Return to machining page
 

Description of my Ball Screw Conversion
I decided to try saving some money and make up my own ball screws from material I purchased from McMaster Carr. I will add the drawings and dimensions to this page as I have the opportunity. My first plan was to try grinding the screw stock to the required dimensions using a 4" 90 deg grinder clamped in the vice of the ShopTask. After messing around with this proceedure for a few weeks I cam to the conculusion that it was too slow and that I was not able to get everything set up rigidly enough to get the finish I wanted.

So, it was off to plan B. I figured if I heated the end of the screw to a nice cherry red and let it cool nice and slow the ends of the screws would be annealed and I could turn them with a carbide bit. Well, guess what. the screws are air hardening steel and I managed to achive was to increase the hardness od about 6" of my screw stock!

Well, plan C. I decided to machine the ends out of some round steel stock and then to attach these machined ends to the screw stock with a coupler sleeve and have a friend TIG weld them in place. I proceeded to turn out the ends and make up some coupler sleeves. On the Y axis I sized the coupler to be a press fit to the machined ends and I used an arbor press to put them together. I made the fit to the screw stock a shrink fit and heated up the coupler and slid it onto the screw with a little help from a hammer. If I was doing this again, I would use larger diameter (aprox 3/4") steel rod and make the couplers and intregal part of the extensions so there are only 2 welds per screw instead of 4. The disadvantage is you have to turn more off the rods, but the extra machining time would be worth the effort. On the X axis I made the coupler to be a loose hand fit onto the screw stock and this proved to be the better method. On the Y axis I had to get the runout on the extensions down to a reasonable number before I had the shafts welded and of course they moved so I had to realign then again after everything else was complete. To weld the X axis we had to come up with a way to fixture the parts during welding as the coupler fit was too loose for the extensions to hold position by themselves. So, we ended up putting the screw in the lathe 3 jaw chuck and using a dead center to align and hold the extension. This method worked far better, requiring me to only align the shafts once and the amount they were off was less than the I got on the Y axis that was press fitted. I ended up with < .005 runout on all the shaft extensions.

Now for a further update after about 4 months of use.
From the day I installed the ballscrews I would get a cracking type noise on the X axis much like the sound you get when adjustable cup and cone ball bearings are adjusted too tight. I attributed this to a ball nut and screw combination on the X axis that was on the tighter side of the tolerances as the X nut did not move as smoothly with the shaft off the machine as the Y axis nut did. Over time this cracking got more frequent and even seemed to cause some binding and lost steps on the X axis. So, I finally decided to pull the X axis lead screw and see if I could figure it out. With the lead screw removed the ball nut now moved smoother than ever and I noticed nicely polished tracks where the ball bearings ride on the screw stock in place of the original black slightly textured surface finish. Looking at the adapter I had made to connect the ball nut to the table I noticed that the threads the ball nut screws into were cut at a slight angle causing the ball nut to be cocked at an angle to the screw. I made a new adapter and this time I made sure the threads were straight by holding the adapter in the 3 jaw chuck and using a center in the tail stock to make sure the tang end of the tap was in line. I reassembled everything with this new adapter and while putting it back together replaced the small allen set screws in the saddle clamp and the CNC belt drive gear with good quality USA screws in place of the soft screws originally on the machine. I also drilled and tapped the CNC belt drive gear for a 2nd set screw to try and keep the gear from getting loose on the X axis shaft. I am happy to report that the X axis is now smooth as silk and has no binds and no more cracking noises.
 

Bill of Materials
ball screws - part#  5966K26  2- required
ball nuts - part# 5966K16 2- 2' lengths required for plan C
You can find these on the McMaster Carr web site by searching for the term "ball screw" with the option by keyword selected. Follow the link to ball & lead screw assemblies that comes up.
1-1/4" dia aluminum or brass - aprox 6" length required
3/4" dia steel rod - aprox 3'

Special Tools required
Tap to thread ball nut to machine adapters
Die for X axis
Die for Y axis

Pictures of the X axis ballscrew and ball nut

If this project sounds a little too big for you consider the ballscrew kit offered by C & G Research.
Click to visit C & G Research

Here is an update on the ballscrew machining that I posted to the Shoptask Users forum on Delphi:
Well, I finally tried doing what I have been chicken to do for over a year and it works fine. Over a year ago I built myself a ballscrews conversion for my Shoptask. At the time the machine was brand new and I was not confident in it's capacity. I asked around about how to get the ends of the screw stock shaped as I needed it and the answers I got was that you could turn the screws in the lathe if you take a real deep cut to get under the case hardening (seemed well beyond the capacity of the machine to take a 1/8" cut into hardened steel) or you could grind them using a toolpost grinder (one person even told me they used a disk grinder as a toolpost grinder, but I could not get enough rigidity to make this work). Well, if any of you have read the story on my website you found out I came up with a third option and machined ends that I had welded onto the screws.

Now I find myself working on another project that I am putting ballscrews onto and feeling more confident in the machine I decided to try machining the end of the ballscrews. So, I adjusted the machine's pulleys to 280 RPM (choosen by the SWAG method) and put a TIN coated carbide insert into the tool holder and proceded to take a tentative whack at it and only went in about .050" deep. Well, I was pleasantly surprised that it cut just fine even thought I was getting a lot of push back on the X axis as it was hitting the side of the threads (I was actually turning the cranks by had to feel what the cutting force was like). So, I backed away from the piece and turned in the Y axis to try cutting the full thread depth all at once. It went through the part I had already cut just fine and continued into an uncut area just as easy. After cutting about 3/4" onto the shaft I backed off and looked at the bit. The carbide was chipped on the top lip, but was still cutting. I changed to a fresh un-coated insert figuring the chipping was caused by the interrupted cut I first made, but not wanting to damage another surface of the coated insert if I was wrong. I proceeded to cut another 3/8" of shaft in one whack just fine and the insert held up just dandy.

Update Again March 2001:

Another project and another method. I was going to do another project and had a piece of ball screw stock that was long enough for the threaded section, but too short to form the unthreaded section the bearings and pulley attaches to. Got me thinking about any way to do it without using an external coupler which whould shorten the useable length of the threaded portion. Decided to try creating a stub that had the proper form for the bearings and the pulley and a 3/4" long X 3/8" diameter stub on the end that needed to attach to the screw. I then center punched the screw and proceded to use a hunker center drill and then a 3/8" cobalt drill bit to put a 1" deep hole in the screw. THe hole was drilled just fine and I then used some 1100 deg F silver solder to sweat braze the 2 parts together. It has worked great so far and is the easiest method I have found to date.