X2 Mill Table and Saddle with linear rails

shooter123456

Registered
Registered
Joined
Jan 20, 2016
Messages
602
After a few months of fiddling with CNC on my X2 mill with mixed results, I decided to try my hand at making a new X and Y axis to get rid of the dovetails. Its the same old story where the dovetails are rough and not too accurate so to get all 8 inches without binding, the dovetails are so loose there is chatter in the middle. In recent months I have found a somewhat happy medium where I can machine light but fast cuts in aluminum (.375" deep, .025 wide, .375 endmill, 60 IPM cutting) while rapiding at 100 IPM. I can hear the motors groaning while in the last 15% of each end of travels and I had to drop to half stepping to make sure they have enough torque to maintain that.

First, the plans and goals of the project.
Goals:
-Increase X and Y travels to at least 14 and 6 inches respectively
-Reduce binding while maintaining rigidity
-Maintain or improve accuracy
-Rapids at 350 IPM

Plan:
-Use aluminum stock to make new table, saddle, and base
-Mount linear rails for accurate and friction free motion
-Make new ballscrew and motor mounting hardware
-Bolt to existing mill base and add extension to maintain rigidity

I am sure the time and effort would be better spent fixing this mill a bit, selling it, and getting something better, but thats not as fun. I figure the worst case scenario, it all goes to crap, the accuracy and rigidity of the new table is garbage, and I go back to where I started out a few weeks of machine work and whatever stock I used for it. Since all the components can be removed and put on something else, there is almost no risk.

I started with a model in Fusion360 and I got it as dimensionally accurate as I could. I found a bunch of 3/4" aluminum plates at a recycling center near me that I am 85% sure is Mic 6 or similar (it machines like cast aluminum, nothing like the 6061 I normally use) and ground. The finish on it was almost mirror like with long swirls and the thickness was incredibly uniform throughout. Checked on a surface plate, most of the surface was within .003" flat and parallel. I used some ink and sandpaper to get it down pretty close on each part.

The components I will use are:

-2 22" 25mm Thomson Accuglide rails with 4 bearings (X axis)
-1 20" 20mm THK HSR20 rail with 2 bearings to be cut in half (Y axis)
-2 IKO MLFG24 19" Rails with 4 bearings (Z axis)

All were ebay deals that I got pretty cheap. The cost was $150, $80, and $68 respectively.

Here is the latest model I am working with. It has been updated as I go and things change.
ZU1upGX.png


Here is one of the hauls I made from the recycling center. You can see the big aluminum plate in the middle with that reflective surface. That part will become the table. I pay by the lb there and paid about $120 for all of the stuff in the picture. The piece on the left is 1.5" thick aluminum, 30" long and 5" tall at the smallest height.
vOkmGUw.jpg


The first part I made was the table. I cut it to a little over 22" long and 8.5" wide. I machined the ends square and I might machine the sides at some point, but fixturing 22" is tough on my little X2. I made a little drill guide to spot the holes for the rails. To do that, I drilled the first hole of the first rail and tapped it, lining up the rail to a scribe line as best I could. Then I secured the first end. I then clamped the other end down and spotted the last hole for the first rail. That hole was then drilled and tapped. Now with both ends secure, I spotted the remaining holes and drilled and tapped. I then loosened all the screws but the first one, clamped the rail against my surface plate and tightened the first 4 screws. I then moved the surface plate down, clamped it so half was against the already secured portion of the rail and tightened 2 more screws. I did this until I got to the end.

Next I got working on the saddle so I could use it to secure the second rail parallel to the first. I used the mill to drill and countersink 16 holes for M8 screws. Once I finish that, I secured the saddle to the bearings while mounted on the first rail. I finagled it a bunch until I thought the second rail was pretty close and clamped it down. I did a similar procedure where I drilled and tapped the first hole, secured the rail, moved the saddle to the other end, drilled and tapped the last hole, secured the rail there, spotted the remaining holes, drilled and tapped, then moved the saddle along, tightening the screws as the saddle was right next to them. Once I finished this, the saddle moved back and forth the full length with little resistance. Checked with an indicator, the rails were within .002" parallel over 22" and within .001" height over 22". Thats good enough for me.

I followed a similar procedure doing the Y axis. The saddle was drilled and countersunk for the M5 screws the HSR20 bearings needed. The rail mount/base was drilled on the mill, then tapped using a screw I cut flutes into because I couldn't find an M5 tap anywhere nearby and I didn't want to wait for one to ship. That worked surprisingly well, but I had to power tap with a drill, otherwise I couldn't get it to start. The Y axis moves with so little friction that lifting the base enough to put a pen under one side causes the whole assembly to slide all the way down. I am still not sure if the 2 bearings on the Y will be sufficient, but if they aren't, I will get new rails with 4 bearings and remachine the saddle and mount. Though for now, it feels plenty rigid. It also weighs at least 25lbs more than the original table.

Here is a picture of the saddle sitting on the Y axis with the X axis bearings installed. The caged balls helped a ton with assembly since the table could be slipped off to get to the screws underneath.
FDWXYnb.jpg


Here it is with the table installed.
b50wBPG.jpg


A view of the goodies underneath.
ntw26sN.jpg


Here is the left end of the X travel.
WFTajNZ.jpg


And the right end.
BpZNFg2.jpg


In those pictures, it looks like the table is dipping a little bit, but it isn't. Its a trick of the light and the slight bending of the desk the assembly is sitting on.

With that part squared away, I got started on the hardware mounts for the X axis. I need two bearing blocks for the ball screw, a motor mount, and a ball nut mount. With the table ready, I took some measurements and got started.

Here is the stock I started with for the 2 blocks.
HCZdcVo.jpg


Rough cut with the stock that will be used for the motor mount. I like these small components because I can use the cut offs from other projects. The X motor mount will be made with the leftover from cutting the ends off the table plate.
7HpUhhH.jpg


I got the first block mostly done. I squared it up as best I could, then cut the 2 important surfaces (top and bearing face) in the same set up to make sure they were as square as my mill is capable of. The machine cut the bearing seat a little small the first time, but I had the toolpath cut with -.002" stock to leave and it cut it for a nice snug fit.
l6o8hs5.jpg


Thats where the project is now. I am looking at 15" of X travel without overtraveling the bearings, and I can probably get away with 2" on each side for getting around stuff, but not cutting. The Y axis has a little over 7" of travel and I had to buy a new leadscrew, since the old one needed to be cut down to fit and would only allow for 5" of travel.

Next, I need to do the other bearing block for the X, the motor mount, and the ball nut mount, then the same deal with the Y. Then I will drill and tap the X2 base to take the new table and if all goes well, the Z will be after that.
 
I am back with an update. Everytime you start cooking with Crisco, something gets thrown in that gums up the works. In this case it was my bandsaw going haywire and not cutting, then after a little fiddling the blade snapped. Cost me a few hours and now I need to wait for a new blade. So it is back to the ol hacksaw while the new one ships.

So I left off needing to finish the Y axis. One rail mounted, the other needing work, as well as holes drilled and countersunk for the mounting bolts on the base. After drilling and milling the holes for the mounting bolts, I realized there was an issue with the Y axis base plate. The edge I squared up was nowhere near square to the adjacent edges, and thats the edge the first rail holes were drilled in reference to. Additionally, since I cut the rail in half, the holes were not spaced the same distance from the end. First some pictures. I like seeing in progress pictures on others peoples pages so I can get an idea how they fixtured and machined every step, so I took a few in progress pictures for this thread. I hope it isn't excessive.

Here is the base of the Y axis after squaring against the left edge with a .0005" DTI and spot drilling the holes.
t2xC4Sw.jpg


All the holes drilled and half countersunk with a 1/4" end mill. This is the first time I managed to get my mill to drill since CNCing it. It needs to run with higher RPMs and rediculously slow feeds like 1 IPM. Otherwise the spindle motor starts to stall and the Z axis looses steps. No worries though, I have nothing else to do so 5 extra minutes to do a drill cycle isn't the end of the world.
G9t3uFB.jpg


All done with that part!
piSKB6q.jpg


Now I needed to figure out how to make sure the second rail would be parallel to the first and the holes are drilled in the right location. I came up with an idea. First I tightened down the first rail against the surface plate to make sure it is square. Then I secured my indicator holder to the bearing with a broken drill bit. Using the model and the theoretical proper distance, a bit of layout dye, and some luck, I scribed a line where the centerline of the second rail should go. Like so:
Xb2DJhb.jpg


Then measuring the rail and the hole location and referencing the squared edge, I scribed lines into the mount marking the locations of the holes. (Ignore the one between the first two... Minor hiccup) But then I needed a way to get the holes to go there. I don't have a center punch or a real spot drill. But I do I have a pretty well equipped machine shop so I fixed that. I made a little center punch and a D bit cutter to start the holes in the right spot.
vUavppE.jpg


Then spot drilled. It went remarkably well. I don't think I have ever drilled such accurately located holes without the CNC. They look like they are slightly off, but they are darn near perfect.
xgXN19V.jpg


So then I drilled out the holes the rest of the way to be tapped. All of this drilling was done on the machine that started it all. My little $60 Harbor Freight drill press that is now 3ish years old and has seen better days. Notice the dent and the broken switch from when there was a cord that got wrapped around it which ended up pulling it off the bench. Whoops :)
Y4EB6cl.jpg


Once drilled, the holes needed to be tapped. As I mentioned in the previous post, I didn't have a tap and didn't really want to wait to buy one. So I ground flutes into a bolt and used that. I did my best to get the picture to show up well. Here is that "Tap".
kovPJgA.jpg


Next up I needed to make another bearing block. That was relatively uneventful. Just some squaring, then milling the circle in the middle, the drilling and countersinking holes for screws. Here are the 2 bearing blocks finished.
a11xwPv.jpg


8meUoUA.jpg


It took me a little while, but I messed with the backlash compensation to get the holes as circular as I could. The bearing seats are within .001" to the best of my measuring abilities (which admittedly aren't great...)

The next thing I made was the X axis motor mount. Initially I designed it to be made out of .5" stock, but I had more scrap from .75" so I went ahead and adapted it to use that. Better than cutting off new stock unnecessarily right?

I started by getting it squared up, and to get the height perfect, after milling that edge to length, I zeroed the machine to that edge. Once squared, I cut the circle, spot drilled, through drilled, then countersunk the holes with a .125" end mill.
nFZchNU.jpg


While countersinking and opening 2 of the holes, the clearance couldn't have been more than .002". I thought it was going to crash but I trusted my measuring and let my hand hover over the E stop and let it do its thing. I would have extended it, but the tiny end mill was already way too far out. If its not a crash, its clearance! And I hope to never have so little clearance again. It didn't touch and there was absolutely no indication of scraping on the surface.
YCsQyFy.jpg


Here is the finished part. I used the flash to make it show up a little better. You will noticed there is a little extra hole in there... I touched off the 5/32 drill after spot drilling, but that height was a little bit higher than the home location. So when I started the mill and ran the program, the first thing it did was return to home, which managed to get it .05" into the part before I managed to slap the E-Stop. It will be covered by the motor and won't cause any issues, so I just continued. No one will know but me... (and all of you guys, but please don't tell.)
TqJ6UBn.jpg


After that, the last thing I needed to get the X axis all squared away was the ballscrew mount. I had a 1.5" cut off from a cut off I got at the recycling center. I dunno if many environmentalists read these forums, but a cut off from a cut off is some pretty good recycling. I marked it, put it on the bandsaw, and thats when sh** hit the fan. The bang was loud, the saw was bucking all over the place, oh man it was rough. Hacksawing it off took probably 15 minutes and it was 91 degrees and 70% humidity in an unairconditioned garage. I miss my saw already.
Pl9fp5G.jpg


Machining it was uneventful. Cut the middle hole, spot drilled and drill the surrounding holes. To make sure the face was perfectly square to the edge that would snug against the saddle, instead of using parallels, I indicated that edge with the Z axis before snugging it down. Here is the X axis hardware. 2 bearing blocks, a motor mount, and a ball screw mount. Some look off center, but that is on purpose. All of those dimensions and clearances were carefully measured and modeled to ensure alignment and clearance. I need to figure out how to get the Y axis ball screw to fit. It will take some more thought before I start on that.
Be0sFUv.jpg


Next I need 2 Y axis bearing blocks, a ball nut block, and a motor mount. That will get the whole thing moving and it can replace the current table to let me make the Z axis. The thought of 15" of x travel with zero binding and 7" of Y travel is what keeps me moving. I plan to also add some limit switches so I can home automatically instead of manually. I am also brainstorming ideas to beef up the column. Currently, I am planning to epoxy fill the column and bolt steel plates to the sides to prevent twisting and dampen vibrations a little. Since it will be using a 1.3 (to 2.25 for 15 mins) motor, it needs to be bulked up a little to use it effectively.

Luckily, I didn't have to do all this work alone. I had some company. I usually make him stay inside since its so hot in NC this time of year and he is black with a big coat, but every time I went inside, he was waiting by the door, and my mom told me when he hears me outside working, he doesn't move from the door and he waits there for hours for me. No worries though, I make him go inside intermittently to get water and cool off a bit.
e1OyFjQ.jpg
 
That's quite the project! One question and two comments ...

Question - Maybe I missed your mentioning it, but do you plan to extend the distance between the spindle axis and the column (ie, move the head outward)? I can't imagine that increasing Y travel would do much good without this.

Comment/concern - The empty light socket on your drill press, turned upward and catching metal swarf, really gives me the willies! Yes, of course, you have it turned off. But switches can get bumped, and metal swarf in the socket could really give you a flashy surprise. I'd suggest putting a bulb, even a dead one, in the socket. Or else completely removing the lamp assembly. Be safe!

Comment/concern - It appears that there are just two linear bearings between the base and the saddle, one at the center of each side. With just one bearing per side, I'd be concerned about applied torque forcing the saddle (and table) to rotate, throwing accuracy off. Have you calculated the forces/stiffness?
 
Last edited:
That's quite the project! One question and one comment ...

Question - Maybe I missed your mentioning it, but do you plan to extend the distance between the spindle axis and the column (ie, move the head outward)? I can't imagine that increasing Y travel would do much good without this.

Comment/concern - The empty light socket on your drill press, turned upward and catching metal swarf, really gives me the willies! Yes, of course, you have it turned off. But switches can get bumped, and metal swarf in the socket could really give you a surprise. I'd suggest putting a bulb, even a dead one, in the socket. Or else completely removing the lamp assembly. Be safe!
Its tough to see in the model, but with the Z axis rails, there is a new head mount that extends the Z axis a few inches out while reinforcing it. I can't remember how much I planned for, but I think its about 2.5 inches.

Regarding the drill press, I hadn't considered that, but my dad pointed that out to me a while ago, and since it was pretty much useless when it worked, I popped that little cover off and clipped and capped the wires. I would have removed it completely, but there is a lot more in there that the cover protects. So there is no risk anymore, and no extra light from it. I actually hadn't noticed that light socket for over a year until you mentioned it.
 
Sounds like you've got everything covered (at least the stuff I'd originally mentioned). Just did an edit and added another concern.

PS - I have an X2 myself, and every time I dream of increasing Y travel, I realize it's not that simple without extending the head.

PS - thise do look like very stout linear rails and bearings.
 
Sounds like you've got everything covered (at least the stuff I'd originally mentioned). Just did an edit and added another concern.

PS - I have an X2 myself, and every time I dream of increasing Y travel, I realize it's not that simple without extending the head.

PS - thise do look like very stout linear rails and bearings.
The Y travel has been my biggest complaint about the mill since I got it so I am hoping this solves the problem.

With my original model, I did a simulation on mill applying a 50 lb moment to the table, and there was very little deflection. At the worst position, it was less than .0005". They are THK HSR20 bearings and the data sheet says rotation about an axis projected from the spindle is rated for 820 Nm which should be more than sufficient for the goal of meeting or exceeding the rigidity of the current system. The X axis is rated way higher (like 28,000lb per bearing) than I would ever ever use, but thats what was available on ebay, and they make the Y axis look puny and insufficient. But in theory (at least as far as I understand the theory), they should be ok. I won't know until I finish it and try a few test cuts, but if it isn't strong enough, I will find another set online, and re do that part.
 
Neat! It does appear that you've done your homework with regard to rigidity. Looking at the third image in your original post, I also understand your constraint with regard to how many Y axis bearings you can use. Please do keep on posting as you go along. This is a very fascinating build!
 
I have done a bit more work and have a few more pictures. I got started on the Y axis hardware. I ran into a design issue where I couldn't figure out how to secure the Y axis ballscrew to the saddle. I ended up machining a mount for it to attach at the far end, just outside the saddle, near where the column of the mill attaches to the base. I also machined clearance for the ballscrew into the Y axis rail mount. It was a little bit off center and skewed, but since it is just clearance and there is more than enough space now, I decided not to bother fixing it.

Here is the cut off I started with. I had to cut it using a miter saw with an old carbide tipped blade on it. Its far from ideal and made a mess, but it worked so no worries in the end. This chunk was found at the scrap yard and its proof that yes aluminum will rust. No idea what alloy it is. I love seeing beat up and rusted out chunks of stock being cleaned up and turned into something shiny.
ROE47DQ.jpg


Here is the finished mount, about where it will be installed. It will be lifted up a little and bolted to the saddle between the X axis bearings.
nfZEY6N.jpg


Here is the clearance cut in the Y axis rail mount.
GzsOx8A.jpg


It looks like there won't be any room for a second bearing block on the Y axis, but it would have been a floating block anyway so I don't think I am losing much there. I also figured out that I need to extend the Y axis motor out a few inches because at the current ballscrew centerline, it won't clear the saddle.

At this point, I need to machine the Y axis motor mount and bearing block, then install and align everything, then it will be ready to go onto the mill. To get the X and Y square, I plan to take a 1-2-3 block and clamp it to the table, then stick an indicator on the mill saddle and indicate on one of the sides, then clamp the base down. Then I will put the table on the saddle, put the indicator on the the table, and indicate on the perpendicular side of the block. I left plenty of room for adjusting them in. Once the table travel is perpendicular, I can secure the bearings. I hope that made sense.

I stuck the whole thing on top of the mills current table to get an idea of the size difference. The new table assembly is significantly heavier than the old one. I haven't weighed it, but the model estimates it should weight about 45 lbs. The new table is about 2 inches taller than the old one, but the Z axis rails should more than make up for that. I plan to add epoxy granite to fill the base, the space between the very bottom plate and the new Y axis base, and the inside of the column to add some weight and help keep vibrations to a minimum.
stU9u8f.jpg
 
I have a bit of an update, though not super exciting since I have been out of town and haven't been able to machine anything. I modeled up the pneumatic tool changer and added it to the working model. I got the top plate cut for that, but thats all the machine work that I have pictures of. Today after work, I want to get the Y axis motor mount and bearing block made and hopefully mount some motors and see how it moves.

I decided to try making some lock nuts for the ball screws since I have never been happy with the ones provided and I can't find M12x1 nuts anywhere. I saw a design for some that use 2 pieces and screws to lock, so I am going to try to make some of those. Here is the design I am going to try. As far as I could tell, you thread them on to the ball screw, then tighten the first one down, then use the screws to lock it in place. Since it is a 1mm thread pitch, the max they will be apart is .25mm.
RioqASt.png


Here is the mill with the pneumatic tool changer added.
upykysf.png


T0 keep myself busy during a long plane ride, I decided to start planning for the next step up for this project. Building an entire machine from scratch. I will start small so I can figure out what goes into a scratch built machine and what shortcomings I need to overcome, without spending a ton on material and parts that might be scratched. The mill will have 9" x travel, 5" Y travel, and 12" Z travel. It will use linear rails on all 3 axes, either the current motor from my other mill, or another treadmill motor, I plan to make the spindle myself, as well as a full sheet metal enclosure. The body will be a welded steel frame and filled with epoxy. Once I finish this project, this one is probably next.

Here is the model:
9haON23.png


Since its hard to see scale, here it is next to the X2 mill.
1kGjwV3.png
 
Back
Top