Improve 3-jaw chuck repeatability

I looking for a way avoid to loose concentricity every time I remove the workpiece from the chuck. Also I want to avoid to remove the 3-jaw chuck. It is fine if I find a solution for diameters up to 20-30mm. My first idea is to hold this kind of ER collet in the chuck:

There are two options I could think of. I could try to scrap the faces of the hexagonal part until there is no runout. And next time put the ER holder in exactly the same position. I suppose that this will take a lot of time.

The second option is to make an ID grinder for the tool post and to cut a little the jaws while they are opened exactly at the same diameter as that hexagonal part. And of course the jaws should apply pressure on something during grinding.

Does any of these two ideas have any chance to work? Because if not then I have more 

Edited By Sonic Escape on 07/08/2023 16:29:32

Mount an ER collet chuck directly on the lathe mandrel concentricity without the effort.

Robert Butler

It's well worth doing a search on the site (use facility on the homepage not the one at the top of these pages) . ' improving accuracy of 3 jaw ' will give you numerous previous posts as this topic has been discussed frequently before.

No 3 jaw will ever be totally accurate!

Regards

Edited By Mike Hurley on 07/08/2023 16:57:14

What did industry use?

Big collets

Just remember that the part, when mounted in a collect, will already need to be completely/perfectly round.

Buy a griptru chuck.

regards Martin

Even expensive 3-jaw chucks in good condition are a poor choice when repeatability matters. Griptru maybe, which are adjustable. A 3-jaw is speedy, but the price is repeatability, which drops off as the jaws, slots and scroll wears.

Better alternatives available. I put an independent 4-jaw chuck on when I need repeatability, because the job can be accurately centred with a DTI. Some machinists always work with a 4-jaw, because, with practice, work can be centred quickly and repeat error eliminated. Collet chucks are the bees-knees for certain workflows, I have an ER32 collet chuck which is very useful at times, but if I was serious 5C collets are better because they can hold shapes other than round. A full set of 5C collets is expensive.

My hex collet block iand 3-jaw are moderately repeatable, but test yours to find out how good or bad yours is in your chuck. I don't think scraping the block is a good idea, because repeatability depends on what's wrong with the chuck : probably better to replace the chuck, or the jaws, or grind the jaws.

Main use of hex and square collet blocks in my workshop is moving work between mill, lathe and bench, not overcoming of my 3-jaw chuck's limitations. Good for what it does, but that doesn't include accurate repeatability.

Dave

I have made my 3-Jaw chuck adjustable, after I determined it has various levels of not accurate(Yes their is the grinding effort of the jaws as well) as you wind and clamp it on various dimentions. I saw an interesting suggestion by well know precision masjinist and that is to use the chuck key in just one slot to clamp it down, dont use the other two slots was his suggestion.

The other thing is 3 or 4 jaw chuck if you clamp a round bar with some lenght you can tap it in place on the far end, that say something about the clampling.

I think if you need accuracy, you also need the ability to readjust it accurate if its not, when re_inserted, a problem of time if you have loads of the same thing to make in a not so accurate machine. So how do one readjust a collet unless its mounted in an adjustable holder-?

Edited By Chris Mate on 07/08/2023 19:47:05

One cannot … that’s why there are very good collets available to those who can justify the cost.

The rest of us just have to find work-arounds.

MichaelG.

Have you actually measured how bad (or not) your 3 jaw chuck is? While you're at it, determine whether it needs to be any better for the work you do. There's no point stressing about micron accuracy if your parts would work if measured with a caliper.

Adding another joint in your part holding is hardly likely to make for better/easier setups....

Edited By Nick Wheeler on 07/08/2023 22:02:08

Most 3 jaw chucks have some run-out unless it is adjustable and why a 4 jaw is used for accurate alignment. Another option is to use soft jaws which are machined in place to suit the diameter of the work which you are holding. (good method if multiple parts of the same diameter are being turned).

I use an ER collet with a morse taper and can move it (with work) from the lathe to rotary table (same morse taper) and back again, without losing concentricity.

Edited By Paul Lousick on 08/08/2023 00:23:07

Soft jaws have the advantage that you can also machine a back stop in the jaw so that if there is a machined lip on the part you are working on it can be repeatedly replaced the same depth into the chuck as well.

Grinding the jaw faces to hold a hexagonal part may not give the accuracy you are expecting because the 120° faces of the jaws may not be accurately machined and also you will have a sharp edge to the ground faces which will damage the faces of the hex part or in turn be damaged by the hex.

Martin C

Guys, I think we need to be a little more rigorous with our terminology here. Most 3 jaw self centering chucks actually have good repeatability, it's the concentricity that is variable for different diameters. If you measure the run-out of a piece of round stock, take it out and replace it using the same pinion and orientated against the same jaw then the run-out will measure pretty much the same as before. You can use this property to make an accurate and repeatable split collet bored in situ.

In Sonic's case, I assume that he has a 3 jaw chuck that is a trial to remove and re-attach: if so I can understand his pain. In my experience ER collet chucks are not that hard and can be turned with a carbide tipped boring bar so I would mount the hex collet block in the 3 jaw, true up the tapered collet seat using the top slide set to the correct angle using a dti. Mark one of the hex sides so that you know which side goes against which jaw and remember which pinion is used to tighten the scroll. You should now have good concentricity and repeatability. Worth a try I think.

Rod

That is a better idea than trying to grind the jaws or the hex! I think I'll try it. Meanwhile I made some measurements with a test bar:

The maximum runout is 0.130mm at 20mm away from the chuck. I repeated the measurements from the video for different levels of forced applied to the chuck wrench. In any case the runout didn't exceeded 0.150mm at maximum distance from the chuck. It is interesting that for a moderate force used to close the chuck the runout is bellow 0.08mm. How good are these numbers?

Edited By Sonic Escape on 08/08/2023 11:16:35

I notice there is a '0' mark adjacent to one of the key socket holes in the chuck. Isn't one supposed to tighten this one first to minimise runout ? (Another of those things I have meant to investigate the effectiveness of but just never seem to get around to doing.)

Gerry

I didn't noticed it. I just tried now to use that socket hole. I got repeatedly 0.2mm runout! With the other socket holes the runout was again bellow 0.150mm. So maybe that mark means to don't use it

Instead of having to turning the hex sides on the ER holder get one that is already round.

Compared to what I get yours are at least 10 times worse. On a piece of ground HSS I'm reading 0.0005" which is approx 0.013mm

One other thing to consider is that by putting a collet block into the chuck you are moving any workpiece further from the bearings this may result in chatter when taking a large cut or in mind of your other thread when parting off. Your chuck jaws also look quite worn so there is a good chance of them being "bell mouthed" where the gripping surface becomes worn towards the end of the jaws reducing their grip

That's the trouble with folklore......

The numbers aren't good. With a ⌀25mm test bar my Chinese lathe and chuck run-out is 0.06mm at 50mm out. It was a little better when new. The exact figure varies a little depending on how the chuck is tightened.

My chuck is inexpensive and tightening it feels gritty. The scroll is not as well-finished as it might be, and could be soft. That my run-out is slowly getting worse is probably due to wear on the scroll and jaws. I'd expect a more expensive chuck to achieve noticeably lower run-out from new, to turn more smoothly, and to resist wear for longer, but they don't last forever! Don't forget a second-hand tool may have been worked hard or abused by previous owners.

I suspect your chuck is worn. With luck it's just the jaws, which can be reground or replaced, but maybe a new chuck is required, the headstock is out of alignment, or the spindle or bearings are damaged.

Check the chuck and spindle are OK before assuming collets are the answer. It's easy to waste lots of time and money fixing the wrong problem. You can guess how I know!

Dave