I finally got curious about the split nut and also wanted to try and adjust out any backlash in the X axis.
So in my reading I found a lot of info about adjusting the star nut thingy where the X axis handle attaches. Supposedly I was supposed to fidget around with this and it would help take up some slack.
I guess the idea was to adjust the preload on the screw or something? I'm not sure but I couldn't seem to do anything with it that provided any kind of result. So...I guess my screw was preloaded correctly.
Moving on to the split nut. It's just what the name implies and I finally figured out why there are two holes just below the locking handle.
I loosened the two screws and turned the handle till it ejected the split nut. This is what mine looked like immediately after it cleared the clamp. That thing in the middle is supposed to be a spring right?
Ahhhh, it IS a spring! In this shot the spring is fully extended. It didn't seem that effective to have the spring totally smashed together.
I compressed the spring by an amount I thought would make it happy and then drove it back into the clamp by turning the handle.
Okay, so how the heck do you get the backlash out? For me it was mostly trial and error. I ran the nut in to the left and then put light pressure on the nut by turing the let clamping screw gently. The problem is, if you over tighten it both nuts get locked down.
Then, run the table back to the right, just a hair, presumably buy the amount of backlash, and then lock down both nuts. I got mine down to .0018 before I got bored with adjusting.
I decided to measure the backlash on the Y and Z as well. I still have the protective sheet over my DRO screen so I just used a silver sharpie and noted the backlash for each axis.
I found the positioning to be "pretty" acurate...at least within .001. However, this brings to light the fact that it's essential to always move the axis in the same direction as when you zeroed out, otherwise you will be off by the amount of backlash.
I also found that if I over traveled by the amount of backlash when moving in the "wrong" direction I would be right on the interned mark according to my dial indicator. It seemed like knowing the exact backlash could be useful information in some situations.