[Homl]

Where can one find inexpensive turn-key linear slide assemblies ready for control motors? Maybe in foot increment lengths. Something that can be easily bolt together for a router table.

Originally I was planing to build my own. By the time I added up all the cost of the linear bearings, lead screws and other component, man was it expensive. On top of that, I have to put in lots of hours designing and building the slide assembly.

I found some fully assembled slides by Techno and a few others, but they are way too expensive.

Inexpensive is what I am looking for. Something reasonably accurate, lightweight and rigid, nothing fancy.

Thanks for the help!

[Sanghera]

What about some inexpensive bushings?

[vacpress]

ebay - theres lots of stuff like that.. gonna be a long wait to get matching componenets tho

[buscht]

http://cgi.ebay.com/ws/eBayISAPI.dll...e=STRK:MEBI:IT

[Homl]

Thanks, but they are too short. For the Z axis, one of the assemblies would work fine. I am looking for something 48" or larger. Are there inexpensive supplier for linear asseblies where you can purchase off the shelf, rather than to hope to get lucky on ebay?

Thanks again.

[DDM]

I don't know how good they are but 8020 and Tslots have a linear bushing setup for their extrued aluminum. I know it's not THK but when I talked to the guy at 8020 it seems that you could probably keep things within 5 or so thou. If you could figure out some kind of pre load to get the slop out of there you would be set. Assuming that you don't get anything too abrasive on the aluminum the anodizing shouldn't wear that much and most of the wear would be on the plastic(not sure what type off hand they use, but you could make your own from modernplastics.com) If that doesn't do it for you there's always gas pipe and roller skate bearings.

[Homl]

Yeah, something close to that kind of simplicity but with a little more precision. Maybe something in the line of a simple aluminum extrusion with a couple of basic guide rods with a decent ACME screw. A thick delrin plate that act as the bearings, the screw nut and the mounting platform.

[DDM]

Yeah other than guide rods could run some money if you went with a precision ground hardened shaft. I know you can get the rod for hydraulic cylinders for a decent price and with using like a uhmw, bronze impregnated teflon or a oiled broze bushing. You could get real trick with the oiled bronze bushing by adding an oil pump to your setup so that the rod and bushing is lubricated all the time and the only thing you would have to check is the oil in the resiviour. Any one else have any comments?

Carl

[pminmo]

McMasterCarr www.mcmaster.com 6061K75 1"OD 48" $58 Look at page 952 of their online catalog.

For 36" and under www.use-enco.com drill rod on sale is pretty good. Just according to how precise you want to be.

[chuckknigh]

Drill rod and even gas pipe work amazingly well, but there's another "off the shelf" option that you may be overlooking. Window frame extrusions. Add some plastic slides, and they should work quite well.

I found a glass supplier in my area, and went into their shop. They have so much of the stuff, even as "offcuts," that it's amazing. That, and some plastic slides, fully supported of course, and they'd make GREAT linear slides.

There is one extrusion I have my eyes on...nearly 4x6" in cross section, and reminds me of a capital letter "E" made out of 8020. I don't know how it works for glass, but it'd be perfect for slides. Stiff, strong, and it would have a minimal contact area.

Thank you, John Kleinbauer, for the original idea.

-- Chuck Knight

[DDM]

When talking about using extruded aluminum products for linear slides having a good cross section is good but when everything is build rock solid it is the straigtness factor that would concern me. You are looking at a max of .012" of straigtness. Sure I could put it on the mill and a dial indicator to see what kind of straigtness I have and then put it in a press and see what I can do to make it as straigt as possilbe. Also If you have 4 different rails that you are using as linear guides they should more or less cancel themselves out if you have a nice tight system which would involve some kind of preload system. From there what kind of wear are you going to encouter. 8020 and Tslots I think both use uhmw for their bushings. If you were to make your own and make them say 10" long you would have greater amount of cancelation of the different straightnesses. But over time you're going to get wear. That's the thing I haven't figured out yet. I don't feel like replacing bushings after say 300 hours of run time because the slop has got too great. If you you could do some kind of preload on them so that over time as it wears it keeps the same characteristics as when it was freshly installed. Maybe some kind of spring setup or air pressure to push the bushings into the extruded aluminum. Ok well I'll shut up now. I've got all of these ideas in my head and not enough time on solidworks to work all of them out.

Carl

[chuckknigh]

Let me ask you this, Carl. You're talking about a deviance from straight, of barely over 1/100th of an inch. That's better than most people can do on hand tools, and wood can move more than that with a change in humidity.

It's wonderful to plan for super precision, but unless you intend to use something like a LASER inteferometer for measuring deviation, a few thousandths is a realistic goal. .01 is just 10 thousandths, which while not perfect, is still pretty darned good. Holding accuracy of a few thousandths, over a span of 4 feet or more, is EXTREMELY difficult, especially for a first time builder.

Throwing money at it can help (i.e. high grade linear components) but you asked about inexpensive options. BTW, UHMW is surprisingly tough and long lived. And, thread wander in anything short of a precision ground screw can throw off your measurements, too...it's not just the straightness of the linear slides.

I'm not saying this level of precision can't be done...just that trying for this level of precision, for a first time homebuilt tool, is probably overkill.

What is your intended use? You've got me curious...

-- Chuck Knight

P.S. Have you considered a parallel kinematic system, rather than a simple X-Y-Z table? More complicated, but errors are not cumulative, like they are on a more traditional table.

[DDM]

Ok Chuck you caught me on one of my rants. If I could get plus or minus .010 on a moving gantry router I'd be more than happy. But my thing is what If I could get more and abuse the hell out of my machine but at the same time be using cheapo linear bushings instead of THKs or something simular. This whole aricle deals with a moving gantry router but can be applied to other styles.

I know aluminum and especially aluminum extrusions arn't the best for building frames and putting your linear motion on. I mean if you wanted to do it right you'd have cast iron and all percision ground components and with that you could probably get that extream percision but you probably wouldn't be touching wood with it.

Ok now that we've looked at the extream end of percision we can step back to the MDF machines with gas pipe and rollerskate bearings. Would you want to base a small startup business on one of these setups? probably not, it's a great learning tool and for those of you just cutting balsa wood for planes and other little trinkets.

Step that up a notch to aluminum framed machines that are using the high priced bearings and ballscrews. Ok this is getting to the point of having a production machine. This is where you run into the issues of money vs. percision. Typically on one of these machines you have two rails and four bearing blocks per axis, maybe except the z. When you want to span 8 feet that get's pretty pricy and this is what this entire thread is about, keeping the cost down while keeping the accuracy up.

My best guess is that you can get the accuracy that you get with THKs and percision ground ballscrews but with using a cheepo linear bushing system, acme rod and keeping it tight. Starting with the leadscrew setup, what is the major problem? Most likely Backlash, Ok, this can be solved by using a preloaded nut, what is the capability of instead of using two nuts and a spring using 3 nuts and two springs. I don't know if this would work but what if you could get say 2 thou more percision. Ok it sounds goofy but what if the machine that did the ACME rod had a thou or two off in several places, by increasing your area of nut coveraged you've elimnated that or at least decreased it. Now taking a look at the linear bearings/bushings. You wouldn't want just one cheap bearing to take a lot of load and still stay accurate, you'd have to be cutting extreamly slow. The best is to use as many rails and bearings spaced as far as possable to keep the table or gantry from torquing one way or the other. This is true for any type of machine whatever it is using for linear slides. So if you are using THKs you would probably want to use two pillow blocks per rail and have two rails per axis. Right? OK now lets get dirty. Take the system that 8020 has for their linear motion and mulitply that by 4. Insted of just using two of your rails for linear bearings, use 4. Why? Because you're gantry is probably going to be more sturdy than your frame now if everything is setup right it will have less of a chance of torquing one way or the other. My other idea was to set a preload between each one of the rails in kind of an x fashon if you can set the right amount of preload between all of the rails with minimal deflection you could possilby get the same results as with THKs at a fraction of the cost. And with this setup there would be no reason to replace your UHMW bushings until they were completely gone because it would always have a constant pressure on them in one direction.

OK it looks like I went of on more of a rant again, sorry Chuck, If I get time on solid works in the next couple of days I'll try and post what I'm talking about so it doesn't seem like I'm such a blubbering idiot.

Carl

[DDM]

Sorry chuck I was going off and I forgot to answer your question. What I'm looking at is starting up my own part time business that would utalize a wood router for just about anything decorative you can think of and possibly adding a 4th axis for pillar work (opps I shouldn't have said that, now I can't take over north Idaho). I want speed, because time is money.

I'm not sure what a parallel kinematic system is but I have a feeling it's along the same lines as what I'm kinda talking about. If you could explain that would be helpful and maybe I might be able to play off of it. I'm looking at going overkill but not over budget and if I can give anyone any good ideas for their own machine I'd be tickled.

Carl

[chuckknigh]

OK, let me address this in 2 parts.

0) Backlash is not the only enemy of precision. If you rmachine is out of square, your precision will suffer. If the thread accuracy wanders (I think it's done on a 1 foot repeating pattern) then your accuracy will suffer. If the heater vent hits one side of your metal machine, and not the other, your accuracy can suffer. Seriously...accuracy vs repeatability is something you might consider addressing.

1) If I was basing a business off of this machine, I'd not use MDF construction, etc. I'd go for (ideally) cast iron, hand scraped dovetail ways...they've got the mass to keep things rigid and precise, and have proven themselves in machine tools for centuries. Of course, they'd cost a fortune, and have numerous tradeoffs involved, not the least of which is rust, in this humid air.

Realistically, I'd just spend the money for some proper linear ways. Barring that, I'd probably look for a way to do a system with drill rod, but fully supported. Trust me on this one, 4' long rails WILL sag unless supported.

I remember seeing one thread, about a man with a small business similar to what you have in mind. He made his own UHMW bushings and put them on a simple framework with either drill rod, or polished gas pipe -- he intended to replace them when funds allowed, but they worked so well that they're still in service years later. It was about a month ago, I think, in the Homebuilt Woodworking Routers section. I'll see if I can find a link.

http://www.cnczone.com/forums/showth...1121#post31121

2) A parallel kinematic system is, in at least 1 form, a hexabot. It's a deformable 6 legged platform which allows 5 degrees of freedom, and 6 in at least 1 variation called a rotobot.

It has a huge footprint for a given volume, but it's among the most flexible machine tools available. Also, like I said, an error in the X axis won't be multiplied through to the Z axis...all 6 legs run independently, and errors tend to cancel out.

These are pictures of some:




Typically in these, the workpiece is moved and the cutter head stays still...but it can work either way. Yes, there are homebuilt examples, but they're MUCH more difficult to design, apparently.

-- Chuck Knight

[DDM]

You the man Chuck, You hit the nail on the head with the basic durability of the machine. I guess a lot of my ideas will work but it's the design of the machine itself and how ridgid and square that makes the difference in the long run. Repetability is a big factor in a lot of this stuff.

I like the whole hexapod design but it's not for me at least not now. I'll have to look through that thread and maybe get some ideas. One quick question though, that would take 6 motors to run that thing and how would you deal with trying to code for a program? Oh well I need to go to bed or else I'm going to be more worthless than normal at work tomorrow.

Carl