View Full Version : CNC conversion lead screw

10-16-2003, 12:13 PM
Hello all,
My current problem is selecting a new lead screw size for the little mill drill I obtained from Harbor Freight. You know the one that sells for about $500. I took the machine apart and was pretty pleased with the fit of most parts, but there was this nasty red grease all over everything and some shavings and burs to take care of. The people from HF were very nice to me by the way.

I have to extend the lead screws for the stepper motors (200steps per rev) I'm using. I need the extra length to keep the hand wheel as well. The lead screws appear to be an acme style screw with an OD of .622" and a thread pitch of 16tpi. (thus that wacky thread dial in 1/16ths). As this is a low budget project that I've already got too many of my own dollars in,ball screws are out. My current thought is to purchase acme rod and a matching tap to make new nuts. I'm leaning towards a 1/2"-10 as it seems that it will make the calculations for setting up the CNC easier and the rod and tap is fairly reasonable in price.I can then also more easily make Inch thread dials. I will have to adjust for the thrust bearings in the table though.

Is there any advantage/disadvantage to using a particular thread pitch or size? Your input is appreciated, just keep in mind I'm cheap, so the laser guided roll formed ball screws are out of the question http://bbs.homeshopmachinist.net//smile.gif.

Thanks, Matt

10-16-2003, 12:31 PM
Just read someone elses post on this in a different forum and he made an interesting point.

The question was why not use ordinary lead screws (instead of ball screws) and use a DRO output to control the CNC. This way you don't need a zero backlash screw.

The answer was that the backlash in the ordinary lead screw will yield poor results even with a DRO since the table will move (the amount of movement is dictated by the amount of backlash) when you make cuts where the tools edge is pulling along the work peice (I think that is called a climbing cut?). When the tool starts cutting the other way it can push or pull the table slightly and distort the cut. My mill has 0.025" of backlash (not a CNC machine).

Obviously, since you are after the low cost solution you need to be aware of the ramifications of using standard lead screws. There may be other ways to help compensate for the issue, too. Maybe someone here has a good plan.

10-16-2003, 01:05 PM
Backlash will be an issue. Take for instance milling a circular hole in a plate. The X and Y axis drives will both have to change direction and there is no opportunity to program in some sort of backlash compensation. However, there is a way to eliminate backlash, at least between the leadscrew and the leadscrew nut. It's an old trick to rehabilitate a worn crossfeed for instance. Make a nut with the same internal thread and material as the original leadscrew nut, about three or four pitches long. Wind this nut up to the leadscrew nut until backlash is just removed. Pin it to the leadscrew nut or otherwise fasten it in place. Then look for causes of backlash in the leadscrew mounting. These are usually easily fixed/adjusted out. You should be able get the backlash well under .001.

I point out that the driving forces will force the two nuts together, not apart, so the fastening of the two nuts only needs to prevent relative rotation, not separation.

[This message has been edited by Evan (edited 10-16-2003).]

Bruce Griffing
10-16-2003, 01:32 PM
What Evan describes is also called a split nut. They are available commercially or you can make your own another way. You just internally thread a block and then nearly split it - like a cotter. Then use a screw to tighten one half against the other to remove backlash. It will need periodic adjustment to take up wear, but it will work fine.

10-16-2003, 03:01 PM
I'm aware of the backlash problems. I know I'm trading accuracy for price. I think your estimate of .025 is larger than I will get, none the less it is still an issue I will have to deal with.

There are a couple of reasons we're building the machine. #1- 2nd mill to demonstrate G&M code basics to students.#2- Portable, I'll be able to transport it to the middle school and hopefully wow kids and get them interested in my classes.#3 Its a great project for my seniors to work on.#4 I'm having a blast doing it. It makes my job more interesting.#5 My wife likes it when I spend my own money on school projects.

10-16-2003, 03:36 PM
Matt: After seeing Evan's auto cross feed in another post, I thought of ways to get at my cross slide leadscrew. Here's one:

You could drill / ream either end of the leadscrew, epoxy a length of ground rod (or a turned coupler with ODs to match the ID of this new hole and the OD of your stepper. If you leave enough clearance for a thin film of glue between the parts (and rotate for good centration), I'll bet it would take a torch and hammer to remove it again http://bbs.homeshopmachinist.net//biggrin.gif

If you decide to try something like this, be sure to choose the OD wisely. I'd hate to glue the thing together and then find that it could no longer pass thru a leadscrew nut.


10-16-2003, 04:37 PM
Thanks for the suggestions. I was already contemplating the split nut idea. There should be room for it. I'm leaning on replacement of the lead screw because I want to get rid of the 16 tpi thread and go to something like 10 tpi or some number that makes more sense... at least to me. I saw evan's drive mechanism (very nice) but making the extension concentric with the screw and lining it up with the motor and suports is probably difficult given the quality of these small mills. I'm also trying to avoid using belts and pullys so some sort of shaft coupling will be made.

Thanks, Matt

10-16-2003, 05:51 PM
This idea was mentioned by Cass, I believe, some time ago, for eliminating play. Hang a weight from a cable, over a pulley, and attach the cable to the table. The backlash will then be zero, as long as the weight is enough to overcome friction in the ways, and whatever cutting forces will be pulling on it. Whenever possible, cut towards the direction that the table is being pulled by the hanging weight, so the play will be taken up against the leadscrew already. One advantage of this method is that the force against the leadscrew, it's bearing, and nut, will be constant, unlike say a spring used in the same way. Another idea now, why not put a spring between the extra nut and the main nut on the leadscrew- that would take the play out of the leadscrew, but not it's bearing. At any rate, that spring would be somewhat compressed, and stay pretty much at the same tension. Same problem though, tight gibs and cutting forces will overcome the tension and allow play at some undetermined point. Maybe the wear rate would be too high on the screw and nuts, though, with either of these methods.
Another interesting thought, if the gibs are loose enough to allow the table to move, there's also the possibility of the table skewing a bit, in x or y axis. If some considerable time is put into eliminating play, then this source of unwanted movement should be dealt with, as well. For me, this effect is most pronounced when hand cranking the table, every turn of the crank rocks the table back and forth once. It's a compromise between too tight gibs and free movement. At any rate, that's also an error showing up on the workpiece. What does one do for that, spring loaded gibs?

Bruce Griffing
10-16-2003, 07:06 PM
I have a Griz minimill that I retrofit with a power feed. My connection to the x-axis was made by removing the lead screw and turning the threads off just at the end. This created a nice cylindrical end. I cross drilled it for a pin and made shaft bored out at the end to mate. With the pin through both everything worked as soon as I had the new bronze bearing at the left end aligned. The Y axis will be a little harder and you should note that the minimill y axis lacks the roller thrust bearing that the x axis has. If I were doing what you are planning, I would plan to fix that also. Otherwise backlash on Y will be out of control.

10-19-2003, 08:18 AM
Thanks for the suggestions again. Bruce, the Y axis surprised me when I took it apart and I found no bearings in it. I'm not sure what yet, but I plan to address that too. The other problem that is actually a bigger backlash problem is the Z. The rack and pinion used in the machine is pretty sloppy. The original setup has a universal joint connected to a worm gear connected to the rack. When you sum up all the slop it is pretty significant although I didn't measure it. I'd like to keep the rack for coarse adjustment by hand, but maybe add a lead screw on the Z to make it more accurate.
Thanks, Matt

Bruce Griffing
10-19-2003, 01:57 PM
You may not be aware, but there is a small Yahoo group on CNC minimills. Here is a link.
As to the z axis - I agree. If you want something other than a two position z axis, you will need a lead screw. If you are happy with a cut/nocut setup, then you could use the stop and make a drive that would forgive an attempt to drive past the stop.

10-19-2003, 08:32 PM
Call Wedin - this is an acme screw / ball screw company. I believe they are in Ohio. They have a great selection and good cost. Both in ballscrews and Acme screws and components. I have done biz with them for a 3/8 ballscrew for a retrofit for a person (one I helped design and build). Cost was great, the help fantastic.


10-20-2003, 08:45 AM
I sent Wedin an e-mail.
Looks like they just may have what I need.