[109jb]

Decided to start a new thread to catalog the build of my round column HF mill. The mill is one that I bought in 1997 and is a model 981. It is about the size of a RF-25 or a Grizzly G1005. Just a bit smaller table. The table is 6.25" x 20.25". I really want an RF-45 size machine, but it isn't in the budget for a while, so even though this is a round column mill, I decided to go ahead and convert this one to get some experience in converting a machine to CNC. I figure I can use this experience when I can get a RF 45 type machine, and if I can get the RF45 machine before I have to sell this one I can use it to make the parts for the RF45 machine. The machine is in pretty good shape, but has seen some storage time during moves and has to live in my barn that I can't afford to keep heated all the time. Here is a picture of the machine before any conversion work was done.



The first thing to do was to get a plan of attack and a decision on what motors to use. For this size machine, I decided that steppers was the way to go. I actually started thinking about this about 6 years ago and back then I bought a stepper kit from Stepper World to play with. I also have a 9"x20" grizzly lathe that I figured I could use the 150 oz-in steppers on. I started a thread a few weeks ago that shows my motor decision. With the help of some fine people on this forum, I have settled on the Keling 570 oz-in steppers and will eventually power them with some good drivers and a 50V power supply. Due to money constraints though I will start with the 570-oz-in motors and TB6560 single axis drives and a separate breakout board and a 24V switching power supply. Here are a couple pictures of the BOB and drivers that I plan to use initially.





From the start I knew I wanted ballscrews and settled on the McMaster Carr rolled ball screws with spring loaded double ball nuts for zero backlash. So, this is the first task I tackled was making all of the ball screw mounts and designing the motor mounts for the steppers. The problem many have encountered when converting their machines is that you first have to take the machine apart, measure everything as accurately as possible, and re-assemble the machine so you can make the parts, and hope your measurements were close enough to have everything work right. I did all that, made the parts and now have a machine that has ballscrews installed.

This first picture shows the y-axis ball screw assembly. You can see the bearing mount at the left , and the double ball nut and the mount that connects to the table saddle. One ball nut is screwed into the saddle bracket. I used blue locktite to hold that ball nut into the mount. rotation of the other ball nut is prevented by the bar across the bottom. The ball screw is welded to a machined end fitting that I made on my little lathe. I held the ball screw in my lathe chuck protecting the ball screw using some aluminum flashing. I then held the machined end fitting in the tail stock chuck and welded it to the ball screw. This worked well for holding everything straight. The bearing block holds 2 angular contact bearings to take radial and thrust loads in both directions. There is a smaller diameter bore that creates a step in the middle of the block that both bearing seat against. There is a second 3/4" thick plate that goes between the bearing block and the machine base to provide for more y axis travel.




This picture shows the X-axis ball screw assemble. Like the Y axis, the threaded end fitting is welded on. The other straight end of the ball screw was machined on my lathe using an indexable carbide tool. The left bearing block is very similar to the y axis one to take thrust loads and radial loads. The other bearing block uses a sealed ball bearing and only takes radial loads at that end. Part of the turned stub protrudes through so that I can still provide a handle for manual milling. If I decide to incorporate that feature. Again one ball nut threads into the block that bolts to the saddle and there is a spring and another ball nut that take up the backlash. You can see a little block that prevents this ball nut from rotating.



This is the machine base and you can see the cutout for the y-axis nut to protrude through. I had to open this up to make clearance for the new bracket. I also gained a little bit of Y axis travel.



This picture shows the Y-axis ball screw assembly in place ready for the saddle to be installed.



Here the saddle is installed with the X-axis ball screw in place. This part has to be mounted and aligned before putting the saddle on the base as the screws install from underneath.



This is the machine with the table installed, the gibbs adjusted, and the bearing preload adjusted.



This was a trial install to see how everything works dry. I still have to grease the bearings and the screws, but checking with the dial indicator, the backlash is zero in both direction and the movement is very smooth. With the gibbs adjusted, I can move either axis just by grabbing the 3/8" diameter end of the screw and turning it with my fingers. Needs a pretty tight grip, but I can do it. With the machine disassembled, I found that those Chinese craftsmen apparently didn't know the meaning of de-burr. There were many sharp edges that were contributing to stiff movement. I dressed those out and the machine movement is better than it has ever been.

This last picture shows the motor mounts that I have made for the machine with the NEMA 23 motors I already had. They are way overkill, but I had the 1-3/4", 1/8" wall tube and the steel plate already. As a matter of fact, I had all of the material I needed to make all the mounts. The tube ID was perfect for the locating step on the NEMA 23 motors and it protects the unsealed angular contact bearings in the bearing block.



As you can see in this picture, the motor for the Y hangs off the table a lot. I wanted to do direct drive, but I may later install a cogged belt drive and flip the motor around to keep the motor within the confines of the table.

Next step is to clean that messy bench and order the real motors. I have the drivers, BOB, and power supply on the way. Shouldn't be long now to get a 2-axis CNC. I will machine the Z axis drive parts using the X and Y axis CNC with machine pauses in the G-code for manual positioning of the Z.

I'll update when I get more.

[Bubba]

Looks good:})
On mine, when I started to do cnc, the X axis all of a sudden started showing lots of backlash. I found the mounting block for the ball nuts would loosen up and actually move. I ended up making a mount with extended "ears" and after getting it aligned, I drilled and pinned the block to the saddle. No more movement!

[HorridHenry]

I don't wish to put a spanner in the works but.....there's only two axis here,are you not supposed to have the turret a third axis?

[109jb]

I don't wish to put a spanner in the works but.....there's only two axis here,are you not supposed to have the turret a third axis?

As I said, to start with, I will do the X and Y axes. Once I get the machine running in X and Y I will use it in CNC mode to make the parts I need for the Z axis.

[109jb]

Looks good:})
On mine, when I started to do cnc, the X axis all of a sudden started showing lots of backlash. I found the mounting block for the ball nuts would loosen up and actually move. I ended up making a mount with extended "ears" and after getting it aligned, I drilled and pinned the block to the saddle. No more movement!

Unfortunately, the pocket where the X axis block mounts is pretty tight. No room to extend the ears in the fore-aft direction. Side to side i was limited to the thickness of the aluminum stock I had on hand. I may be able to drill and ream for one dowel pin in the centet between the 2 bolts. I'm sure I will have the machine apart again. Once I get it running I will figure out where to mount limit switches, wire anchors, etc. and will likely put some kind of lube system on for the ways. I'll see what I can do about that dowel then.

[hoss2006]

Very nice double nutting.
Pictures of the mill put the drive/motor selection into perspective, the table isn't quite as huge as some may have envisioned before.
570's should have no trouble with it especially when you go to 50V.
Hoss

[109jb]

I was planning to just put a y-axis way cover on to protect the y axis ball screw, and since the x-axis screw is under the table I wasn't going to do anything there. I will still do that much, but the more I think about it, I think I would like to put some kind of wiper on the ball screw nuts. I have seen it mentioned about using hot melt glue as a wiper. Does this work well? Is it as simple as lubricating the screw and putting the hot glue on the nut?

[Bubba]

You can see how I did some for mine here:
My CNC Homepage

[109jb]

You can see how I did some for mine here:
My CNC Homepage

Those look good. I like those kind of home brew solutions. I also liked your ball nut setup. Mine should work, but I'm not sure if I have enough preload forcing the nuts apart. I only have about 60 lbs which I am worried won't be enough. If you don't mind, I think I'm going to modify mine to take a setup like you have. I'll wait until I have the electronics hooked up so I can CNC the parts, but I really like your solution since I should be able to get just about any preload I want.

[arizonavideo]

Just crank the ball nut spring until it is flat. I bet that will be more like 150 pounds.

You only need the smallest amount of give in the spring like .002" or so, the metal flex in the spring even when flat should have enough travel.

[Bubba]

As far as imitating my setup, have at it. I put it up there to share with anybody that wanted to use it. Also it might give someone an idea on how to better improve it!

As for your preload, attached is a chart that I borrowed from the Original Industrial Hobbies site back when Arron had the company. As you can see, you don't really need massive amounts of preload force. As I remember (senior moment) I have between 80 and 90# of preload on my 5/8" ball screws. Remember, you might flex a spring pack on a roughing cut due to cutting load, but the loads on finish pass are quite light. If you have an excessive amount of preload, you will shorten the life of your assembly.

[SCzEngrgGroup]

As far as imitating my setup, have at it. I put it up there to share with anybody that wanted to use it. Also it might give someone an idea on how to better improve it!

As for your preload, attached is a chart that I borrowed from the Original Industrial Hobbies site back when Arron had the company. As you can see, you don't really need massive amounts of preload force. As I remember (senior moment) I have between 80 and 90# of preload on my 5/8" ball screws. Remember, you might flex a spring pack on a roughing cut due to cutting load, but the loads on finish pass are quite light. If you have an excessive amount of preload, you will shorten the life of your assembly.

Pre-load should exceed the highest machining force. A double ball-nut assembly will provide zero backlash ONLY as long as the spring pressure exceeds the machining force. Typical pre-load for a 5/8" screw is about #150. Even if you double that, on a hobby machine you won't live long enough for wear to be an issue.

Regards,
Ray L.

[109jb]

Thanks guys. With the setup I have now, the problem I have is that the floating nut can only be locked in 1/4 turn increments, and if the return track is in the wrong spot, I have to do a 1/2 turn there. On the X axis it worked pretty well to where I could almost max out the spring. On the y axis, I can't get there. Also, the springs I have will only give me a max of I think it was 80 lbs even then. Once I get the machine running I will make parts similar to what bubba has on his site. I will probably modify the design a little to match what I already have but it will be pretty similar. I have an idea how to make it to where I can fine tune it and have both return tracks pointing up. I'm going to shoot for about 150 lbs on the preload. Thanks again.

[109jb]

I forgot to mention that I got the stepper motors yesterday. Although I didn't have any direct contact with anyone at Automation Technologies Inc., I have to say that I wouldn't hesitate to do business with them again. I ordered on Monday morning and received the motors Tuesday. The online transaction was smooth and shippinng fast. They are only about 50 miles from me but I ordered online anyway and am very pleased with the results.

[109jb]

Well no further progress due to out-of-town business trips and 1,100 sq. ft. of hardwood floor that the wife says takes priority over my hobbies. There was also 1,100 feet of tile and old hardwood that had to come out first.

Right now I have all of the bits to complete the 2 axis setup and have my motors mounted. I just have to wire them up which I will do on a temp basis to test everything out.

Since I couldn't do any hands-on work on my mill, I have been working on my design for a control panel to use with EMC2. Here is what I came up with:



I then bought a $5 USB keyboard and a USB mini-keyboard with built in trackpad. The mini-keyboard is for use in the shop with EMC. The integrated trackpad will eliminate the need for a mouse and mousepad.



I think it will work great with the control panel and a small format monitor of about 8 to 10". Here is what the control panel might look like with a 9" monitor integrated with it. The mini keyboard is just about 1/2" narrower than this combo would be. I'd just put a little shelf for it and have a nice combo.



The cheapo keyboard I bought is a donor for my control panel. The first test was to see if EMC would tolerate the 2 USB keyboards hooked up at the same time. I tried it on my laptop running EMC and it can with either one or both keyboards controlling the software. So I liberated the circuit board from the $5 USB keyboard and spent an evening in the hotel figuring out the pinout of the circuit card for the keystrokes.

The AXIS interface for EMC2 has a bunch of keyboard shortcuts but I wanted to modify some of those shortcuts. I spent some time figuring it out which you can read about here.

One thing I am concerned about is if simply using a dpst pushbutton will properly activate the keystroke combination's. For example, when we do "shift-a" to get an "A", we press "shift" first then "a", whereas a dpst pushbutton will do "shift" and "a" simultaneously, or maybe even slightly in the wrong order. I asked how to get a little delay off of one side of a switch like that (capacitor?) over on the General Electronics Discussion forum. (link) I could remap all or most of the shortcuts to single keystroke combinations, but would rather not have to.

Anyway, I am back home now, but I still have about 400 sq ft of that hardwood flooring to install this weekend and will be out of town again next week. Hopefully I can eek out enough time to test my CNC setup before I go out of town again.

[109jb]

Well I had to quit on the flooring project at 10pm last night and tonight, so I did some work on the mill.

Last night I was able to get it temporarily wired up and did the stepper setup and a little tuning in EMC2. I was able to get 100 IPM rapids with the Keling/Automation Technologies 570 oz-in steppers running at 24V and 3 amps using the cheapo TB6560 single axis drivers. I had absolutely no problems with the setup and the machine seems to run very well.

Tonight I went out and ran a little test and cut some chips. It isn't much, just a run with a 3/8" end mill to knock off the top of a 2-1/2"x2-3/4" 6061-T6 block and then a little circular pocket. Right now I only have the X and Y axes running and I put in M0 codes to pause the program so I can position the Z axis. Below is a little you tube video of the first program cuts. I'm pretty happy with it. So far I have about $300 in the conversion

[ame="http://www.youtube.com/watch?v=g4CGC_xWl5w"]DIY Harbor freight CNC mill drill - YouTube[/ame]

[Bubba]

Looking good:}) Your almost there.
I did notice that you do not have a chip shield on the back of the saddle to minimize swarf etc dropping on the Y axis screw. My original one got tore up and I went to the local gasket shop in town and got a piece of 1/8" red rubber and made one for about $5. Saves a lot of crap going down on the lead screw.

[109jb]

Looking good:}) Your almost there.
I did notice that you do not have a chip shield on the back of the saddle to minimize swarf etc dropping on the Y axis screw. My original one got tore up and I went to the local gasket shop in town and got a piece of 1/8" red rubber and made one for about $5. Saves a lot of crap going down on the lead screw.

You caught that I was anxious to get the machine running on the steppers and just forgot to put it back on. I hadn't really planned on cutting anything last night, but that little chunk of aluminum was sitting right on the mill table and calling out to me "Mill me.... Mill me...." So I threw it on the vise and wrote that little test program and the rest is history. Right after the video was made I saw that I forgot to put the chip guard back on and wound up cleaning the aluminum chips off the back end of the y-axis screw. Its on there now.

[109jb]

Well i finally got my z axis setup installed. Thanks to Arizonavideo for relaying the idea of reversing the quill spring so that using the stock fine feed makes sense. I had previously planned to make a ballscrew setup but i think this will work fine. Now I just have to wait for the third TB6560 single axis driver board and I'll have 3-axis control.

[skipper]

Great build I am planning on doing the same to my mill some time soon (still a newbie). What do you mean by reversing the quill spring? Do you simple turn it around and what does that do? Do you still get back last in the z using the fine feed adjustment and not a ball screw?
Can you show more details of your machined ball screw ends?

Thanks!