Grinding the Unimat three jaw chuck

I bought a second hand Unimat 4 from a colleague in somewhere back 2019. I gradually dug into the learning process of working on the lathe with metal and the various tools.

At a certain moment I turned a brass part, flipped it over and turned the other end and much to my surprise the two ends did not line up. There was a clear step visible of about 0.2 mm. I installed my dial indicator with which one can measure the concentricity of parts up to 0.01 mm (0.4 thou of 4 "tenths"). The spindle over the chuck was okay with a runout of about half a notch (0.005 mm or 2 "tenths"). The outer rim of the three jaw chuck also ran beautifully concentric (even if not being so very critical) but when I inserted a brand new 2 mm centre drill this happened

Almost 0.2 mm runout is hopelessly inaccurate for a lathe.

So I started reading about the ways to improve that. Grinding the jaws was almost the first word that I found on the internet. Grinding means you introduce some sort of grinding tool into the open jaws of the running lathe and the three jaws will rub against that tool and the surfaces of the jaws will eventually be ground to the same distance from the centre of the spindle. There are numerous videos on the subject but it proved kind of hard to find one that offers a method suitable for a very small, non-professional lathe. A very important finding of all videos is that the jaws should be ground under load. It is no good to grind them just like that as they have some play in the scroll of the chuck. Loading the jaws by simply putting a part in the chuck is no good as you can't get tot the jaws with the grinding tool. Now what? There are basically two solutions for that: the ring method and the spacer method. The ring method relies on threaded holes in the chuck jaws. You bolt a ring to the three holes and tighten the chuck inwards. Simple. Uh, oh, my chuck does not have holes in the jaw. The jaws are extremely hard and a feeble attempt to drill holes myself hardly scratched the surface.

So the spacer method is left.

The spacer method needs three equal spacers to be inserted between the outer edges of the jaws and then the jaws are tightened inwards. So the first job to do is to make these three spacers.

I had a 3 mm thick steel washer from which I sawed the spacers. I milled them in my Proxxon MF70. First one side flat for each of them individually. Then I put all three together with the freshly milled reference surface down in the vise and tapped them to sit absolutely true. Then I milled the tops of all three, aiming for a width of about 8.5 mm. After removing them from the vise I deburred them and measured them. There was some variation though and I sanded them on 320 grit emery paper with a bit of oil on it. With some patience I got them within 0.02 mm

After these photo I took the widest and sanded once more and reduced the difference to 0.01 mm. Close enough for me.

In the preceding days I had ordered this special Proxxon tool. It is advertised as a grinding tool holder but in fact it can be used for any job that needs a Proxxon drill to be combined with anything in the lathe. For €15 it cannot be a bad investment.

For reference: Proxxon no 24098

Below you can see how it holds my Proxxon drill in the lathe. It just a matter of adjusting the height of my Quick Change Tool Post holder to get it on centre

Good, that it said, it was time to get the three spacers in place. Needs to be said that it really required some dexterity to hold three spacers in the right spots and then tighten the jaws. But after some trial and error I got them right with by first clamping the first spacer very lightly, then with some pressure inserting the other two and when all three were correctly placed tightening the jaws real hard. Take no chances, these parts will rotate quickly and one of them flying out will not only ruin your setup but may present a serious hazard. So do a trial run with the chuck loaded this way and make sure to keep all body parts well away from potential trajectories. My three spacers stood the test.

Now it was time to do the real work. Preparations took all day, the actual work was don in five minutes.

It looks simple enough. A few notes though.

• The grinding tool I bought is from Dremel. It ran as good as true so I did not need to dress it with a diamond dressing tool. Later I saw it was officially only for non-ferrous metals, but it behaved extremely well.
• Take care that the opening of the jaw is big enough to accept the grinding tool. If it does net fit your spacers need to be bigger.
• On the other hand see to it that to touch nothing else than just the tip of the jaws.
• The grinding tool should be into the Proxxon motor tool's chuck as much as possible but on the other hand stick out enough to reach the most inward part of the chuck jaw tips.
• COVER YOUR LATHE! The grinding generates quite a bit of dust. This dust seems to be detrimental to the ways of the lathe. So try to collect it as much as possible with old cloths covering the ways.

A first measurement

Even with the jaws full of grinding dust I could not not resist the temptation to see if matters had improved. And really the results were encouraging, from 0.19 mm to 0.03 mm is a good improvement.

I disassembled the chuck, cleaned it thoroughly making sure every trace of metal parts or grinding dust was removed.

On the left a shot from all three jaws. The grinding surface on the most left jaw is clearly narrower than the other three. It is likely that it was protruding just a tad less into the chuck, thereby causing the excessive runout. Also not that I rubbed the ground surfaces very lightly with a sheet of 1200 grit of emery paper with a few drops of oil on a flat surface. Doing so I took the sharpness off the edges of the ground surfaces.

I assembled the chuck and put it back on the spindle. To my dismay the runout had increased to 0.10 mm!

After some thinking I realised that the Unimats spindle has a double thread to hold the chuck. So I unwound the chuck, going very slowly at the end of the thread until I felt a click when the chuck jumped in the other thread. I then retightened the chuck and measured again. And now watch:

And now for the final measurement