rotary table alignment

[picclock]

I am just about to do some fairly accurate  work using my 6" Vertex rotary table and I thought I would give it a checkout. I've never got very accurate results from it, always good enough though I've had my suspicions. I set up a dti on the inside edge of the central MT2 taper, the one I normally set the centre position with. The runout was 7 thou when rotating the table by hand   .  I thought I must be doing it wrong and put an MT2 /B16 chuck adapter in it to measure the outside of the adapter but the results were the same - ouch. I then checked the outside edge of the table which has a runout of 1.5 thou - better but hardly great. 

There is no noticeable play in the central bearing - probably <0.5 thou. So my question is how can I get an accurate table alignment ?

I'm OK with chopping up an MT2 adapter and fitting it permanently so that I can indicate off of the true centre mark I have made, of coarse I have to find the true centre of rotation first, and I have no idea how to do that short of moving a cylinder around to try to get it right. I have a dummy MT2 adapter with an end which can be machined. If I were to mill the edge of that the centre of the milled part should be the centre of rotation. Then I would have to try to key the adapter so it would be in the correct angular position .. . 

Unless there is a better way  .. .

Really appreciate all the help members of this forum give - without which my engineering (I use the term loosely) experience would be far less rewarding.

Many thanks for your time and patience

Best Regards

picclock

[hopefuldave]

7 thou" runout sounds a bit excessive to me too... is the table sat square on the bearing, or do you get a height variation as well?

The MT2 adaptor sounds sensible to me, if milled in situ it would be concentric, at least - then once you  have it in place, centre-pop both adaptor and table for future installations?

 If you wanted to correct the central MT2 socket, you could temporarily loctite the adaptor in place for milling and set-up and then heat the adaptor until the loctite lets go once indicated in (not incredibly hot - might be a good idea to put a central tapped hole in the adaptor so you can jack it out with a tube and bolt)

If you took the rotary table to the lathe face-plate, assuming it's big enough - otherwise dismantle it to the point where you can get it on the lathe? The table should unbolt from the wormwheel (I assume the Vertex is what Soba cloned, and the Soba's wheel does!) and could be bolted to a flanged stub mandrel to go in the 4-jaw if you don't fancy the idea of the complete rotary table whizzing round while you bore the central socket I expect your lathe could handle a 6" disc!

Once on the faceplate / in the 4-jaw and indicated on the newly-machined reference, you could then rebore the taper and follow it up with a MT2 reamer if you're lucky enough to have one - it would be hard to make the runout worse than it is!

I've found some of the far-eastern rotaries can be very accurate, sometimes the same manufacturer gets it wrong, though - I have a 4" tilting table where the tilt pin isn't parallel to *any* other surface, by a few degrees - eventually (tuits permitting) I'll put it on the lathe's boring table and - well - bore it straight! Apart from the tilted pivot (and the crude angle setting) it's pretty accurate, well under a thou" variation in height , bore centre as the table rotates etc. - I was thinking I'd put another mod on it to make it a "sine rotary table" on the tilting axis while I've got it apart? That would mean drilling and boring for another pin parallel with the pivot and an accurate distance away - relatively easy while the table's bolted to the lathe!


Just my ape knee's woof,

Dave H. (the other one)

[Bluechip]

Hi Picclock
 
Just a small point ... I have a Groz 2MT to B-something chuck arbor ( supposed to fit a Jacobs #34 drilling chuck )
 
And the Bxx end is as pi55ed as the proverbial newt ... some .011" eccentric. 
 
I would verify your arbor in something like a lathe mandrel ( I found out in the Myford mandrel ).
 
Other than that, other Groz bits I have seem fine. As Dave H says, some manufacturers have an off day ... 
 
Trust nowt until you know it's good ...
 
Check it in some other 2MT socket.
 
Now you've frightened me. I have a 6" Vertex ... never checked mine ..
 
Dave
 
 

[picclock]

It occurred to me after I posted that I could 'just' machine a recess, say 4mm deep, in the top of the MT2 socket in the table centre. This would leave it all mostly untouched, yet still give me my accurate centre for setup. I could use a carbide mill and only remove 10 thou or so for a depth of 4-5 mm, easily enough for centring. I don't use the MT2 socket for anything other than centring so nothing would be lost, and in the future should I ever need it, a strip down and re-bore of the MT2 taper could be arranged.

The table surface is remarkably parallel to the table in both planes - I checked when it arrived, and have done so since  -  it just never occurred to me the centre MT2 taper would be eccentric.

Many thanks for the advice.

Best Regards

picclock

[Pete.]

If you can tip the head of your mill over and use a DTI and some maths to get it the same as the taper-angle for MT2, you could mill out the 7 thou runout and still keep the taper. The adapter would sit a bit lower but you're only taking off .0035" so it won't be much.