[ajm291]

I recently got an X2 from Harbor Freight which I' immediately tore down and cleaned up. I began putting the table back together and noticed that I was getting binding when I brought the Y axis to the front of the saddle. I tweaked things for hours, but couldn't get the satisfaction out of the fit so I began inspecting things.

I took the table apart and ran a dial across the ways of the saddle and noticed that there is a .003 increase from the back of the saddle ( this is where I zeroed the dial) to the front of the saddle (where the dial read .003) This is consistent on both ways. They are higher at the front of the saddle and lower in the back. I'm using a granite counter top as my reference which I measured with the same dial to verify that it is flat (and it is)

I also noticed that the ways are lower on the outside of the saddle and higher towards the dovetail. Is this intentional?

Is there anything that I can do to get the ways flat to .001 or better? Like more aggressive sanding using the method described in crevice reamers way smoothing guide?

[TroyO]

What was the point of reference that you read a difference of .003 from?

For adjusting the fit/finish of the basic bearing surfaces the sanding and/or "lapping" methods only really work for very small (.001 or less) problems and not much more.

There are tons of messages on this forum and others about similar issues. (One of which was mine, I will see if I can dig it up, LOL.)

For now, take a look at this... and it's references linked at the bottom of the article.

Hand scraper - Wikipedia, the free encyclopedia

That may be what you will have to do to fix it. I had good luck using signifigantly less precise methods than what those links show, but I still stuck to the basic idea. The tooling can be expensive and fancy but you can get by using an old worn out file and a magic marker. Again, compared to the artistry of "real scraping" mine was at best a third graders tukey-from-tracing-a-hand but it still ended up way better than the way the machine arrived, LOL.

Here you go.. found the link to my post about scraping using a file... hey, it worked! So, look around for hand scraping and see what info you can find, but DO be darn sure when you go that aggressive at it that you are 100% certain you are removing the RIGHT metal, LOL. It's hard to put it back... ;-)

http://www.cnczone.com/forums/763376-post15.html

[Johann ohnesorg]

Call horrorfreight, they should send you a new machine base. To scrape 0,006" from the base will drive you nuts. It would take for ages.

[TroyO]

Yeah, that is why I was asking what the reference was... if he was referencing say from the granite surface up to the ways, in other words he was out of parallel from the "feet" to the ways it doesn't really matter.

Actually... I'd think for that one you would want to check the dovetails by putting a rod on either side and checking parallel.

Like this ----> o\ /o

And measure the front and back of them to see how far they were out from each other.

The next measure might be... flip the base upside down and rest the ways on a set of matched paralels. It should be easy to see if they are sitting flat on both paralells. If it is, that first dovetail paralelism check is where the problem is.

Hmmm, I will knock up a quick CoC because I don't like my explaination... LOL

[TroyO]

OK, Crap-O-Cad attached...

[ajm291]

Thanks for the quick replies and links. This is all new to me so my approach and terminology are most likely incorrect.

Let me explain how I got the .003 value. I have a dial mounted to a dial holder with a metal base. I put dial holder base on the granite and the dial tip on the way and zeroed the dial. The shaft of the dial was pressed up against the top of the dovetail. I moved the dial holder from the back of the saddle to the front keeping the shaft against the top of the dovetail while I moved the holder. There was a positive increase on the dial one third of the way from the back of the saddle and it continued to increase until I got to the front of the way where the difference was .003.

To me this made sense since this is where I was seeing the binding problem. I was assuming that since the way was higher in the front, the space for the gib was reduced which created the binding issue.

I'll try the approach with the rods and parallels next...

[SCzEngrgGroup]

Thanks for the quick replies and links. This is all new to me so my approach and terminology are most likely incorrect.

Let me explain how I got the .003 value. I have a dial mounted to a dial holder with a metal base. I put dial holder base on the granite and the dial tip on the way and zeroed the dial. The shaft of the dial was pressed up against the top of the dovetail. I moved the dial holder from the back of the saddle to the front keeping the shaft against the top of the dovetail while I moved the holder. There was a positive increase on the dial one third of the way from the back of the saddle and it continued to increase until I got to the front of the way where the difference was .003.

To me this made sense since this is where I was seeing the binding problem. I was assuming that since the way was higher in the front, the space for the gib was reduced which created the binding issue.

I'll try the approach with the rods and parallels next...

What you need to do is shim the base to get the horizontal part of the ways to read parallel to your surface plate in both X and Y, as there's no reason to believe the bottom of the base is machined parallel to the ways, nor does it matter in use. Once you have the ways "levelled", then measure the dovetails at several heights, and see if they are level and true. Finally, use the rods to ensure the dovetails are parallel to each other.

Regards,
Ray L.

[ajm291]

I understand everything in the previous post with the exception of the quoted text below.

...measure the dovetails at several heights...

I'm having a difficult time understanding what this means. Can someone help me understand this better??

Once the ways are leveled in both X and Y, does this mean to just measure a bunch of points on the dovetails to ensure that they are identical? If they aren't, then I should know what the problems are correct?

Thanks!

[ajm291]

I played around with the table this evening. I put the parallels on it, but they indicated that the ways were in good shape. I put the table back together and things fit a lot better this time around (I'm sure that I did something wrong to cause the binding in the first place) The Y axis is as smooth as butter moving from back to front now. It's like night and day compared to the way that it was moving/binding a few days ago. Thanks for the help!!

[luv2ride]

Man I hope that will be my case when I tear mine down and re-assemble! Glad it is working for you.

[TroyO]

Ohhh crap, I may have missed something entirely....

By chance when you were looking at the binding issue, were you using the hand wheel to move the table or were you sliding it manually?

The reason I ask is a common "binding" problem is because the leadscrew mount isn't centered with the hole in the leadscrew nut. Loosen the two bolts that hold the the mount to the mill base so they are just slightly snug-ish (Term of art.. ;-) and run the saddle in as close to handwheel as it goes and THEN tighten the two bolts back down.

That way the mounts can shift a little to let the mount get perfectly in line with the leadscrew nut before you tighten them down. If it's off centered it will get tighter as it reaches the end of the travel because the leadscrew is driving at an angle.

I hope I described that well enough... it may explain why it seemed to work better after re-assembly. You may have just lucked in to better leadscrew alignment the second time round?

[luv2ride]

That is a good point. I know I had to tweak mine a little bit on re-assembly a couple of weeks ago.