I use the toolpost / side of the tool and finish with DTI

You can do the job 98% by eye

It's like sex

The more you do it the easier it gets

Depends on what kit you have to hand. I generally use two centres and a DTI

Regards,

IanT

+1

And practice!

I position the job roughly by eye, sometimes using a centre in the tailstock as a reference, sometimes with the lines on the chuck, at this stage not tightening the jaws.

Easiest and most accurate to measure offset with a DTI, but it can be done by eyeballing the gap between a cutter and job.

With two jaws horizontal, set the DTI. Then rotate through 180 degrees to measure the difference. Turn the keys in opposite directions to halve the difference, tighten one while slackening the other. Rotate back through 180 degrees, measure the new difference and halve again. Then set the other two jaws horizontal and centre them with the same process.

At this point, the job should be more-or-less accurately centred. Spin through a full turn to see how good or bad it is, and repeat the 'halve the DTI difference' with two chuck keys as necessary.

Tricky bits;

• During adjustment making the jaws tight enough to hold the job without being so tight as to make halving difficult.
• Learning to turn two chuck keys together and in the right direction! (I still get this wrong occasionally)
• Tightening up at the end without disturbing the centre setting.

I cracked it by spending a few hours doing nothing but centring various objects in my 4-jaw. Initially I had to think every move out, now it comes naturally.

Dave

Edited By SillyOldDuffer on 01/09/2020 11:44:34

By eye – use the tool or tool holder to give you an indication. Then, digital DTI. It is easy (if you have one). Set to zero at jaw 1, rotate to jaw 3. DTI will give you a + or – error. Halve it, and adjust jaws to correct. Likewise jaw 2 and 4.

Practice on a bit of round bar first. As others have said "prctice makes perfect."

Never felt the need for 2 chuck keys & lots of practice works wonders.

Tony

Instead of 2 centres I have a bit of 3/16" rod about 12" long with a point on one end. Grip the non pointy end in the drill chuck in the tailstock and apply DTI towards the pointy end. Saves swapping from chuck to centre. One of these days I'll get round to loctiting a bit of bigger stuff to the non pointy end so that the DTI can be applied when the eccentricity is higher. I'll put a flat on it so it doesn't roll off the shelf, but not run the DTI on the flat.

When applying it to the centre pop, I wind the tailstock out until the handle is at the back before locking the tailstock, The weight of the handle then keeps the point in the mark

Just tried Hopper's method only to find it took longer. Despite understanding the logic I slowed down just because the method is slightly different from what I'm used to. Hard to break old habits I guess. I shall persist : Hopper's way must be easier!

Dave

This is also my preferred method, have also used the two chuck keys method.I find this one the easiest.

George.

Here we have quite neat way of centring square bar in a 4 jaw.

Bill

A quick way to get very close without needing a DTI uses the cross slide dial to measure the error and a support block between the bed and chuck jaw to hold the jaw exactly horizontal.

Start by putting the job in the chuck at your best guess to central. Choose a jaw to start with and set it horizontal using the block. Arrange some sort of rigid probe on the tool post with a vertical face to contact the work. For example a hefty rectangular tool set backwards in the tool post with its end dead vertical works fine. If you have a QC post the vertical face of an empty holder may well work. The driil carriers from Dickson systems work fine for this. Nice to find a use for it!

Run the cross slide forwards until the probe lightly touches the job. If you don't trust your feel or if doing a small job in a bigger, less sensitive machine, trap a piece of cigarette paper or 1 thou feeler gauge just enough to be stiff to pullout. Note the cross slide reading.

Pull back, rotate the chuck by 180° and repeat. Set the cross slide to halve the difference and adjust the chuck jaws so the work just touches the probe.

Rotate 180° and verify that the cross slide readings are very close. Adjust again if more than 10 thou out.

Repeat process with the other pair of jaws.

Given a modest level of care its easy to centre the job to better than 10 thou TIR first time round. With practice right first time may become normal for anything other than the most exacting jobs.

Go round again for the final adjustment if needed. When finishing off you may prefer to leave the probe sitting very close to the work and make the final tweak using feeler gauges instead of the dial. Expect something under half a thou TIR to be easily achievable.

I imagine the process works well with a DRO.

This was the method I was shown for square stock. Took 30 odd years to twig that it would work well for round work too. By which time I had a nice stock of dial indicators.

Clive

In that case it's not surprising 'cos I wrote the MEW article you saw. My link was the actual article.