Tuesday, May 26, 2009

 

Accelerating RepRap

Accelerating RepRap from Adrian Bowyer on Vimeo.

The experimental four-dimensional GCode interpreter is now five-dimensional. Four dimensions are so three weeks ago...

It has an extra DDA variable that is the feedrate. This means that if you send it:

G1 X85.3 Y75.8 E18.3 F2495.4
G1 X85.5 Y75.5 E18.7 F2575.9

For the second move it accelerates linearly from 2495.4 mm/minute to 2575.9 mm/minute as it moves from X85.3 Y75.8 E18.3 to X85.5 Y75.5 E18.7. That is to say, it treats the F values in exactly the same way as it treats X, Y, Z, and E. (This is not how GCodes conventionally work, of course, so you can turn this behaviour off by commenting out a line in the firmware that says

#define ACCELERATION_ON

)

Of course you can have the machine move at constant speed as well if you want. You do that simply by setting the feedrate first:

G1 F2575.9
G1 X85.5 Y75.5 E18.7

That will do a move at a constant 2575.9 mm/minute, as you would expect.

The latest Java host software in the repository now supports all this (again you can turn it all off by setting Accelerating false in the preferences).

There is one final firmware upgrade needed to finish this work off: buffering the moves in the micro-controller (as Zach and co. are doing with their code). At the moment there is a barely-perceptible pause between moves, which means that accelerating to high speed hits a momentary barrier at the end of the acceleration. As the whole point of all this is to make the machine move much more smoothly, that's got to go.

Accelerations allow the machine to work faster and more accurately at much lower motor currents, or maybe even with smaller cheaper motors. Or maybe even with motors we print ourselves, which Forrest is working towards...




Comments:
I'm assuming that your Darwin isn't nearly as noisy as your videoclip seems to indicate. Is that correct?
 
i think it has something to do with the frequincy of the noise mic's really pic it up well.
 
Yeah, I know the mike on my camera picks up and amplifies the softest sounds into something that sounds like a rock concert. :-s
 
It's a bit noisy, but the camera makes it sound like a tractor... The Darwin at the lab in the University is a lot quieter, for some reason.
 
Adrian

I would check what the darwin is stood on and whether it is acting as a sounding board ie material resonance.

Putting rubber feet on might help decouple the darwin from what it is stood on.

There again you might have these things covered already.

Cheers

aka47
 
Actually one of the things on my list to do* is to print some sprung feet for the machine to reduce sounding board effects.

*This list is now so long that I am forgetting some things on it faster than I am doing others :-)
 
I must admit to have been toying with the idea of designing a sort of flange with a cone on top that developed into a hexagonal receptacle for a steel M8 acorn nut.

The idea being to put a cheap rubber door stop (Hardware shop, B&Q, Screfix, etc) in the bottom to make screw on shock absorbing, noise suppressing, feet.

The more I think about it though using a long nut would allow you to have adjustable screw on feet that you could lock off with a standard nut.
 
Post a Comment

<< Home

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

Subscribe to
Posts [Atom]