nice rostock with leadscrew drive

[cambo3d]

I've been wanting to convert to leadscrews and then this showed up in my email. I like the leadscrew design and the direct drive extruder, not to mention being all aluminum! trinity labs makes really nice printers this one should be no exception.

https://groups.google.com/forum/?fromgr ... Ce754W_DO8

http://www.youtube.com/watch?v=0p6PEwD7HMg

[Polygonhell]

Still not sold on lead screws for printers, they really limit the maximum speed unless they have a very steep pitch.

[cambo3d]

Still not sold on lead screws for printers, they really limit the maximum speed unless they have a very steep pitch.

while fast speed is good, i also prefer leadscrew for better accuracy. I believe he is using 10start lead screws, which do have a steep pitch to them.

[Polygonhell]

I think the accuracy advantage is overstated, there is only so much you can do with squirting plastic out of a hole. Most printers have many other issues that swamp the belts contribution to error.
Plus you have to have some mechanism to take the backlash out of whatever nuts you area using on the screws.

To be fair I can see some value in a rostock design because of the relatively long belts.

[cambo3d]

Most printers have many other issues that swamp the belts contribution to error.

exactly. these belts depending on which type have some stretch to them and wear out prematurely. I've used both types of drive systems and prefer lead screws over belts.

[Polygonhell]

My mills obviously use leadscrews (ballscrews) and I wouldn't use anything else for that purpose, but mills have substantial side loads, and printers don't.
Coming from CNC first to printers I thought much as you do, but speed actually matters to a printer, for a number of reasons.
You have to assume that the maximum speed you can run a small stepper at reliably is about 1000RPM, and at the voltages we generally run at that speed the torque at those speeds is negligible. A 1mm pitch leadscrews is then limited to about 16mm/s, a multistart screw might get you to 10 x of that, but there is a secondary issue.
The maximum reliable rate most firmware can run an Interrupt is at about 20KHz, so to hit that 1000rpm, you can only run abut 4x uStepping.

It's doable, but if you look at the people who start with screws for builds on the reprap forum, you'll see many of them revert to belts.
Even the very expensive FDM machines use belts, it's a perfectly reasonable solution for printers.

As I say the issue with belts on the rostock design is the length of the belt, but they could be considerably shorter if a different motion system were used, the lowest point of the carriage motion is well above the base.

[geneb]

Essentially it boils down to the fact that it doesn't make any sense to measure with a micrometer if you're going to mark with chalk.

g.

[cambo3d]

I was talking about accuracy in terms of linear movement, not just for 3d printers. Having an all in one cnc machine that could do both milling and 3d printing would be great. With belt system on a rostock you wont ever get there. The chalk thing is another issue. If your machine is accurate enough for milling then adding an attachment for 3d printing should be easy. Might not print as fast as a belt but lets get real here, I printed a 200 micron print yesterday that took nearly 16hrs at 30mm per sec. The quality was great but speed not so much. Any cnc machine can do that speed.

[Polygonhell]

I was talking about accuracy in terms of linear movement, not just for 3d printers. Having an all in one cnc machine that could do both milling and 3d printing would be great. With belt system on a rostock you wont ever get there. The chalk thing is another issue. If your machine is accurate enough for milling then adding an attachment for 3d printing should be easy. Might not print as fast as a belt but lets get real here, I printed a 200 micron print yesterday that took nearly 16hrs at 30mm per sec. The quality was great but speed not so much. Any cnc machine can do that speed.

Sure but what were your travel speeds and will the mill match the acceleration?
And FWIW yes they matter, try setting your acceleration to about 100 and travel speeds to 25 mm and try running a print with a number of travel moves.
My mill tops out at about 60IPM roughly 25mm/s, but the acceleration is MUCH slower than the printer, it's moving a lot more weight around.
Qubd has a hybrid device, but I believe they swap to belts when printing, unlike there original design.
The reprap forum has a lot of stories of people who started with converted mills, most find their subsequent dedicated printers produce better prints.

The concern I'd have with the machine in the link isn't actually as a printer, the use of prebuilt linear motion systems will probably result in good quality, looking up the specs they should hit ~100mm/s, but I have to question how the balljoints will hold up to any sort of milling.

[cambo3d]

I was talking about accuracy in terms of linear movement, not just for 3d printers. Having an all in one cnc machine that could do both milling and 3d printing would be great. With belt system on a rostock you wont ever get there. The chalk thing is another issue. If your machine is accurate enough for milling then adding an attachment for 3d printing should be easy. Might not print as fast as a belt but lets get real here, I printed a 200 micron print yesterday that took nearly 16hrs at 30mm per sec. The quality was great but speed not so much. Any cnc machine can do that speed.

Sure but what were your travel speeds and will the mill match the acceleration?
And FWIW yes they matter, try setting your acceleration to about 100 and travel speeds to 25 mm and try running a print with a number of travel moves.
My mill tops out at about 60IPM roughly 25mm/s, but the acceleration is MUCH slower than the printer, it's moving a lot more weight around.
Qubd has a hybrid device, but I believe they swap to belts when printing, unlike there original design.
The reprap forum has a lot of stories of people who started with converted mills, most find their subsequent dedicated printers produce better prints.

The concern I'd have with the machine in the link isn't actually as a printer, the use of prebuilt linear motion systems will probably result in good quality, looking up the specs they should hit ~100mm/s, but I have to question how the balljoints will hold up to any sort of milling.

not saying there isn't any other design considerations, but it is one step closer.