threaded rod instead of belts

Has anyone built a delta printer using threaded rods to move up and down the arms? I know they are slower than belts but they could possibly be more precise. Moreover, moving the hotend from end to end only results in very little travel of the belts/rods in the z axis.

I am not sure if anyone has. I will line out the design issues below that make their use uncommon.

Imagine that the pitch on your screw is 2mm. (That is a very generous number. You can go bigger but it will start to get expensive.) Now, the smallest pulley I have is 36mm in circumference. That means that you are going to be 18 times slower using generous numbers. Some of these printers to push 1000mm/s. Switched over they now will top out at 56mm/s. If you aren't really diligent in your design then you will probably end up with a much slower top speeds. The positive is that you will be at least 18 time more precise. Unless you beef up the rigidity of the machine substantially most of this extra precision is going to be wasted. Additionally, even if the positioning precision is realized you still have the inconsistencies in the plastic filament which will lead to +/-4 microns error. (That is as good as it gets unless you can transduce the mass flow rate in your hot end with more precision.)

The other issue is backlash. This is minimized by gravity being in the same direction as each threaded rod. If you accelerate too fast the forces may overcome gravity. You would have to solve this by either using a low acceleration or anti-backlash nuts.

I felt bad about posting the reasons for not going with a threaded rod so I went as far as to design what I believe would be an acceptable threaded rod solutions. Check it out. [forums.reprap.org]

Convetional threaded rod delta:


The picture is from this page: [3d-delta-reprap-southafrica.blogspot.sk]

Thanks for the replies. I was thinking of getting something along the lines of 5tpi (approx. ~5mm). This would bring it up to at least 100mm/s. The reason why I thought this would be a good idea was because it's a happy medium between accuracy and speed. 1000mm/s is nice but that's only good for jogging. I don't think many people will actually be printing at speeds that high. I saw your lisa concept, all I asked was a quick yes or no answer but didn't realize how much I would get back. It is ok to be critical, no need to feel bad!

Where does the lack of precision come from in belt systems, and is it manageable?
Screw drive is precise but has it's own precision aspects --backlash --
and screws with usable lead for speed, can be real expensive.

What is the max rate travel for extrusion? Laying down plastic?
Are non-extrusion move speeds important?

There is also the limitation of stepper motors and stepper drivers.

If the motors were at the top --- the drive screws could be shorter -- not reaching down all the way to the bed?

The lack of precision from a belt is just from the fact that get more movement for a stepper step. What we mean by precision here is really just movement/step.

I have been able to print at around 250mm/s for short burst. However, I print at 60mm/s when I want it to be perfect every time.

Non-extrusion move speeds really don't matter a lot. A print speed of 100mm/s and a rapid move speed of 1000mm/s will only save you a few minutes per hour of printing. That and it puts more strain on the machine.

The screws could be shorter but you still need to be able to tension the screws. The price I found for the screws I want is $10/ft.

@vinhdinh: I thought you would like to see what exists because of your question.

[youtu.be]

It is actually printing too but you will have to wait on those videos.

this is amazing! and you told me it wasn't worth it! hows the print quality? I am excited to see the next videos

nicholas, I realized I have some rods that are similar to what you are using. I got them from some old ibm tape drives. only problem is that they are 3/4" thick, I assume I'll need to beef up the motors for sure if I were to attempt this. on the flip side, I think these rods will be more than stiff enough to keep it rigid. and they come with delrin nuts

I'm not at all sure you would need bigger motors. A lot would depend on how much friction the follower nuts on the screws turned out to have. The thread pitch would also play a big part in how well they work.