Hi Raymond .
I own a similar machine to the lux (dovetail column i hope ?) and also had some issues but had a play around and found that the column had gunk under it and was not truely at 90 deg to the table .
Now you say that it trams within a thou @ 225mm circumfrence and about the same if you wind the head up the column – what are you using to test against ? an angle block ? Make sure it is 100% parrallel to the axis you are using as a datum .
When you raise and lower the head the locks must be loosened so the HEAD would follow the inside of the dovetail and not the flat base it is meant to bear against when locked – could be mis -leading .
So try clamping the locks at each end of the travel so the head is bearing against the flat parts of the column and confirm the readings .
Have you tried to tram the quill with the locks on the column dove tail as loose as when you checked the vertical movement of the head for trueness ?
You could try to tram the quill to column and see if they are parrallel but this is a mission and would need some tools made . also the spindle needs to be locked so it cannot rotate but can move freely up and down .
I'm sure they use a taper gibb set up on the column and , maybe there is an issue there ?
Have you tried placing a block of wood under the quill then using the downfeed apply a small amount of pressure before locking the column locks ? – see if there is any change .
I would also ensure my table locks are tight before doing any vertical measurement – just to be sure – after all that what you are ultimately using as a datum .
IAN