I needed to do and I decided to go ahead and use it that way and tapped it in for zero run out on that workpiece before tightening the bolts, and then realized how nice that was, its worked great since then.

Paul T.
www.power-t.com

PeteF
03-05-2011, 03:47 PM
Paul, that is the correct way to mount a chuck anyway ... well except machining the register undersized bit ;)

Even if the resister is quite a tight fit, the final step when mounting should be to loosen the bolts, chuck a bar in the jaws and indicate the high point. Then use a mallet to shift the chuck on the backplate. Of course the tighter you've machined the register, the less it will move when you do this. I think a lot of people don't know about/or do this final step, but even how you torque the bolts up will influence the runout ever so slowly. It's splitting hairs, but why not, it's probably the chuck most of us use the most so why not get it as close as the design allows!

When I first turned my register I took it down to precisely 100.00 mm and it needed to be hammered onto the chuck. However the chuck ran horribly. When I indicated the register it turned out that it wasn't perfectly concentric. I put this down to the carbide insert tools I was using as they weren't machining this CI too well at all. So I swapped out the carbide for a sharp HSS tool and took a very slight skim off the register to correct it. However it is now of course half a poofteenth undersize, but I'm happy to say not enough to consider machining it off and starting again, but I was very lucky as it wouldn't want to be any smaller.

philbur
03-05-2011, 06:24 PM
You can actually take the register right off with no consequences. With four bolts gripping the chuck onto the backplate it's going nowhere. The register on a 5" chuck I have is 1mm undersize and has given no problems in 25 years. As Paul points out it makes it a poor mans Adjust True.

Phil:)



........ but I'm happy to say not enough to consider machining it off and starting again, but I was very lucky as it wouldn't want to be any smaller.

philbur
03-05-2011, 06:38 PM
Pete, chasing 0.01mm concentricity on a 3 jaw scroll chuck is a pointless exercise. My 25 year old 5 has a concentricity error upto 0.1mm and Ive yet to have a situation where I need to do anything about it.

Phil:)


With due respect Phil, it certainly is when you mount it!

PeteF
03-05-2011, 07:55 PM
From what I've been able to ascertain the 0.01 mm runout is about as good as I'm likely to get for a 3 jaw and I'll just leave it as is. I'll just confirm more carefully that's what it is today, counterbore the socket head bolts and call the job done. I hadn't realised until today that the chuck's master pinion is marked so I'll check my measurements more carefully using that pinion to initially tighten the jaws.

Phil, you're right, the register isn't needed, but if it's turned accurately on the lathe the chuck belongs to, and it's a firm fit into the back of the chuck, the chuck should automatically turn to the limits of its tolerance. The "bolt trick" I mentioned above can then be used to help take out some of those tolerances. In theory all the bolts should be doing are stopping the chuck falling, rather embarrassingly, onto the lathe bed and spoiling your day. They shouldn't be part of the chuck registration per se. What I'd be concerned with is in the event of a crash ... not that I'd ever do that of course :rolleyes: if you're relying on 3 pissy 8 mm bolts for accurate registration, the chuck could be shifted off centre in the crash. Even done up tight I reckon a crash could move a chuck against it's backplate if that's all that was keeping it in position! I don't know, but that's just my gut feeling. The chucks are certainly designed to sit against a register, so I'm guessing the manufacturers do that for a good reason.

Pete

PeteF
03-06-2011, 03:57 AM
Phil, just to let you know I got the TOS chuck properly mounted and finished. I indicated a Compac 215 dti onto a ground shaft, tightened with the master pinion, and carefully set up the DTI to be correctly reading. The 215 is 0.002 mm per division and the needle barely moved! I'll be conservative and say 1/2 a division, so 0.001 mm runout. Obviously as you pointed out there's no guarantee that figure will repeat at any other diameter, indeed it almost certainly won't. However the object of the exercise was to get the chuck mounted as concentrically as possible and it's fair to say I'm extremely happy that it's that close, at least at this diameter of work. The spec sheet calls for a maximum of 0.025 (IIRC) runout throughout its range, which is still quite good in my book. If I need any better and I'll use the 4 jaw or collets as you mentioned. I'm certainly very impressed with this chuck and only wished I'd thrown my old one in the bin (ie sold it on eBay) sooner!

Pete

philbur
03-06-2011, 05:58 AM
As long as you are happy Pete, that's the main thing.

Phil:)

PeteF
03-06-2011, 06:44 AM
Who wouldn't be happy with that amount of runout from a non-precision 3 jaw! ;)

philbur
03-06-2011, 07:41 AM
You could have achieved the same result with the chuck you threw away.

Until you check at different diameters and different distances from the chuck your single reading tells you very little about your new chuck.

Phil:)


Who wouldn't be happy with that amount of runout from a non-precision 3 jaw! ;)

PeteF
03-06-2011, 08:15 AM
You could have achieved the same result with the chuck you threw away.

Until you check at different diameters and different distances from the chuck you single reading tells you very little about your new chuck.

Phil:)

That's complete nonsense Phil. Firstly you have no idea what my old chuck was like. In fact it was bell mouthed and quite worn out. It would have taken quite a lot to grind and rebuild it, and even then the scroll would still be stuffed! Using this new chuck is an absolute pleasure!

Secondly the manufacturer of this new chuck tells me in the spec sheet what the maximum runout is for the chuck throughout the range. Assuming the runout on the diameter I'm measuring now is at one extreme of the range, unless the manufacturer is lying on the spec sheet, the MAXIMUM runout I can now expect is the runout I'm measuring myself plus the runout the manufacturer quotes. So in contrast to what you assert I know quite a lot about the chuck now, as the WORST case scenario is 0.025 mm (the manufacturer's runout figure for the range, plus my runout figure of 0.001. Again that's worst case. In reality, even if the chuck was at the limit of its spec, the chances of me dialing it in at the very extreme of the chuck's runout limit is pretty remote, and there's a better chance it will be +/- from the current runout instead of simply + or simply -.

philbur
03-06-2011, 08:22 AM
Pete have you already forgotten about the poor mans Adjust True method.

Phil:)


That's complete nonsense Phil. Firstly you have no idea what my old chuck was like. In fact it was bell mouthed and quite worn out. It would have taken quite a lot to grind and rebuild it, and even then the scroll would still be stuffed! Using this new chuck is an absolute pleasure!

PeteF
03-06-2011, 04:18 PM
Not at all Phil, indeed I noted it above as the final stage of mounting a chuck. However I didn't replace the chuck because it wasn't turning concentrically, I replaced it because it was well and truly rooted and was missing a set of jaws. The lathe is over 40 years old and it's quite possible this is the original chuck sold with it. I have no idea how concentric that chuck turned, I probably would have measured it when I first got it, but as you said, if that was a real requirement I wouldn't be using a 3 jaw. HOWEVER, when replacing it why wouldn't I try to get the chuck as concentric as possible? It may reduce the occasions when I feel a need to use a 4 jaw or collets. Indeed it's come up so well I now know it definitely will!

When I first started the thread I actually hadn't read the spec sheet to see what the manufacturer was claiming. You know how it is; read the "instructions" as a last resort ;) Once I did, it was more accurate than I expected, since I never consider a 3 jaw to be a particularly accurate device. Subsequently I've spoken to another guy I know who, by pure coincidence, happens to have precisely the same machine with precisely the same chuck. His is a much older chuck, but he's also getting quite low runout figures.

Anyway, that's my 2 cents worth on the whole exercise. As I said, I've never considered a 3 jaw to be a particularly accurate tool, but by simply doing the job carefully (chances are it will be on there the rest of the time I own the machine, so overall a small investment in time) it's come up MUCH better than I expected. I have no idea how the chuck will hold up over time, but I've heard good things and so far I'd have no hesitation in thoroughly recommending the TOS chuck. Furthermore, if anyone is struggling with a worn out POS chuck with missing outside jaws ... just like I was, I'd rate replacing the chuck and fitting a new one accurately as right up there with a QCTP in terms of bang for the buck. Again, I'm now just kicking myself I hadn't done this much sooner!

Pete

J. Randall
03-07-2011, 12:26 AM
Pete, I don't think anyone is getting on you for getting a new chuck. I think what Phil has been trying to get across to you is you still don't know what you have until you finish checking your new chuck at different diameters. Then you may find out it really is as good as you think it is.
James

PeteF
03-07-2011, 01:33 AM
Hey James ... and what I'm trying to get across is the chuck manufacturer has already done that on my behalf and provided the figure in the spec sheet. That is the maximum runout figure that chuck will exhibit with any diameter ground rod chucked in the inside jaws (amongst many other specs they provided across the face etc etc). At some dimensions it may well be less than that figure, but never more unless it doesn't meet spec.

Pete

J. Randall
03-07-2011, 04:08 AM
Absolutely correct Pete, If the spec sheet is correct, and if it is that is great. But you still don't know that for a fact until you put it through its paces at other diameters. Until you verify, which you started to do with one diameter, it is just specs on a piece of paper.
James

PeteF
03-07-2011, 04:54 AM
Absolutely correct Pete, If the spec sheet is correct, and if it is that is great. But you still don't know that for a fact until you put it through its paces at other diameters. Until you verify, which you started to do with one diameter, it is just specs on a piece of paper.
James

Errr, umm, well yes I guess it is just specs on a piece of paper. However there are a lot of things I don't know for a "fact" and simply have to trust the manufacturer is telling the truth. I will often read the labels on food packages to read what the contents are, but sure as heck don't send all my food out for testing before I try and eat it!

I think some here may be missing the whole point of my post, and that was to try to mount my new chuck as accurately as possible, however I didn't want to chase gains that simply couldn't be obtained. Now it's bolted on it will either be concentric enough for the particular job I need it for or it won't, and if it's not I will remove it and use either collets or the 4 jaw. I'm sure as heck not going to sit there with a bunch of rounds measuring the runout at different diameters just to find out if a reputable manufacturer is actually a closet liar that's for sure!

Pete

gwilson
03-07-2011, 10:55 AM
I NEVER believe anyone's spec sheet anymore. The only way of knowing how accurate a chuck is,or how square a square is,is by testing it correctly yourself.

In the case of squares,I have a number of granite masters I've been lucky enough to find(good ,old USA made).

I'm suspicious that these days they just print op a stack of those spec sheets and throw one in with the item.

DougC_582
03-07-2011, 04:10 PM
I have a Chinese bench-top lathe with a 3-jaw scroll chuck, with the usual three pinions spaced around the body. I have noticed lately that the chuck will center considerably better if all three pinions are tightened to the same torque.

I looked around online after discovering this and did not find it mentioned anywhere, or in the few machinist books I have. I can only suppose that maybe tightening only one pinion pulls the ring/scroll gear off center?

The lathe was bought new so the chuck is not worn or damaged. If you have a similar situation it is something else free to try.

PeteF
03-07-2011, 06:01 PM
Doug, yes that's correct, just like some chucks (this new one of mine for example) have a master pinion marked, in my case with a "0", that will result in more accurate and repeatable movement of the scroll.

J. Randall
03-08-2011, 12:12 AM
Pete, everyone was just trying to answer your question, Is it worth trying to refine it anymore? The answer was in order to know you have to check it at different diameters. You don't want to do that, and that is fine, if you want to get your answer off a paper that was included with it, really what was the purpose of the post to start with?
James

PeteF
03-08-2011, 12:45 AM
Sorry James I'm not going to go any further with this. I posted a question I thought was quite reasonable. I have already said that at the time I posted the question I hadn't read the instructions/spec sheet. The specs also apply ONLY to the CHUCK and not to the combination of chuck with the backplate. The chuck itself could be guaranteed to be accurate to 2 micron at any diameter for all it matters, but it doesn't float in mid air, and when it's actually mounted a typical runout figure could be many MANY times that figure. I certainly didn't know as I have only ever mounted one other chuck before and it was a 4 jaw. Now it's mounted I've found it to be remarkably accurate, and continues to be so at the diameters I've subsequently used. If some want to poo-poo the whole exercise, feel that 0.10 mm chuck runout is all anyone should ever strive for, that chuck manufacturers are actually liars and insert spec sheets for no reason, or try to turn the whole thing into some bizarre argument, I'm afraid I'll need to leave them to it.

Pete

Edit: incidentally if you re-read through the entire thread again I think you'll agree the only person to actually try to answer the question I asked was RC when he posted this:

Probably within 0.05mm would be good.... The TOS 3 jaw I bought years ago (little 100mm one) had horrendous runout, but was probably within spec...

J. Randall
03-08-2011, 05:15 AM
Great Pete, glad you went ahead and used it at other diameters and found it satisfactory, all that stuff about food labels was a little bizarre. I bet you have good chuck that will give you lots of service.
James

luthor
03-08-2011, 06:56 AM
Pete, I have a new 125mm Tos 3 jaw that I will eventually mount on my little Hercus, it is what Tos call "superior" class with runout quoted as less than .025mm. Your chuck must be the same class or you were extemely carefull or lucky when mounting it.

PeteF
03-08-2011, 07:09 AM
I'm not sure. I was quite careful, but the final little bit was more ass-than-class I feel! This is the chuck anyway.

https://www.machineryhouse.com.au/Products?stockCode=C704

Pete

philbur
03-08-2011, 07:27 AM
I disagree, if you re-read through the entire thread again I think you'll agree that I answered your question in post #2.

Your question in post #1 was:

"Is it worth continuing to try to refine it more or does that sound about what I should expect?"

My answer in post #2 was:

"0.01mm is way better than you have a right to expect from a low cost 3 jaw"

Phil:)


incidentally if you re-read through the entire thread again I think you'll agree the only person to actually try to answer the question I asked was RC when he posted this:

philbur
03-08-2011, 07:34 AM
Exactly, that's why you need to check the run-out at different distances from the chuck. The chuck certificate is meaningless if you fail in mounting the chuck axis square to the spindle axis.

Phil:)


The chuck itself could be guaranteed to be accurate to 2 micron at any diameter for all it matters, but it doesn't float in mid air, and when it's actually mounted a typical runout figure could be many MANY times that figure.

Mcgyver
03-08-2011, 01:00 PM
somewhere in this thread i think read about the chuck perhaps not being a tight fit on the backplate :eek: That's a problem, maybe needing a new back plate. iirc the correct way is a slight interference fit/shrink fit. the back plate is of course is finished while installed on the lathe so should be perfect to the spindle. Set pin 1 in hole 1 so you can repeat the set up.

The reason is the bolts are insufficient to locate the chuck to the backplate and it will move. Don't believe me? Ever heard of tramming in a b-port

Here are good instructions

http://www.lathes.co.uk/latheparts/page7.html

After that rely on the chucks accuracy, soft jaws, adjust true mechanism, alter sequence so everything is done at one set up, tapping with a hunk of copper or use a four jaw or collets or faceplate - which are able to achieve more concentric accuracy than the three jaw.

philbur
03-08-2011, 03:28 PM
It is a well recognised and accepted method of making a 3 jaw adjustable. If you crash badly enough that it moves on the back plate then it takes all of 1 minute to reset the chuck true. In addition the movement will potentially have absorbed energy that might otherwise have gone towards bending your spindle.

Other items relying on a couple of bolts and friction include: the cutting tool in the tool holder, the tool post on the compound. and the compound on the carriage, why make an exception for the chuck.

Phil:)


somewhere in this thread i think read about the chuck perhaps not being a tight fit on the backplate :eek: That's a problem, maybe needing a new back plate. iirc the correct way is a slight interference fit/shrink fit. the back plate is of course is finished while installed on the lathe so should be perfect to the spindle. Set pin 1 in hole 1 so you can repeat the set up.

The reason is the bolts are insufficient to locate the chuck to the backplate and it will move. Don't believe me? Ever heard of tramming in a b-port

Here are good instructions

http://www.lathes.co.uk/latheparts/page7.html

After that rely on the chucks accuracy, soft jaws, adjust true mechanism, alter sequence so everything is done at one set up, tapping with a hunk of copper or use a four jaw or collets or faceplate - which are able to achieve more concentric accuracy than the three jaw.

Mcgyver
03-08-2011, 04:52 PM
Other items relying on a couple of bolts and friction include: the cutting tool in the tool holder, the tool post on the compound. and the compound on the carriage, why make an exception for the chuck.
)

because you want it to stay put over the long haul, the others are temporary. I disagree that's it acceptable...error in the three jaw will vary over the scroll in any event. You don't have to crash anything, shouldn't ever happen with a manual lathe but overtime it will shift just like the mill head does...if there's lots of play that's how much it can shift. In the privacy of ones home you can do anything you please, but its not good engineering practice to have a permanent assembly relying on bolts and friction rather than a positive register such as plate to chuck, pins etc.

PeteF
03-08-2011, 05:21 PM
Yes I agree, I think over time a chuck would move about about if it wasn't a firm fit on a register. I don't know for sure, but that's my theory anyway.

It's not normally necessary to get a new backplate BTW, you can just machine off the register and start again, and can typically do this a number of times for a backplate before it gets too thin. How many times will depend on how wide you make the register, but even quite deep, you should get at least 4 goes at it.

In my particular instance, the fit on the register is only VERY slightly undersized. Initially I had it at precisely 100.00 mm (at least at the places I was measuring it), and it was a press fit into the chuck as it should be. However the carbide wasn't cutting evenly (I'd never struck that before) and therefore it wasn't perfectly round. I swapped it for some sharp HSS, blued the register and took a skim off it, just enough to remove all the blue. That bought it down to 99.87. Not a huge amount undersized, but as I mentioned I wouldn't want it any smaller. It's still a slight push fit in the chuck. HOWEVER I wasn't especially careful with my drilling of the holding bolt holes. I incorrectly presumed there would be enough clearance in the hole to mean that it wasn't especially critical. So I marked the positions with 8 mm transfer screws and drilled them, but not with any extreme care. Of course in doing so they were slightly off position, to the extent that even after opening up to larger clearance holes, I had to take a few licks with a round file to shift 2 toward each other. I'm still not quite sure what happened there, as I said I didn't use extreme care, but didn't exactly use a portable drill and sharpie either! My "fix" was a bit of a bodge, but since it was before counterboring them at least nobody can see my mistake :D The end result is that 2 of the bolts are now very firm in their bores along one side, and when combined with the register means in my case that chuck isn't going anywhere.

However I agree, the register should be a press fit onto the chuck, and anything less is simply a bodge. ... of course I COULD always claim drilling the holes off position was deliberate to counter my undersized register :p Would anyone buy that story?

Pete

philbur
03-08-2011, 05:34 PM
It is a well recognised and accepted method of making a 3 jaw adjustable.

Phil:)


because you want it to stay put over the long haul, the others are temporary. I disagree that's it acceptable...error in the three jaw will vary over the scroll in any event. You don't have to crash anything, shouldn't ever happen with a manual lathe but overtime it will shift just like the mill head does...if there's lots of play that's how much it can shift. In the privacy of ones home you can do anything you please, but its not good engineering practice to have a permanent assembly relying on bolts and friction rather than a positive register such as plate to chuck, pins etc.

philbur
03-08-2011, 05:39 PM
Only 0.13mm. So how much movement on the backplate would you class as unacceptable.

Phil:)


That bought it down to 99.87. Not a huge amount undersized,

PaulT
03-08-2011, 05:49 PM
. . . but its not good engineering practice to have a permanent assembly relying on bolts and friction rather than a positive register such as plate to chuck, pins etc. . . .

So I guess we should all throw our Kurt (and Kurt copy) vises in the ocean because they rely on bolts and friction? (yes, I know you can use the keys against the table slots, but not if you want it dead square)

If used with judgement there are times when this method is not only ok, its preferred as you can get it as accurate as you want or need for a particular job.

If you are the only user of the machine, you'll know when its taken a hard hit a probably needs realignment. If you're not the only user of the machine, put an indicator on it to check it before use.

The same approach has worked fine for me for many years on my "poor man's" tru-adjust chuck. If I need dead on I check it and adjust as needed, otherwise just fire it up. I've never seen it move by itself, but I've been careful and have never crashed this chuck.

Paul T.
www.power-t.com

PeteF
03-08-2011, 05:50 PM
Only 0.13mm. So how much movement on the backplate would you class as unacceptable.

Phil:)

As I said Phil, that's about as small as I would want it. It was an error, but still a slight push fit so one I felt I could live with. Had the carbide been cutting as it should, and the initial register round, then my initial 100.00 was what I was looking for. I shouldn't have persisted knowing the tools weren't cutting correctly as I paid the price. Incidentally I tried swapping inserts and all manner of speeds and feeds and couldn't get them to cut properly. Strangely I bought another backplate from the same retailer in the US and it machined beautifully using the same tools! NFI why the difference. Chinese QC?

I have no desire to make a bodge "adjustable" 3 jaw. That's not how a properly designed adjustable works, and they have adjustment screws to shift the chuck and help maintain its precise position. I already have an adjustable 3 jaw ... it's a 4 jaw ;)

Pete

philbur
03-08-2011, 05:50 PM
It is also temporary, you can use the adjustment to re-center at different diameters, I do.

Have you ever had a parting tool dig in, that is a manual crash in my book.

After 25 years with a 5" three jaw with a reduced register on the backplate and no slippage I would say experience says otherwise.

What's a bridgeport got to do with anything.

Phil:)


because you want it to stay put over the long haul, the others are temporary. I disagree that's it acceptable...error in the three jaw will vary over the scroll in any event. You don't have to crash anything, shouldn't ever happen with a manual lathe but overtime it will shift just like the mill head does...if there's lots of play that's how much it can shift. In the privacy of ones home you can do anything you please, but its not good engineering practice to have a permanent assembly relying on bolts and friction rather than a positive register such as plate to chuck, pins etc.

PeteF
03-08-2011, 06:06 PM
Phil, with due respect what you're saying now seems at odds to what you said previously.


My 25 year old 5 has a concentricity error upto 0.1mm and Ive yet to have a situation where I need to do anything about it.

I think the bottom line is we all have our own personal tolerances and limitations of what we consider acceptable. Given that I was going to be using this particular chuck a lot I wanted to make it as accurate as I possibly could, but that's not to say everyone wants to or should. As has been pointed out many times already by numerous people, a 3 jaw isn't a precision tool and the concentricity will vary depending on the diameter of the jaws, simply due to the scroll design. However I have no regrets about spending the time I did on getting this accurate ... even if the final little bit was more ass than class and probably nothing more than a coincidence :D

Pete

philbur
03-08-2011, 06:45 PM
Not at all. Across its complete range of clamping it can be out by as much as 0.1mm. That's why, when I need a degree concentricity I use the centering adjustment to center the work piece. It's partly due to the ability to adjust that has made it unnecessary to do anything with the chuck over the last 25 years. The adjustment can also avoid the need to de-mount the 3 jaw, mount the 4 jaw and then back again.

The adjustment completely overcomes inaccuracy in the scroll etc. You can have concentricity better than a collet, and at any diameter. With a tight register to prevent mythical slippage on the backplate you get what you get. So my 25 year old nackered chuck on a reduced register is actually more useful than your spiffy new 0.001mm runout (at a single diameter) chuck on a tight register.

I have an article from "Model Engineer" regarding the modification if you would like a copy. The subject comes up every so often, usually resulting in a similar discussion to this one. I decided to make the modification based on an article written by Tubal Cain. If it was good enough for him it was good enough for me.

Phil:)


Phil, with due respect what you're saying now seems at odds to what you said previously.

luthor
03-08-2011, 07:21 PM
Pete, what make of lathe is this chuck mounted on and what type of spindle bearings does it have to be able to maintain a repeatable 1 micron TIR ?

Mcgyver
03-08-2011, 09:34 PM
So I guess we should all throw our Kurt (and Kurt copy) vises in the ocean because they rely on bolts and friction? (yes, I know you can use the keys against the table slots, but not if you want it dead square)


Paul, you mustn't have understood the point or you're just being silly. A vise on the table is frequently moved about, removed etc - its a temporary set up.....and when it really, really counts you'd indicate first because you know under load it can shift from the last time it was set up, right? A chuck on the other hand you want to do once right and forget about it.

Maybe a particular user takes such light cuts it'll never matter....but it being proffered that many install backplates this way does not mean its the right way to do it. It may well just be an excuse for getting a proper shrink fit with a larger size item; something that is not completely trivial especially for a beginner probably without large mics. Rohm or Pratt or Yuasa saying do it would be credible, general hearsay is, well hearsay.

The relevance of the bridgeport was as i said
it will shift just like the mill head does which is a a much more robust bolt & clamp mechanism than a lathe chuck/backplate, yet still it readily moves about.

As i say, have at it in any manner you choose, and if the lathe is small and light cuts are taken it may not ever matter, it won't to me. I can see how one might argue this does create a 'poor mans adjust true', there's merit in that for some I suppose....but for the benefit of others reading it was worth pointing out its not the correct way to install a back plate. There are so many other ways to achieve concentricity to whatever degree is desired, and that it is poor engineering to assembling things like that without a register makes it an argument I'd not give much weight to. Each to his own I suppose. :)

PeteF
03-09-2011, 04:25 AM
Pete, what make of lathe is this chuck mounted on and what type of spindle bearings does it have to be able to maintain a repeatable 1 micron TIR ?

It is just a Hercus/SB9 plain bearing. I was using a compac 215 DTI (0.002 mm graduations) set horizontal to be correctly measuring. I think from memory it was a 12 mm ground bar I was indicating onto, however I'm away for work at the moment so can't confirm that. The DTI was indicating under 1/2 division movement, but I had a bit of a gut feeling it may have been sticking ever so slightly, I haven't noticed that with this indicator before, but called it 1/2 division movement to be conservative.

Actually now you ask, if I get some time I'll set the same piece of silver steel in a collet & compare. Normally however I've found them to be extremely accurate.

Pete

macona
03-09-2011, 04:40 AM
McGuyver is not referring to things like vise to table locations where it is intended to be demountable. He is referring to parts like the joints in the back jaw of a kurt and the base where they have a key to maintain registration or dowel or taper pins on a taper attachment where it bolts on to the carriage. Or the fixed jaws on my fixture plates. You could never trust mere friction to keep the jaw square to the axis of the machine, thats why the have precision bored bushed holes in the plates and fixed jaws.

PaulT
03-09-2011, 01:19 PM
McGuyver is not referring to things like vise to table locations where it is intended to be demountable. He is referring to parts like the joints in the back jaw of a kurt and the base where they have a key to maintain registration or dowel or taper pins on a taper attachment where it bolts on to the carriage. Or the fixed jaws on my fixture plates. You could never trust mere friction to keep the jaw square to the axis of the machine, thats why the have precision bored bushed holes in the plates and fixed jaws.

There is no absolute on this, on I would claim that on the majority of components that use alignment pins or keys, the purpose of these pins is to align the component so that it is then clamped via friction in the correct location. Most of the times these pins are not designed to take any shear force to keep the component in alignment, the clamping is intended to do that.

Any side force on the alignment bushings and pins is bad as it diminishes their accuracy, you don't want this to happen, the clamping force needs to be large enough so that the pins never see any shear forces.

Is this always the case? No, sometimes the alignment devices are also designed to take shear loads. But I would claim this is not intended most of the time.

Take the case of the keys on the bottom of a Kurt vise. By design they don't fully fit the table slot width, you push the vice until they flush with one side of the slot and then clamp.

So clearly if the keys were meant to keep the vice from moving, they would only work in one direction. You wouldn't want a lot of force on those keys and the table slot surfaces, it would just eventually diminish their accuracy.

You will see this on mass producted items also. A 4 speed transmission I rebuilt had the side cover plate aligned with roll pins. The location of the cover plate is very important so the shifting fingers are actuated correctly. But roll pins won't take hardly any side force at all, they are basically rolled up pieces of sheet metal.
The roll pins get the side plate in perfect alignment during installation but the bolts on the plate then clamp the plate with enough force to keep it from moving from the shifting forces, the roll pins won't help with that at all.

Paul T.
www.power-t.com

PeteF
03-09-2011, 08:04 PM
Take the case of the keys on the bottom of a Kurt vise. By design they don't fully fit the table slot width, you push the vice until they flush with one side of the slot and then clamp.

So clearly if the keys were meant to keep the vice from moving, they would only work in one direction. You wouldn't want a lot of force on those keys and the table slot surfaces, it would just eventually diminish their accuracy.


Hmmm, your vice analogy now raises another question in my mind; how tight should the locating keys for a mill vice be? I just fitted a new vice and the keys with it were the wrong size. So I ground them down but then had then very slightly undersized so I needed to push the vice in one direction to properly locate it, just as you describe Paul. I wasn't happy with that so made some new (much longer ones) keys that are now a slight push fit in to the slots. I feel that is a much better arrangement and although I didn't bother hardening them, I think for the number of times the vice is removed and replaced, the wear won't be a significant factor.

Pete

PaulT
03-09-2011, 08:14 PM
Hmmm, your vice analogy now raises another question in my mind; how tight should the locating keys for a mill vice be? I just fitted a new vice and the keys with it were the wrong size. So I ground them down but then had then very slightly undersized so I needed to push the vice in one direction to properly locate it, just as you describe Paul. I wasn't happy with that so made some new (much longer ones) keys that are now a slight push fit in to the slots. I feel that is a much better arrangement and although I didn't bother hardening them, I think for the number of times the vice is removed and replaced, the wear won't be a significant factor.

Pete

I have a lot of fixtures for different parts that align with pins or keys to the slots on my table. I like these to be a very close slip fit, so no wear occurs on the table slot during insertion. I then push and hold the fixture up against the rear slot surface, tighten the hold down bolts and I'm good go to.

This method gets you pretty square and I also use it with the plate I use to quickly mount my vice.

For anything super critical (like making a new fixture plate) I'll sweep the vice jaw with and indicator and tweak it dead in if need be.

Paul T.
www.power-t.com