I have a question, the answer to which is likely obvious to most, but I’d rather invite commentary before I go all in: I plan to convert an i3 I have here to an all leadscrew platform (i.e. for X and Y since Z already is using T8 leadscrews). I understand that there is an intrinsic reason to keep the extruder light, especially with a system that has belts since heavier weights require more force to stop the carriage and change direction, and particularly since belts are slightly elastic. However, if you have a leadscrew in X, is the weight less of an issue (it’s an issue anyway, but is the effect less than what you would see with a belt)? I am trying to decide if it’s worth installing bowden or direct drive extruder. I prefer direct drives just so I can handle flexibles, and I suppose then print speeds will suffer. Thanks in advance.
Leadscrews are terrible for our application – belts are not elastic, generally. They have steel/kevlar cables in them and are meant for positioning and power transmission. The only reason you should use lead screws generally is for milling or some other thing that requires additional force to be placed upon the machine. Even then, generally the machine is not built to handle those stresses mechanically. Leadscrews slow everything down by quite a lot due to the motor needing to turn a LOT more for a specific distance to travel.
People belt drive their system all the way up to dual extruders with motors on the carriages…and they’re fine that way.
@ThantiK thank you. Is there an argument to drive one axis and not the other (e.g. Y not X)? I keep reading conflicting reports as to accuracy. Specifically for printing alone and not for milling, that is.
The only reason you might want to drive say…the Z axis with a leadscrew rather than a belt is the fact that you can turn the power to the motor off in between layer changes and that it doesn’t move all that often. The gearing that the leadscrew introduces allows the Z axis to stay in position due to motor detent torque.
There is no reason to drive X or Y with a leadscrew in 3D printing applications. Whoever is telling you this is probably either a machinist (who generally believe they know “3D Printing” because it has some shared base mechanics) - or someone who has looked into being a machinist.
There MIGHT be a reason to drive the X or Y if you had a gigantic moving mass…I’m talking like 5kg or so, but that’s a lot of weight.
@ThantiK thank you. I am curious to also understand the mods I see to some i3s on Thingiverse that do this. I can understand leadscrews in a Delta for example like the ScrewDelta. This is all for my ongoing education
@Samer_Najia – lead screws on a delta are the dumbest idea ever. I think it was done simply to prove it could be…it’s not a good idea just because someone did it.
I agree. Leadscrews might make sense for a machine that is doing both 3d printing AND milling… I’ve thought about a combo machine with my Big-E… but I will try belts first since milling plastic isn’t that hard. The combo machine would produce fast, thick, ugly layers, then mill the walls smooth… sounds like a software nightmare though. Like I have time to try crazy stuff like this 
Brook
If a person does use lead screws, they need to make sure that they use anti-backlash nuts and a fast motor that they can control the output position of well enough. The closed loop brushless DC motor projects on kick starter right now are the best chances for that.
If a person uses lead screws, they might want to have either a hand crank for the axis or a disengage option so they can reposition the hot end and remove the work or they will want to hit home when it is done printing.
A bent leads crew even with a companion smooth rod will introduce flaws into milling or printing projects.
The stepper motors usually used in 3D printing would result in slow prints.
@NathanielStenzel thanks. This is what I was looking at doing http://www.thingiverse.com/thing:1369041 as well as http://www.thingiverse.com/thing:561143. Could you comment on the viability of these things?
@Samer_Najia both designs look like they use normal stepper motors which would cause a speed problem. You would at minimum need to shop around for a stepper motor with a speed 4x the speed of most stepper motors to overcome the speed limitations of the screw pitch if you use a normal screw pitch. If you use a 45° screw pitch, you do not gain strength and may actually lose some due to friction. I might be overlooking it, but I do not see anti-backlash nuts in either. The point of anti-backlash nuts is to make sure the work tool does not wiggle any along the length of the lead screw.
Faster stepper motors (which will not need as many steps per revolution to get the results) and decent lead screws and anti-backlash nuts (usually comprise of two nuts with a gap forced between them) may add a bit of cost. Look in the CNC communities and you might get an idea of the prices.
What do you think you will gain by using lead screws? What do you plan on doing?
@NathanielStenzel I am trying to learn more about this stuff mostly. I was looking at experimenting with a printer I have here that itself is in need of a board upgrade. I found a version of the T8 nuts on Banggood that are anti-backlash and could implement those. I also have a set of higher torque motors I can use, but it sounds like I also need them to spin faster than nominal as well. In any case for the most part this is all educational…trying to get better at this.
The new parts that you would need will cost money and probably speed. It might be worth simply asking in another thread how i3 users manage to print flexible filaments and study their printers.
Higher torque motors of typical speed would help a belt drive get the power it needs to run faster with a weight to move. Higher speed (3x or better for typical pitch) motors with typical torque would be okay for a screw drive system. I suggest looking up the maximum speeds of direct drive systems and Bowden tube systems for flexible filament. I suspect that the ability to control the filament itself may be as much of a speed bottleneck as the motion control.
An alternative way to get more speed from prints is to try different slicers or different slicer settings or different part layouts. It is amazing how much difference in speed you can get that way.
If you do decide to try this out, fail or succeed, please share the results with the rest of us. Knowledge is power. The more knowledge we gain, the more we can advance the 3D printing field.
Who knows, maybe you to trying this and wanting more speed may lead to you using a closed loop brush less DC motor system and then the information gathered from your tinkering will get even more interesting.
Now that’s encouragement. I will order the backlash nuts. All the other parts are from my spares. I will try it with ordinary steppers and work my way from there.
Another thing I would recommend asking is how much speed you can get from making thick prints of flexible filament and if there is any problems with minimum cooling times (thus minimum layer times) required for them. I have no experience in flexible filaments myself.
Good luck on your experiment. I would not have the energy or resources to try such things myself. Well…not without blowing a load of cash and I am a cheapskate.
Oh if this gets expensive the whole printer gets scrapped. I got it as a gift 2 years ago and it long ago has been supplanted. I don’t mind spending $200 for this test, plus I am using this as a test bed for some other enhancements
Oh. Just for the record, I am not trying to encourage or discourage the project. I am simply trying to make sure everything is considered and that you have the knowledge or have an idea of what knowledge must be gained to take on the project. I am a firm believer in informed decisions. Hopefully all the information I provide is spot on.