Alignment Question for Machinists

[Lavcat]

I have never used a cross table/x-y table before. This afternoon I mounted a cross table (Proxxon KT 150) to the base of my drill (Proxxon TBH). The table mounts with four bolts and nuts to the T-slots of the base.

I figure it is important that the direction of motion of table be square with respect to the base, but I'm not sure to what tolerance? I've not seen anything in books or online that speaks to this. What I believe is most important is that the direction of motion be square with respect to the work that is on the table.

But to check my table alignment I mounted a dial gauge on the column and indicated across the left to right range of motion. The surface I indicated does not look like a precisely ground surface, but it is the best that I saw on the back edge of the table. I was right on at both extremes, but I had a deviation of -3, +5 mils running across the travel.

I suspect that this alignment is quite good enough, and as good as I can get. But I'd love to have confimation from someone who knows about such things.


Edit: Bonus question -- what is a good oil to use for lubricating the table? 3-in-One oil?

[tamu]

I would say 8mil of difference is way too much. on the other hand it really depends on the job your doing too. if its just for rough jobs then it wont matter much but I would make sure its as square as possible. If it was me I would get it with in a couple of thous.

maybe you can jack the low end up with some shims and try to get it square?

we use good old 20-50 oil.

[Lavcat]

My description was probably not very good. I meant square in the horizontal plane. I have not yet measured vertically. There is lots of literature about getting tables square vertically so I did not mean to ask about it.

Also since I am within a fraction of a mil on either end I did not worry too much about being more off in the middle. I attibute the error to the poor surface on which I am indicating.

[Lavcat]

Measuring vertically the table is square to within 5 tenths, which is one graduation of my gauge.

[Conrad Hoffman]

The rotational alignment of the table to the drill press doesn't matter at all. It's only when you put a vise on the table that you need to orient the vise correctly. I wouldn't want to see more than about 0.0005" over a smooth fixed jaw, and that's for a 6" jaw. Still, it all depends on what you expect to do.

[Lavcat]

Squaring the vise is next! Will report back hopefully in a couple days.

[tamu]

Originally Posted by Conrad Hoffman

The rotational alignment of the table to the drill press doesn't matter at all. It's only when you put a vise on the table that you need to orient the vise correctly. I wouldn't want to see more than about 0.0005" over a smooth fixed jaw, and that's for a 6" jaw. Still, it all depends on what you expect to do.

true but if you want to clamp the work directly on to the table then you need the table to be true so say you can drill 5 holes on a straight line. but I guess you can do that on a vice too.

if you still want to align it straight you could use a bit of rod in the chuck and poke it through the t slot. take measurements in several different spots and adjust until you get the same reading.

[Conrad Hoffman]

I'm confused! The T-slots are always parallel to the table travel. If you bolt something directly to the table, you still have to indicate it in, just as with mounting a vise. The attachment of the whole X-Y assembly to the drill press is still of no concern.

[tamu]

yeah sorry Im making it hard to understand. I was talking about clamping things directly onto the table.

What I do is to get a couple of flat bars that fit perfectly in to the slots, while having some of it sticking out and push the work right against them then clamp it at that position. it gives me a near perfect allignment against the table. saves lots of time too.

[bear]

Trying to get the outer dimensions of a SQUARE shape using an X-Y table is a problem.

The reason is that the screws (acme threaded lead screws) have slack WRT the motion of the table. It is NOT repeatable.

People take these sorts of machines and do all sorts of mods to get better tolerances, including putting ball bearings on the crank end of the screws... There are tricks to getting less "lash" wrt the "nut" (usually bronze) that rides the screw...

Or you can mount some of the modified digital dial calipers on the side of the slides for a poor man's DRO.

The general trick is to take out the lash by carefully rotating the crank end till you feel the slack taken up, and then using the turns counter on the crank to get to the spot you want...

I think this is what you are asking about...?

With a low cost X-Y table you will need to mark ur spots and just use the table as a positioner... also if it doesn't have LOCKS for the ways, you'll need to modify it so that it does, otherwise you can't really get from point A to point B without the possibility of the table surface moving...

Btw, and fwiw, you can buy a full milling machine in the USA, used, for <$2000, and sometimes WAY less than that... occasionally there are small ones too... the Chinese benchtop mini-mills may suffer from the same issues to some extent.

There are reciprocating ball screws that can be bought and subbed in, for a very precision result. In the small size they are not very expensive... fyi.

That's what the home DIY CNC'ers use...

_-_-bear