View Full Version : Alternate Linear Rails??

10-01-2003, 03:19 PM
I am about to start on a steel framed machine and was wondering what material I can use for linear rails. I was thinking of using 3/4" drill rod or something of that nature with bronze bushings.

I can't affortd THK rails and such. My tolerance requirements are not 0.001"... they are more like 0.050"

I intend to use the machine for 3D sign cutting and 2D plasma burning of 1/4" plate steel. The machine frame size will likely be about 3' x 4' or so.

What size of machine frames have you guys found the most useful?


10-01-2003, 04:35 PM
Figure out how big the stuff is you want to cut. Make it as big as the largest size.

That said, smaller is much less complex, and the odds of you finishing the project go up inversly relative to the complexity of the machine.

We all want to see people succeed, so start small. :) :)


10-01-2003, 06:56 PM
Real Thompson shaft isth that expensive...I can buy 1" Diameter for 90 Cents an inch, and the bushings (sealed rollers) are 27 bucks each....if you go smaller, the price goes down..;)

10-01-2003, 08:41 PM
Price is relative. What is too high for you?
I just bought sets of 8' , 4' and 14" THK rails w/bearings for about 750 total on ebay. And I didn't shop at all.
I got some 53" THK rails w/bearings I'm trying to unload now (got a few too many in one deal).
I understand that some people build their whole machine for under $300, sometimes using plumbing pipe and skate wheels for the linear bearings.
It all depends on what you want and how professional you want it to be.

10-01-2003, 09:51 PM
Hey Samault,

Well I am extremely price conscious right now because I'm currently unemployed.:( I'm trying to do a Chuck Knight style machine, but want a little more precision than black pipe rails and skate bearings. Drill rod or simple BMS bar is a little more expensive than black pipe, but still an order of magnitude cheaper than linear rails.

Like Chuck, I also scavenged steppers from old laser printers. I got four of the 100 oz. in. motors that I intend to apply to relatively fine pitch ready rod (5/8-11 or 3/4-10). I'll use delrin split nuts as Chuck did to reduce backlash. The fine pitch ready rod provide for some torque multiplication which may be essential for these small motors.

I should also mention that this is my first homebuilt machine, but I am quite familiar with machine control firmware and hardware. I intend to utilize some parts from the scavenged printers to build my own driver boards, and will build my own opto-isolated parallel break-out board.

The z-stage will be dual function to hold both a PC-690 router (whichI already have) or a plasma torch (which I will rent when required). I would like the machine to be rigid enough to mill aluminum with a very light feedrate.

The plasma work and metal work are both for hobby work as can be seen on my web site:


Obviously, I will probably document my CNC router construction :-)

Thanks again,

10-01-2003, 11:01 PM
I have my own "style" of machine? Yeesh...I've got a reputation, already! :-)

Hi, Eldon. Glad to see you over here. The focus is a little different than on CAD_CAM, isn't it?

The type of machine I built is heavily inspired from John Kleinbauer's designs...and of course there are changes I would make to it, even now.

The best advice I can give you is to build *something* right now, get it working, and then build your real machine. Even in the construction of something very small and simple, you'll learn more than you would just reading about it, and planning your machine. There truly is no substitute for *hands on* experience.

Go get some MDF, some black iron pipe, some skates from Goodwill, and build *something.* Hit the scrap bins at Home Depot and Lowes!

After you've made all of your mistakes on the cheap machine, then you can incorporate all the improvements you would want to make, into a bigger and better machine. You'll probably save money in the long run...and if you build it like me, the cost will be so negligible that it won't be of any significance.

-- Chuck Knight

P.S. BTW: Sorry I haven't emailed you lately. I've got my boards back, and I've been drilling aluminum today, building a mount and adapting an enclosure. I'm going to try to finish wiring them tonight, and maybe my machine will FINALLY move tonight, or tomorrow. We'll see...

P.P.S. While waiting for my boards to come back, I couldn't just sit idly by, twiddling my thumbs. I started a conversion project, based around this. A $20 X-Y drill press vise. Should prove quite useful for higher precision work.

10-02-2003, 12:14 AM
That is some excellent advice from Chuck Knight.

I am reminded of similar advice from my telescope mirror-grinding days:

"Don't build your last telescope first. In the time it takes you to grind an 8 inch mirror to final form, you could have ground a 4", and a 6" and then the 8" mirror." :)

10-02-2003, 12:16 AM
A word of advice. Stick with readyrod 1/2" or smaller. The larger diameters don't work well..too heavy for the poor little stepper to spin (too much friction in the nut too) and you end up going slower than with the smaller rods. Trust me on this one. Been there.


10-02-2003, 12:28 AM
This is interesting........ I appreciate all the input on the mechanicals. I didn't realize the friction coefficient was so high on the ready-rod vs. the ball screw technology.

My experience in motion control comes from working here:


In particular look at the Model 5200 and 5250.

I was responsible for building up a z-stage controller from third party solutions providers and providing laser power control. Laser power control was achieved by a servo feedback mechanism in which measured laser power was the servo, and the plant was a polarizer plate controlled via stepper motors. The z-stage rode on a ballscrew which was spec'd by mechanical engineering.

I am just now learning the intricises of the mechanics behind the machine :D My knowledge of the electronics will suffice.... for now. I don't have the bucks to do the multi-frequency servo control which was employed on these machines. We were working to micron tolerances on these BTW.


10-02-2003, 01:10 AM
Assuming your steppers are like mine, the shaft diameter is 1/4". This is an important detail -- this style of machine has a direct coupling between the shaft and the lead screw.

The easiest way to make a coupler is to use a high pressure air hose, from Home Depot. It's the clear tubing with the spiral fibers in it...you'll know it the second you see it. Since you only need a few inches, sometimes they'll even give it to you as a "sample" piece!

You get the tubing with 1/4" ID, cut an inch or so, and just stick it onto the end of the shaft, and the end of the lead screw. Tighten down some worm clamps to hold it in place, and you've got a very reasonable, flexible connection...this helps, because with a rigid connection, alignment (within a few thousandths) matters.

I used 5/16" allthread, because I could tap a 1/4" ID tube, and thread it in -- made the connection even more secure. I don't require the worm clamp for the leadscrew...though I have one, just for the sake of overengineering. :-)

And, BTW, my machine has so little friction, with the skate bearing and iron pipe, that I can turn the lead screws with 2 fingers. It's effortless...and the 18tpi thread will make sure that the forces available are FAR in excess of anything I'll require for cutting.

-- Chuck Knight

10-02-2003, 01:12 AM
To get around the small power of your motors (100 oz-in) try using a motor on either side of the ball screw shaft. That way you have two motors on each shaft (at least X and Y.) There is nothing you need to put them in sync or anything as long as they are same models used on both sides. I have no experience with it, just read that somewhere, so I may be talking through my hat.

I notice in your pictures there are no motors mounted yet. Do you just use trained monkeys who turn the handles real fast on que? LOL.
Seriously, I'm all for planning these things out. I know too many people who just have a pile of parts and never finish their project because it wasn't well thought out. "The mind is the final weapon, all else is supplemental." - John Steinbeck.

10-02-2003, 01:29 AM
I've been working on the electronics, tonight. I finally got my boards, and have only 3 of them, so I'll be switching back and forth till I get another set...that's why I'm trying to take my time on the electronics. I want it to end up being done right, AND to end up being modular enough to swap the driver box between the routers, with relative ease.

The vise project is on hold, till the other router is completely finished -- hopefully, now that my boards are here, that'll be tomorrow. The image is just a "stock" image that I keep on my hard drive for these types of forums. It shows the vise at a fairly good angle, and most people can "see" the potential for the vise the moment they see the picture. Anyway, I have all the parts to do the vise project...but the other one has taken priority for the moment.

I have created some nylon sleeves that thread into the original 14mm ACME nuts on the vise, and which I've tapped to take something of higher quality...standard allthread. Chinese vise ACME thread is not of good quality, to put it mildly...hardware store allthread and a nylon nut exhibit less backlash! Since this will be a tiny (4x6") router, I'm using a 5/16" leadscrew, which should prove more than adequate. And besides, I had some left over! ;-)

Anyway, the inserts for the vise are finished, I'm experimenting with a slightly different split nut design that I think will reduce backlash even more, and the motor mounts are sitting on the other workbench, right now. One router at a time...

Trained monkeys...I like that! :-)

-- Chuck Knight

10-02-2003, 11:36 AM
Hey Eldon, I also worked for ESI back in the late 70's. My brother-in-law worked there in the maint. dept for many years as well.