Turning an Offset In A 3-Jaw

[rake60]

Somewhere I saw a formula for shimming one jaw of a 3-jaw chuck to cut an offset. The formula was "Offset = 2/3 Shim Thickness"
The valve eccentric on my last model called for a .050" throw. I hate indicating in a 4-jaw, so I decided to give it a shot. I needed a .075 shim. What I has on hand was some 1mm brass stock. Two pieces stacked = .078"
Maybe I was just holding my mouth right, but the total run out of the piece was .101" Close enough for me. It worked out perfectly.



I'm sure it's an old trick, but one I just found. There will be no more indicating
those little eccentrics in my basement!

Rick

[mofugly13]

I had not heard of that trick, thank you!

[mklotz]

The 2/3 formula is wrong. You can calculate the correct shim using the program from the ECCENT archive on my page. A more secure method utilizing a tube is described there too.

[rake60]

OK Marv, I got your DOS program to work.
It comes up with .075" packing for a .050 throw in .625 stock with .5 wide jaws.
.050 = 2/3 of .075 Are you sure 2/3 is wrong?

Rick

[Joel]

Neat trick.

[BadDog]

Yes, 2/3 is definitely wrong as a general solution.

[rake60]

OK you've got me experimenting.
Using the ECCNTUB program. It came up with .07798 I started turn a piece to that when it dawned on me that was essentially what I had with the two thicknesses of the 1mm brass shim. With that I had .101 total run out or .0505 throw. A half a thousandth I can live with.
I have only been a machinist since 1978 so I am still learning. EVERY DAY for the past 19 years.

Rick

[speedy]

Thanks for the rough n ready solution Rick.
But....I reckon by the time that you had consulted Marvs Eccent program, found the correct shim and set up; I could have set up the ecentric in my 4 jaw and had a cup of tea/beer or coffee . Learn the 4 jaw, it`s a breeze.

[Evan]

At small diameters it will also depend on the type and shape of the chuck jaws. The work holding face is usually curved and how effective the shims are depend on if they bridge that curve or conform to it.

I was just using the same technique a couple of nights ago to make an offset temporary dead center to use in making a taper pin.

[wierdscience]

Neat trick,used it before myself,it's great for small offsets.

Coolest Idea along these lines was a friend who had a chuck with two-piece jaws.He had an extra jaw he made with a built in screw adjustment.You could swap the jaw out,make a rough setting,chuck a part and check it with a DTI,then advance the jaw accordingly for the offset needed.Once set he could lock the jaw down and turn as many pieces as he needed.

I've been wanting to build one since I saw it 10 years ago,but it is still on the roundtoit list

[A.K. Boomer]

You wouldnt believe the combo's you can get by just swapping the jaws all around and also starting them into the scroll at intermittent sequence--- if you record all the offsets you may get close enough for building things that your going to create -- if your trying to match up to a standard size its going to be few and far between, dont trust it like you would a normal chuck-up, sometimes the material will be on the corners of the jaws if its a big off set and it also makes for different spaced holding which is not as stable, if you git hit in the forehead cant say as i didnt warn ya.

[mklotz]

OK Marv, I got your DOS program to work.
It comes up with .075" packing for a .050 throw in .625 stock with .5 wide jaws.
.050 = 2/3 of .075 Are you sure 2/3 is wrong?

Rick

The relevant equation looks like:

p=1.5*e-r+0.5*sqrt(4*r*r-3*e*e+2*e*w*root3-w*w)

It may have come out to be 2/3 in your particular case, but it's obvious from the equation that, given other choices for the parameters, it won't always be 2/3.

Despite having written this program, I think this is a downright awful way to do eccentrics. The tubular approach described in the archive is better, but a 4jaw is still the right way to do it.

[rake60]

There's no doubt that a 4-jaw is better, but if there's a quicker, easier way
to make a non-critical part I'll use it.

Actually this became a major topic of discussion at work today. Of course one of the engineers set off to prove it won't work. He came back a couple hours later and said on the small scale I'm working on at home it does works.

Now if I were making IC engine parts or something more fussy I'd probably use the 4-jaw. Unless I shimmed it in the 3-jaw first and got the perfect answer from the indicator.

Rick

[Alistair Hosie]

good idea thanks for sharing will try that myself some day. regards Alistair