Setting lathe compound to 90 degrees

[SteveM]

I managed to pick up the matching P&W 3C lathe to match my mill.

Didn't really have a pressing need for it, but figured it would be handy.

The compound on this thing has a LOT of travel, so it will be useful for turning tapers longer than I can turn on the atlas.

It has such a long compound because it has no traveling carriage, as do most jeweler and watchmaker lathes, so the compound is what is used for plain turning.

What are the "best practices" on how to set this to 90 degrees for turning parallel to the centerline?

Would I just indicate off something like a drill blank held in a collet? Is there a more accurate method?

An interesting thing about this lathe is that the base is cast integral with the feet at each end and the bed. There are only three mounting holes, two inside the headstock, and one next to the tailstock foot (you can see it in the picture). On the underside, the bolt holes go through pads so that the bed is actually suspended slightly above the bench and the lathe rests only on the three points, so there is no way that the bolts will twist the bed.

Steve

[whateg0]

90 degrees to the bed? Attach a DTI to the compound and run it across the face of the chuck or faceplate. Parallel to the bed, do the same but indicate the inside of the bed ways.

Dave

[pete]

It depends on how often you move that compound, but always having to indicate it back in to being parallel to the ways would get old pretty fast. If it were me I think I'd try and figure out a simple machined plate that auto indexes the compound back into being parallel using datum surfaces that are already on the cross slide and the compound if that's possible. Since I've never tried it then it's hard to say if something like that would give you enough repeatable accuracy to make it worthwhile. It would get you very close at least so it would still speed up the indicating if that was needed.

[spro]

Hi Steve. Looking at your lathe's tailstock, it appears to have a long barrel. The casting is also configured so that the long compound can really get close to it. So it seems a gentle taper is to be turned towards the headstock with the compound right closer to 90 than other lathes. 90 +/-degrees to the cross slide not the lathe ways.

[whateg0]

It depends on how often you move that compound, but always having to indicate it back in to being parallel to the ways would get old pretty fast. If it were me I think I'd try and figure out a simple machined plate that auto indexes the compound back into being parallel using datum surfaces that are already on the cross slide and the compound if that's possible. Since I've never tried it then it's hard to say if something like that would give you enough repeatable accuracy to make it worthwhile. It would get you very close at least so it would still speed up the indicating if that was needed.

I would think for a lot of things that might be ok, but I don't know that I would trust it to get the compound truly square. Then again, it's probably as good as sticking a 1-2-3 block on the chuck and butting a qctp up against it. I guess my thinking is that since this is done in place of a traveling carriage, it needs to be as close as possible, or else you'd end up with a taper on every part.

Dave

[SteveM]

I guess my thinking is that since this is done in place of a traveling carriage, it needs to be as close as possible, or else you'd end up with a taper on every part.

Yes, that's exactly it.

These lathes were probably not doing a lot of long turning across an entire part. This is the same setup as most watchmakers lathes, and what's the longest critical part on a watch?

I won't have it set up until I finish a complete reconfiguration in my shop, so reporting an actual result is going to be a long way off.

The fixture idea may work - I'd have to make one and see how it comes out.

At the very least, it will probably get me to within a couple of thou and then we indicate from there.

Steve

[earlgo]

Here is a thought from left field. Perhaps you could make a fixture that mounted on the bed ways, but slid easily. A DTI or equal could be mounted on the fixture so that it ran along the edge of the cross-feed dovetail. You could then adjust the cross-feed and then slide the DTI until you were satisfied. The sliding gizmo would prevent a lot of cranking during setup.
Perhaps this suggestion will trigger a better idea.
--earlgo

[John Hasler]

I put a radial pin in the edge of the disk and screw a small slotted block to the carriage in the correct spot for the pin to stop on. I'd then bring the pin into contact with the block, fine tune the angle using a dti and a precision bar between centers, lock down the block, check again with the dti, and be done. Another pin and block would set you up for facing. Make the easily pins removable if you might need to go past them.

[GlennW]

Chuck up a piece of aluminum bar stock, make a cut, and then mic both ends. Adjust and repeat until the measurements are equal.

That's the method i use on my cylindrical grinder.

[Harold_V]

What Glenn said. That's how one sets taper with the little Hardinge lathes, which changes every time the carriage is moved. It's the only truly reliable way to eliminate taper with those machines. Other methods (like dialing in a long shaft) will get you close, but that method doesn't address minor variations in things like how the indicator relates to the centerline.

I was never a fan of those machines.

H