Set screws: cup to dog point

[historicalarms]

Your idea on chucking up in a hand drill & touching to a grinder or belt sander has worked for me on many occasions, on bit of advise tho is to thread a nut all the way onto the screw before touching the grinder. quit often the grinding operation will fold a thread or two back where the grind stops and with a nut on your screw you just unscrew it to re-form the beginning thread on the screw.

 

[DPittman]

Lots of good advice already but I had a bunch of old cup screws so I chucked one up in the drill and took it to the belt sander. I wasn't particularly careful and a better job could be made but it was simple and I think effective enough. Maybe not good enough for some applications. The first picture is the one I ground next to a factory made one.

 

[trlvn]

Lots of good advice already but I had a bunch of old cup screws so I chucked one up in the drill and took it to the belt sander. I wasn't particularly careful and a better job could be made but it was simple and I think effective enough. Maybe not good enough for some applications. The firstView attachment 5210View attachment 5211 picture is the one I ground next to a factory made one.

Thanks for the demo. I was hoping to try this today but life got in the way. Maybe tomorrow...but maybe not.

Craig

 

[PeterT]

I'm not too familiar with your lathe but I will point out another consideration. Maybe it pertains to you, maybe not.

There are screws to adjust the annular gap between gib strip & dovetail or slide tightness if you prefer. They may be spaced out in an array if they are the parallel, non prismatic type gib strip (which are adjusted on the ends). And then there are 'carriage lock' screws which actually an abomination of the same gib tightening screw thing. They push on the gib strip which in turns imparts friction on the dovetail which then <cough> locks the slide. Its quite common, at least in offshore lathes. The only problem with it is that the act of tightening can also impart movement to the slide as it imparts torque to the gib strip. So you set your dial to 1.000, 'lock' the slide, it now moves to 1.005" & you would not be wiser as you make your cut. Hard to do accurate parts with this potential variation. I learned this watching my DRO wander as I engaged the lock. I also noticed little bite marks on the gib strip itself.

So I went out of my way to make a frictionless lock system. I used bearing balls on the ends of the screws attached with Loctite. These engage in a 60-deg center drilled cup on the end of a small brass foot. The other side of the foot foot has the same angle as dovetail. The net result is torquing the screw imparts 99% perpendicular force to the gib strip nad little or no displacement to move the table or slide. I can probably scrounge up a picture if this doesn't make sense.

 

[trlvn]

OK, I'm on the right track now. I made a simple version of the holder @John Conroy showed. Just used some 5/16" aluminum tapped 8-32 and then slit from each end with the hacksaw. If I tighten the 3-jaw as hard as I can, the screws usually stayed put.

Just need to repeat the same thing with the 10-32 screws.

@PeterT my little Atlas 618 does not have a proper lock on either the cross-slide or the compound. These set screws are the only way to 'lock' that direction of movement. In fact, only 2 of the 4 screws are actually bearing on the male part of the dovetail at any one time. In the following (old) picture, you can see the gib screws along the right side (one is out of sight). The ones at the top and bottom are just pushing on air!

After I get the other gib screws replaced, I'm going to fire up my diy milling attachment and see if I can do any better. Which shouldn't be hard--the first attempts were...ragged.

Craig