Friday, October 27, 2006

 

More on filament and threaded rods...

I took off a few days this week and drove down the coast to visit a few potential outfits that could potentially make parts and supplies for reprap.

Noll, Inc. down in San Luis Obispo offers a good range of Acme threaded rods and associated parts like thrust collars and bushings in a size range that would suit our reprap developments. The advantage of using acme rods are that they are accurate in the micron range, unlike the studding that we have been using.

The Acme thread rods are done in stainless steel and run in the range of $10-15/foot in the diameters and thread pitch ranges that we could use. They also offer thrust collars in anti-backlash designs for about $30. This stuff isn't cheap. Overall, however, it might be more cost-effective that studding in that we are having to do all sorts of things to force studding to do what we want that we could avoid having to do if we just went straight to Acme threaded rod.

Here are some details that I didn't know about before, though.
I also visited another potential supplier of filament. I'm not going to mention their name yet because I haven't got firm prices out of them yet. They don't want to quote till they have some data sheets for CAPA. I'm planning on getting those out to them in a few days. I can tell you that they were talking about processing costs in the range of $1.10/lb, though.

I feel that it will be a good idea to develop relationships with several suppliers of filament rather than dealing with one exclusively so that we will have some flexibility on delivery schedules and quantities as the demand for filament picks up in the next year or so.

Comments:
I'm actually very curious as to whether it will turn out to be practical to use computer control to correct for errors in the axes of the machine... it could be that the level of precision in the acme rods is unneeded in the final run-down. They're certainly wanted in manual tools, and they make CNC machines easier to code for (and CNC machines are already so expensive with everything else that the good drive rods don't make it noticeably worse), but for something like this, software compensation may be the better solution. Perhaps taking an optical sensor, mounting that next to the head, and using a calibrated pair of lines to take measurements of the distance between the lines according to the axes, and then fitting a spline to the results and using it to correct in software... Could be a noticeable edge in terms of production cost over the more standard machines.
 
Right now, I'm on the same page with you, viz, not using acme threaded rods given what I know now after my visit to Noll. I've had a lot of experience using the standard sort of studding that you can buy at any good hardware store, though, and I must say that the experience is not completely positive by any stretch of the imagination. I thought that it would be nice if I could avoid having to make that complicated sliding joint that I had to make to take up for having studding that wasn't quite straight. Acme threaded rods from Noll would have solved that problem.

That said, I'm not totally sold on any particular approach. I'm just trying to make sure that I know enough about the plusses and minuses of the different approaches that I can make informed decisions about what I want to do. :-)
 
i'm sure you can find a much better deal than that.

personally, i'm looking into ordering some rods from Enco. They have plain steel rods + nuts for about $40-50 with shipping.

Granted, they are probably not micron level accurate but they will be a helluva lot better than your generic threaded rod from the hardware store.

One good advantage Acme threaded rods have I read in the cnczone forums and is related wear and tear. The rods are designed to be used back and forth many times whereas your average threaded rod is not... its designed to hold and have high friction. This leads to having them wear out much sooner and thus you'll have to replace them. :-/
 
***One good advantage Acme threaded rods have I read in the cnczone forums and is related wear and tear.***

I don't know about everyone else, but I see the threaded rods as more of a 'repstrap' axis than a 'reprap' axis... and plan to move to a rack&pinion or gear based system once I have the capabilities...

I'm not looking for great accuracy with my repstrap, i'm not even looking for 1mm (but i'll take it!). I'm just looking for something that'll get me to the next level, and if I need to start with a 6" gear for my rack and pinion to be accurate enough, I'll deal with it and use it to make a 3" gear rack&pinion with the same/better accuracy!

All I'm looking for out of my threaded rods is for them to work and to reliably produce something... even if they've got 'bad spots' or if there's a spot on the worksurface that is inherintly flawed, it's an evolutionary step.

If 'off the shelf' threaded rods aren't good enough to get the initial repstraps working, then we've failed to produce a system that can refine itself and make more accurate copies... The reprap designs that will continue to rely on threaded rods may need to go with quality rods for the accuracy, but personaly, I don't want to put that much reliance on external components. I want to control the accuracy of my repraps by the parts that I'm making...
 
I think that you can account for backlash in the drive nut in the software. I'm getting past a lot of problems with that sort of thing in gearboxes by putting the encoder directly on the studding rod.
 
***I don't know about everyone else, but I see the threaded rods as more of a 'repstrap' axis than a 'reprap' axis... and plan to move to a rack&pinion or gear based system once I have the capabilities...

I'm not looking for great accuracy with my repstrap, i'm not even looking for 1mm (but i'll take it!). I'm just looking for something that'll get me to the next level, and if I need to start with a 6" gear for my rack and pinion to be accurate enough, I'll deal with it and use it to make a 3" gear rack&pinion with the same/better accuracy!***

I've been thinking along the same lines for quite some time. The threaded rods in Tommelise are repstrap stuff as far as I'm concerned. I may change my mind about that later, but for now I'm still focussed on evolving a relicator that has the absolute minimum of material that can't be fabricated by the replicator itself.

About the only concession to outside manufacture in my mind is the use of tempered float glass as a working surface. It is dirt cheap, for practical purposes optically flat and as near to a universally available material as can be had on this planet.
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]