[devers6]

(Moderator - This may not be the appropriate forum for this post; if not, I apologize. I wanted this to get high visibility, and this is an active forum. If it is better suited to some other location, please feel free to move it)


I've got an idea I just can't seem to get out of my head, so I thought I would outline it here to see how it plays.

Basically, I've been frustrated by the lack of DIY-friendly linear motion components to make it easy to assemble a home-precision level CNC machine. Sure, you can haunt Ebay for months and pick up THK slides and such (I built my 24" x 48" router that way), but what happens is that you end up designing your machine around the parts that you can find, rather than designing a machine to your needs and then buying the parts.

Aluminum extrusions like 80/20 are great because they give you lots of mounting options, and I found the cost to be acceptable for the basic framework (I used 1030 - 1" x 3" - for my frame, and it was a little over $.50 an inch, and now you can get seconds on Ebay for even less). But one thing that always bothered me was that there didn't seem to be a general-purpose extrusion that lent itself to the construction of a linear way system. You see modular linear motion systems that include ways, carrier, motor mounts, motors, ball screws and everything that seem like just the thing to base a machine around, until you find they cost anywhere from $15 to $30 an inch.

Given the popularity of sites catering to the home CNC hobby, and particularly the success here at CNCZone, it would seem that there is a burgeoning market for reasonably priced components that could be used to form linear way systems in home machines. Sort of an "Erector Set for CNC"; or more practically an online "CNC Supermarket". What is needed is a kind of consensus for just what kind of parts are needed, and some way to produce them economically.

What I propose is kind of an "open source" extrusion profile design (or several); where the different consumers represented here at CNCZone could hash out the parameters of what would be appropriate for the various configurations that home and light business users would need. These design ideas could be vetted by the experts that frequent these forums, to keep out "rookie mistakes" and to refine the designs into something that could be economically manufactured by the many extruder facilities out there. No one would be allowed to "own" these designs, to patent them or restrict them in any way. However, like the various commercial distributions of Linux, those with the wherewithall and the desire could take it upon themselves to have one or more of these profiles manufactured; with the expectation that there would be at least some kind of willing market. Perhaps the polling facilities here at CNCZone could be used to gauge buying interest, to prod some entrepreneur into taking the leap into having a run made and then offering them in the kind of lengths that would be friendly to the DIY group.

When I first thought of this, I was anticipating tooling costs in the tens of thousands of dollars, and minimum first-run quantities of thousands of feet. But that wasn't enough to kill the idea - I thought if the right design could be created, the costs might just work out. I was pleasantly surprised to find, on further research, that extruding startup costs are far less than I thought. It seems that uncomplicated dies in small circle sizes (less than 6") can be cut for around $1000 or so, and even more complex profiles like hollow shapes will still fall probably under $3K. It looks like a minimum run is something like 500 pounds, which would probably equate to around 500 - 1000 lineal feet, and I'm assuming the recurring costs would be primarily driven by the cost of aluminum; maybe $2.00 - $3.00 a pound. So for a shape that worked out to about 1 lb/ft, you might have a final cost of perhaps between $3 - 10$ per lineal foot for an extrusion that exactly fits your application. I'd pay that kind of money for an extrusion that met my needs. In fact, I might even spring for the first manufacturing run and see if I could make a few bucks.

So, tell me what you think. Here are a few ideas I have for custom extrusions that I would like to see:

- An extrusion in the form of a Nema-23 and Nema-34 motor, with mounting holes sized ready for tapping or to drill out for a through fastener. A 2" length of one of these would make a dandy motor mount. You could hog out one side to allow access to a shaft coupler or bearing.

- An upside-down T-section standoff profile for mounting the shaft for the open-type linear bearings (priced those lately? $$$)

- A linear way profile that would mate with a complementary carriage extrusion, to build a sliding system that enclosed the lead or ball screw and could be available in exactly the length your machine requires.

- Maybe even a carriage profile with T-slots to hold skate bearings at the appropriate angle for the gas pipe guys.


I'm sure you all have ideas as well. I'd really be interested in a discussion on this. Maybe I have the economics all wrong, but undoubtedly there is someone out there who knows and can set me straight.

Thanks,

Dave

[devers6]

As a follow-up, here is the kind of thing I am talking about. Obviously, this is a proprietary design; but it demonstrates the kind of things that are possible with extrusions. Think about the kind of profiles you wished they stocked in 4 foot and 8 foot lengths in a rack at Home Depot, draw them up and post them here. Maybe wishing can make it so.

[devers6]

Here are a couple of profiles for Nema 23 and 34 motors

[Hack]

Perhaps we should be looking at it a different way. Perhaps we should try to design a new bearing to fit existing extrusions? Just a thought.

Dan

[devers6]

It's funny - any time I see a novel extrusion these days, I try to think of a way to turn it into a linear slide! John Kleinbauer (a true hero to the DIY-CNC crowd) has proposed a number of designs using 'storefront aluminum'. I haven't seen an example of the profile, so I can't get as excited about it. The problem always seemed that existing profiles were "almost there" - they would work great if it was a little thicker here, or if it didn't have that projection there. That's why I thought why not design the ultimate slide profile and then figure a way to get it made at an affordable price.

But you are right, the most expensive part is the longest one. If we could find a good existing profile for the length of the ways, and then design a carrier for it (maybe even an extrusion, but it would be much cheaper if you only had to buy a foot or so of it). It seems that a lot of people here have made the gas pipe and skate bearing concept work for their machines. Maybe a great starting point would be a design for a carriage system that would make the mounting and alignment of the skate bearings a no-brainer. I'm sure it would be popular!

[CJL5585]

How about some extrusions that would allow one to build a machine out of aluminum plate. I am thinking corner brackets, etc. that would require one to only drill holes thru the plate on a drill press.

Just an idea. That is why my machine is aluminum and wood instead of being all metal.

Just a thought.

Jerry

[pen25]

great idea but as was asked by hack why not design a bearing or wheel? spring loaded use three sides of the extrusion or at least one linear teflon runner. come to think of it i wonder if fag has something that would fit it.


interesting info here..
http://www.fag.com/content/en/produc...ck_rollers.jsp

[carlnpa]

I'm with Hack and Pen on this one, develop a bearing to work with existing 8020 profiles. I made some out of uhmw and they work well here starting at post #23 http://www.cnczone.com/forums/showth...5&page=3&pp=10. Maybe develop a plastic extrusion or bearing carrier to mate with the 8020 profiles.

[pen25]

i wonder how well the 8020 rollers work. i still think the bearings i scored off the garage sale make your own at 3.50 a pop will work fine for what i have planned. the slide material was even cheap just have to cut and drill it. carl thx for the link to your thread. looks like your machine is really cleaned up. need to load some pics of parts you cut with it.

[Hack]

Below is a quick cad sketch of an extrusion idea that could use the V Roller bearings on an eccentric shaft.

Sorry for the lack of detail.

[Madclicker]

This will wear the extrusion very rapidly. When one side of the bearing is rolling the other is grinding.

[Hack]

It would have to consist of a minimum of 3 bearings offset from center a bit. Obviously one bearing as you mentioned would drag on a side as the other rolled. I had actually shown (although very poorly) 2 bearing in the picture above, but it would take a minimum of 3 to keep alignment.

Dan

[carlnpa]

Hack made a machine using the 8020 uhmw system for bearings. I am cutting plastic parts in UHMW, Nylatron GS/GSM amd other moly filled nylons. These plastics are very tough, long wearing and low friction. I have alot of respect for them, they are intended to handle much larger loads and speeds than these cnc machines will ever put on them. I'll make up some bearings for a y and z axis in UHMW for a machine like Hack made in the next few days and post results here. I have also been playing with UHMW holding skate bearings, but I'm still not convinced the added complexity is worth the small reduction in friction.
Pen - attached are pictures of some of the parts I make on these machines. Both machines will cut these parts, the K2 is just faster and more accurate.

[spalm]

I don’t really like the idea of aluminum rails unless something softer that aluminum is riding on it (like plastic). Looks like the slides in post #1 have Teflon strips installed. So maybe try to extrude a bearing holder to use cheap bearings to ride on angle iron or something else that is strong (i.e. iron) and cheap. I wish we could extrude iron.

Steve

[PEU]

I always remember this basic rule: The harder material will grind the softer one.

As Steve said, plastic bearings may be able to travel along an aluminium extrusion without problems, but these should be easilly interchangeable for the same rule above.

An idea is to use instead of a 40x40 extrusion, use a 80x40 one, so you have 2 parallel slots.

This way you have your 3 points: 2 are the slots, and the 3rd is the mid section which can add tension to the triangle setup.

My 2 cents (of peso)


Pablo

[AndyL]

I've been looking through the 8020 catalogue all evening (pablo - you're a bad influence!). Had a question...

Is there any reason one couldn't use the 'off the shelf' linear bearings from the 8020 system in place of the DIY'ers rollerblade/pipe system?

Thinking it would be pretty easy to build a gantry router completely out of the 8020 system... But the question becomes, would those bearing suffice?

Andy

[joecnc2006]

I've been looking through the 8020 catalogue all evening (pablo - you're a bad influence!). Had a question...

Is there any reason one couldn't use the 'off the shelf' linear bearings from the 8020 system in place of the DIY'ers rollerblade/pipe system?

Thinking it would be pretty easy to build a gantry router completely out of the 8020 system... But the question becomes, would those bearing suffice?

Andy

Many have done it, just do a search you should find examples of them using it.

[mdreitzusa]

here is a pic of the rollers that i made.they are two 1 5/8" disks cut from an old cutting board.then pocketed with a 7/8" forstner bit for the roller blade bearing.and though hole is 1/2".they ride on square aluminum tubing.the v groove was made with a 90 deg. sracper.