109JB's G0704 conversion (pics now, plans soon)

Well, I finally sold my round column Harbor Freight mill (link) last week so it is finally time to get my G0704 up and running on CNC. I had measured, designed and made all of the parts for the G0704 on the round column mill, but had to wait until it sold so that i could buy the steppers, drivers, and BOB. I got all but the BOB and now have the mill mostly assembled. It has gone together and been taken apart probably about a dozen times over the past few months, but hopefully this is the last time before it makes some chips of its own. I already have some parts I need to make for my other numerous hobbies.

The summary of the conversion and the components, along with some future plans is as follows:

- Automation Technologies (Keling) 570 oz-in steppers on X, Y and Z
- Automation Technologies (Keling) 6050 drivers
- 48V 13.5A Dell server power supply
- Chinese manufactured parallel port BOB
- Salvaged 2.6 GHz pentium 4 PC
- USB Game controller for a pseudo pendant
- LinuxCNC (I am going to give Mach3 a try this time around though)
- Thompson rolled ballscrews with double nuts on all axes
- Angular contact bearings on all ballscrews
- Machined aluminum end plates on all axes
- Lapped ways on all axes
- Way lube system started (ready to plumb)
- Home/limit switches planned using proximity switches
- 4th axis planned (already have a 6" rotary table)
- Possibly a different spindle motor and belt drive

As for the machine tour, I'll start at the end and then go through the conversion.

Below is a picture of how it sits in the barn right now. Well almost. In this picture I had forgot to put the Z-axis way cover on which has since been corrected.
http://i521.photobucket.com/albums/w...n/image14.jpeg

Below is a pic of the Z-axis ball screw setup. I found that the Thompson ball screws are too hard for the lathe tooling I have to be able to cut threads, so I welded on a chunk of 5/8" diameter steel rod on the end of each ball screw and then proceeded to machine that. The threaded end just has a 12mm straight section followed by a 12mm fine pitch thread on it, That's all. You'll see why later. In this picture you can see that the double ball nut arrangement has been installed.
http://i521.photobucket.com/albums/w...n/image01.jpeg

Here is a closer view of the double ball nut. Starting from left to right in the picture. First is the "fixed" ball nut that is threaded into the aluminum slide mount the attaches the Z-axis slide. This part is in 2 pieces with a tight fitting key. This was done that way so that it would fit through the top of the column while assembled on the screw. The second ball nut is next. Its threaded end is simply protruding into a 1" diameter bore in the end opposite of the slide mount. on the right end is an aluminum plate that has a stack of wave washers between the plate and the ball nut This plate is attached to the slide mount with a pair of 1/4" diameter bolts to provide preload to the ball nuts to eliminate backlash. The preload on all axes is about 120 pounds.
http://i521.photobucket.com/albums/w...n/image02.jpeg

Here is the column top cap with stepper standoffs and the end of the ball screw. On each of the ball screws, I have a wave washer in the assembly stack to provide preload to the angular contact bearings without having to monkey around with double nuts.
http://i521.photobucket.com/albums/w...n/image04.jpeg


Here is how my motor couplers are made. These are home brew Olham couplers made from aluminum with a HDPE puck. The motor side just has a 3/8" reamed bore with a 1/4"-20 set screw to lock it on the shaft. The ball screw side has a 12mm threaded bore and after threading, I split the coupling. I have two 1/4"-20 SHCS that clamp the coupler onto the threaded end of the ball screw like a split collar. With the threads, I can put the SHCS's in just snug and thread the coupling half onto the ball screw. I then tighten it up to compress the wave washer for preload and then lock the SHCS's. Seems to work great. We will see how it works in operation soon.
http://i521.photobucket.com/albums/w...n/image15.jpeg

Here is the y-axis ball screw setup. Same principal as on the z-axis, just different geometry. Also, I cut the threaded end off of the floating ball nut to reduce the length of the assembly.
http://i521.photobucket.com/albums/w...n/image05.jpeg

And the Y-axis motor mount/thrust block. Same angular contact bearings, but here, I have a little aluminum puck with a recess that acts as a pseudo labyrinth type seal to keep crap out of the bearings. The angular contact bearing on that side protrudes from the mounting block a bit to accommodate the seal/puck. You can see on this one that the wave washer is between the puck and the coupler.
http://i521.photobucket.com/albums/w...n/image07.jpeg

Carriage installed, Y-axis hooked up, and X-axis screw in place.
http://i521.photobucket.com/albums/w...n/image09.jpeg

At this point the table is installed and the end plate at the floating end of the ball screw is installed. On the floating end of the x screw, I did machine a 12mm diameter into the ball screw to fit the bearing.
http://i521.photobucket.com/albums/w...n/image10.jpeg

And the opposite end awaiting the coupler and motor.
http://i521.photobucket.com/albums/w...n/image11.jpeg

Here are a couple pics of the Y and Z axis with motors and couplers installed. Somehow I didn't get a picture of the X axis at this point.
http://i521.photobucket.com/albums/w...n/image12.jpeg
http://i521.photobucket.com/albums/w...n/image13.jpeg

Well that is about all for now. I think this setup is going to work pretty well. Each axis moves very smoothly after lapping. Next is to wire up the electronics. I already have the PC set up for dual booting LinuxCNC and Windows XP Pro with Mach3 and LinuxCNC installed. I am still waiting for the BOB to arrive but will have the power supply and drivers temporarily wired for testing tomorrow. As soon as the BOB gets here I can start moving under CNC control.

I will post my plans of the parts for this conversion when I get it tested and have a chance to clean up the drawings a little.

Once the mill is up and running I plan to convert my little 9x20 lathe to CNC.

Enjoy,

John B.

looks awesome I am really interested in you design. If I can not achieve the accuracy I would like I may have to do a double nut with preload like you.
Are you concerned about rigidity of the y axis with only 2 stepper standoffs?

Actually, there are 3 standoffs. Just can't see the one on the bottom in the close up of the Y-axis motor. If you look at the first picture you can kind of see it. With 3 I think it is fine. I had to go with 3 to clear the carriage. Instead of rotating 45 degrees, I rotated something less than that so that it would clear the carriage and also be able to access the mounting nuts for the bearing block.

What method of lapping did you use? Scraping, Sanding, etc.

I have the same rags, Sams Club?

I look forward to following this.

For lapping, I just used Mothers Mag and Aluminum polish because I had some on hand. Since it is pretty non-abrasive, it took more than with a real lapping compound, but I was just looking to make things move a little smoother. I applied the mothers to the slide and then adjusted the gibs to where I could slide it but with some noticeable drag. I would guess I did about 20-30 strokes, cleaned, and repeated about 3-4 times on each slide depending on how it felt. Each slide had definite areas that were tighter than others. I just did enough that the slide felt the same along the entire length. Didn't use but about 1/8 to 1/4 of the small tub of polish total. I then cleaned everything in my parts washer.

Quote:

---

Originally Posted by BAMCNC.COM

I have the same rags, Sams Club?

---

Sorry, I just found this funny as hell and I couldn't find an emoticon for it.

Possibly just too much beer this evening...

Didn't have much time to work on the mill this weekend, but I did get my power supply converted so that I can use it on this project. This is a 48V Dell server power supply rated at 13.5 A. The nice thing about using this is that in addition to the 48V output it also has 12V and 5V outputs that can be used for other things. Initially I will use the 5V output to power my BOB and the 12V output for my proximity switches I will use for home/limit switches. On my last conversion I had to have a separate 5V power supply for the BOB and would have had to have another 12V power supply for proximity switches had I put them on. On top of that it is much cheaper than the typical power supplies used for CNC at $40.

Here it is after "converting" it.

http://i521.photobucket.com/albums/w...n/image16.jpeg

The power supply is a "hot swappable" power supply. When used as it was intended, there are generally 2 of these on a server so that either one can be removed and replaced without powering off the server. The power supply is slid into a rack on the server and it powers on once the AC and pin connections are made. It has a typical jack for a PC power cord for the AC, 2 blade connections for 48V and a 24 pin connector for everything else. To convert it you have to figure out what the 24 pins do. Without going into the boring details, there are 12V pins, 5V pins, ground pins, and pins that turn the unit on. After a couple hours of probing pins with a volt/ohm meter I was able to get it to power on. I then soldered wires to the pins to tap the 48V, 12V, 5V, and grounds that I needed. Had to solder because I didn't have matching connector. I then added some screw type terminal strips to make future connecting/disconnecting easier.

Here is a shot of the terminal strips showing the outputs I will have.

http://i521.photobucket.com/albums/w...n/image17.jpeg

Once I have the machine up and running with at least a few hours of test time on it I will post the details of wiring this power supply along with the drawings of the mechanical parts for the conversion. I am hesitant to post them now since it hasn't been tested yet. Especially the power supply since, well, it is electrical and I'm no electronics tech.

Well I got the BOB today so went out and wired up the steppers, drivers, etc. I got it running and did some tuning. Here is a crappy video from my phone of those results.
https://www.youtube.com/watch?list=U...&v=MxRA1i9OzPk

In the video, the machine is running 180 IPM. I have it set there right now because the X axis has some occasional stalling of the stepper at 240 IPM. I checked around and I think it is my gib strip that is slowing that axis down. It seems tighter on the skinny side than the fat side. I can get the x-axis to run a lot faster just by loosening the gib, but then I have some table slop I don't like. I guess I will do some more lapping on the x-axis gib. The Y and Z will run up to 240 IPM without any problem.

On another note, my home brew oldham couplings are working like a charm as are the double ball nut arrangement. I checked backlash on each axis and I have about 0.0010 on X, 0.0002 on Y, and 0.0005 on Z. With the gib problem I have on the X, I think the backlash on that axis will come down a bit one I get it fixed. I also did some repeatability checks by setting an indicator up with each axis at one end of its travel and then running the axis full travel back and forth at rapid speed 10 times returning o the start point. I couldn't read any error on the indicator, so at 180 IPM it doesn't look like I am losing any steps.

The Dell server power supply worked like a charm and I had it running for about 3 hours tonight, so I think it will be fine. At some point, I will put a volt meter on the output while running to see if there is much voltage drop under load.

Things still left to do before making chips:

- Put electronics into an enclosure
- Tidy up the wiring
- Add home/limit switches
- finish oiling system
- Fabricate a way cover for the front of the y-axis

Carbide bits will easily cut the case hardened ball screws.

What diameter of ball screw did you use on the Z axis? Your backlash numbers look very good.

Would you happen to know how heavy the head is?

Quote:

---

Originally Posted by H500

Carbide bits will easily cut the case hardened ball screws.

What diameter of ball screw did you use on the Z axis? Your backlash numbers look very good.

Would you happen to know how heavy the head is?

---

Yes the case hardened screw can be cut with carbide. That's how I cut the floating end diameter. However on these screws cutting down to 12mm did not get beneath the case hardening so I wasn't able to cut the threads because I don't have a carbide threading tool. Welding the ends on was easier for me to do anyway. Just cut off a piece of 5/8" rod, weld it on and then machine it. Doing it this way I was actually able to just use a theading die.

All of the ball screws are 5/8 diameter with a 0.200" lead

Don't know exactly how heavy the head is but I'd guess 40 pounds or so.

Heat the screw end untill dull red and let it cool slowly. Placeing a can on the end to hold the heat would let it cool slower. It should be soft after that, then thread.

Guys, I know all about the different methods I could have used, but when you don't have a set of oxy-acetylene bottles, don't have a carbide tipped threading tool, but do have 5/8 round stock and a welder you do what you gotta do. It worked perfect welding a little piece on the end anyway so what difference does it make?

On another note, I did a bit of checking and confirmed that the x-acis problem is a bad gib. I set up an indicator ad shown in the picture and at the right end, I am getting about 0.004" of play when I shove the end of the table fore-aft. At the other end the play is 0.000". So I apparently have a gib that needs to be 0.004" thinner at the skinny end or I need to shim the fat end.
Attachment 171429

How are the gibs adjusted in those? Like a Sieg with 3 adjustment screws and flat gib, or a Taig with a tapered (wedge) Gib and 1 screw.

I think your welding method was a good idea, and I bet a lot of people would be doing it if they had thought of it. Did you harden the steel after machining?

The G0704 has tapered gibs. I think for right now, rather than pulling the table off again, I will just get some allen screws and jam nuts and use the threaded holes for the table locks to adjust the gib for now. I will still use the stock adjustment screws to position the gib left to right.

Later on I will pull the gib out and use some wet or dry sandpaper spray glued to a granite block to adjust the taper on the gib. I may just machine and scrape a new gib as I think this one will protrude too far out of the left side of the table once I adjust it.

As for the welded end on my ball screws, I didn't harden it afterward. Didn't see a need.

I really have to get some home/limit switches on this thing. I was playing around with it last night and ran the table a little too far and BAM!!!. My previous machine I had set for only 72 IPM, I was only running 24V and 3A on 5A motors so the power was way lower, and hitting a hard stop wasn't a big deal. That was actually how I homed it. I ran it to the hard stop and then backed of 0.100". Not going to be doing that with this machine. The good news is that I have the proximity swithces I am going to use already. The bad news is that I fried one testing it due to a brain fart on my part so now I only have 2 switches and have to order another one. I'm just going to order 5 more since I will need 2 when I do my lathe and want a few for backup. They are only $4 each so no big deal. Just have to wait for them.

I also trammed the mill in finally last night (I was up all night tending to the 14.5 lb smoked brisket and 12 lb turkey I am doing for today). The tramming came out great. The the column is square to the table within 0.0005" over 5" right now in both directions without having to shim. The head tilt got the spindle is dead on. I haven't run the table yet to see how flat it is. I have read about some problems others have had in this regard, but most of my stuff isn't too critical anyway. If I have something that critical I will use a fixture plate trued before use anyway.

Another thing I noticed is that you can't really get full travel on the Y-axis the way the column is. I will be making 2" to 2-1.2" spacers for the column and the head. Haven't decided if I will use aluminum or cast iron for them. Probably Aluminum since I have a 6" round of 6061-T6 that is about 3 feet long that I can make them out of. That is something I will make a trued up fixture for.

Well have to get ready for company. Everyone have a Happy Thanksgiving.

Quote:

---

Originally Posted by 109jb

...... The bad news is that I fried one testing it due to a brain fart on my part..........

---

Ah ha haven't heard that one before:p:p

Nice thread, love the pics and the detailed descriptions and this looks like the start of a very interesting and good build...thanks for sharing:cheers:

Quote:

---

Originally Posted by 109jb

Guys, I know all about the different methods I could have used, but when you don't have a set of oxy-acetylene bottles, don't have a carbide tipped threading tool, but do have 5/8 round stock and a welder you do what you gotta do. It worked perfect welding a little piece on the end anyway so what difference does it make?

---

As long as you didn't anneal the working area of the ball screw and machined the part after welding ( for concentricity ) it shouldn't make a bit of difference.

Quote:

---

On another note, I did a bit of checking and confirmed that the x-acis problem is a bad gib. I set up an indicator ad shown in the picture and at the right end, I am getting about 0.004" of play when I shove the end of the table fore-aft. At the other end the play is 0.000". So I apparently have a gib that needs to be 0.004" thinner at the skinny end or I need to shim the fat end.
Attachment 171429

---

Tapered gibs have to be hand fitted by scraping. That only after verifying that the matting ways are parallel and proper. The big problem is verification of the ways, scrapping in a gib isn't that big of a deal.

Quote:

---

Originally Posted by wizard

As long as you didn't anneal the working area of the ball screw and machined the part after welding ( for concentricity ) it shouldn't make a bit of difference.

---

Yep. The weld affected zone isn't where the nut will be running and the 5/8" rod was welded on first and then the machining of the end was done afterward. However, on my round column machine I also welded a piece on but the piece was machined first. I then chucked the ballscrew in my lathe chuck and the end piece in the drill chuck on the tailstock to hold it in alignment for welding. Doing it that way also worked fine.

Quote:

---

Originally Posted by wizard

Tapered gibs have to be hand fitted by scraping. That only after verifying that the matting ways are parallel and proper. The big problem is verification of the ways, scrapping in a gib isn't that big of a deal.

---

Yep. If/when I make one I will machine it close and then scrape it in for final fit.

Not much done today as we are used car shopping. All I got done was a little test of the proximity switches I am using for my mill. I bought the cheapest ones I could find on e-bay at $3.86 each which includes shipping. I have heard differing stories as to the repeatability of these switches, so I decided to do a test. Got one mounted on my X-axis in a crude fashion and wired up to my BOB and ran the test. You can see the results in the video below, but in summary, I am getting repeatability of 0.0002" or better. I tried it a few times in the video and about 100 or so additional times and each time it was within 0.0002". Probably about 90% of the time I could not see any difference in the indicator reading. Not shown in the video, but a concern of mine was how much overrun there is if the limit trips. I found that at a rapid speed of 240 IPM, there is about 0.012" of overrun how I have my machine set up right now. It doesn't come close to crashing into the switch.

http://www.youtube.com/watch?v=IGO47l62T2o&feature=channel&list=UL

I just came in from the barn and have a couple of reports. I had been using a usb to mini USB to power my BOB from the computer because it was easy and I wanted to get it running. Tonight I wired up the 5V from the power supply and found that the 5V from it doesn't have enough current to run the BOB. The info tag doesn't list any 5V output, I just found some pins that were putting out 5V and tapped into them. I guess since it wasn't listed, I should have known that the 5V was not going to have much current ability at all. The instructions for my BOB don't specify how much current it needs, but I checked around and other BOB's need about 500mah. In any case, the 5V on the P.S. won't turn on the BOB. No biggie. Just a little snafu. The 12V outputs are rated at 4A, so I will just wire up a 7805 voltage regulator to get the 5V and the current needed for the BOB. Right now the BOB is all I have that needs 5V and is all I plan to have that needs 5V.

While in the barn, I tried, but couldn't resist the urge to mill something. I didn't get a picture, but it came out very well. I milled a slug spacer out of 3/4" 6061-T6 bar that I need for turning a chuck adapter on my lathe. It has an ID of 1.54" and an OD of 2.4". I cut it using a spiral cut path with a 3/8" 2 flute end mill. Feeds speeds and depth of cut were:

2300 rpm
10 IPM
0,050" DOC

Nothing too impressive, but it cut well and didn't have any problem. The ID was cut first and came out great, within 0.002" of the target without any fiddling with the program (+ or - 0.005 was all I needed for this piece.). The diameter is round within 0.001" measured at 0, 90, +45, and -45 degrees. The OD had some chatter because the 2-1/2" wide bar was hanging off the end of the vise. The finish looked good until the remaining material at the bottom of the cut got too thin which is what caused the chatter, but I expected that. Nothing that a quick pass on the lathe couldn't fix.

I have to say that this little machine does extremely well for its size.

Not much work will be done on it for the next few days though because the wife has some home projects she wants done.

Good video and good explanation thanks. ...