Delta arms with traxxas joints

Hi,

I just received my delta arms from tridprinting.com with traxxas joints and I see that the sideways motion of the joint is only about 30 degrees which is only enough to reach the the outer limits of the frame but isn't able to go further. This limits my print surface from 50cm diamter to 30. Is there an effector that adds another degree of freedom the joints or am I limited until I buy magnets?

Cheers

Quote
sungod3k
Hi,

I just received my delta arms from tridprinting.com with traxxas joints and I see that the sideways motion of the joint is only about 30 degrees which is only enough to reach the the outer limits of the frame but isn't able to go further. This limits my print surface from 50cm diamter to 30. Is there an effector that adds another degree of freedom the joints or am I limited until I buy magnets?

Cheers

Check out the Cherry Pi3 system of steel balls in cups held under tension by Spectra line and springs Works really well.

The Cherry Pi thread is lower down the Forum section!

Doug

Im using parts of the cherry pi already, so I had this on my list but with this being my first delta I decided for the easier complete kit of arms, all glued and ready. This makes of course a good case for magnets for my next upgrade but for now im stuck with traxxas.

Quote
sungod3k
Im using parts of the cherry pi already, so I had this on my list but with this being my first delta I decided for the easier complete kit of arms, all glued and ready. This makes of course a good case for magnets for my next upgrade but for now im stuck with traxxas.

Yes but have you seen that the Cherry Pi doesn't use magnets any more?

Yeah but im not cutting of the traxxas joints I think I have to build a joint like on the original rostock.

Quote
sungod3k
I just received my delta arms from tridprinting.com with traxxas joints and I see that the sideways motion of the joint is only about 30 degrees which is only enough to reach the the outer limits of the frame but isn't able to go further. This limits my print surface from 50cm diamter to 30. Is there an effector that adds another degree of freedom the joints or am I limited until I buy magnets?

I don't understand the problem. The maximum radius of a circular bed on a delta printer is limited by the size of the effector and how far apart the towers are, because the effector needs to remain clear of the towers and associated belts. That being so, you don't need more than about 30 degrees of sideways movement at the Traxxas joints to reach any part of the bed, unless your ratio of diagonal rod length to delta radius is lower than is commonly used.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

This should clearify. With the traxxas joints I only reach the frame the dark green bed with the blue arms. But I had originally intended to use the light green bed with the red arms.

I understand your picture, but I still don't understand why you have a problem. At the position you have shown with the red arms, the carriage they are connected to should be is well above the effector, so the actual angle of the arms relative to the carriage and effector is less than 30 degrees - unless you have chosen unusually short arms.

My Mini Kossel with Traxxas joints reaches beyond the edges of the bed at the point midway between the towers with an angle of less than 30 degrees at the joints, as you can see here.



Edited 1 time(s). Last edit at 02/18/2015 12:33PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote

the carriage they are connected to should be is well above the effector, so the actual angle of the arms relative to the carriage and effector is less than 30 degrees

I dont think the vertical position changes the relative angle of the horizontal movement. Its limited to the same angle horizontally not matter where it is vertically.

In your foto is that the maximum position your arms can reach? If that is the case you could have a bigger bed if you had more reach right?

Quote
sungod3k
I dont think the vertical position changes the relative angle of the horizontal movement. Its limited to the same angle horizontally not matter where it is vertically.

The vertical position makes a big difference. Imagine an infinitely tall printer with infinitely long diagonal rods. There would be zero angle in the XY directions at all the Traxxas joints.

Quote
sungod3k
In your foto is that the maximum position your arms can reach? If that is the case you could have a bigger bed if you had more reach right?

The bed is as large as is can be without the effector hitting the drive belts When the effector is close to a tower, the rods to that carriage are almost vertical. The only advantage to having more than about 30 degrees movement at the joints that I can see would be that I could use a triangular bed with the vertices of the triangle between the towers - if such a bed were available.. As it is, the nozzle can reach several mm beyond the edge of the bed midway between the towers.

Have you assembled your printer yet, and found that you really have a problem? If so, what is the ratio of diagonal rod length to delta radius on your printer?

btw I'm not saying that other sorts of joints may not have advantages, just that Traxxas joints appear to me to be adequate (just) in respect of the amount of motion they allow.

Edited 2 time(s). Last edit at 02/18/2015 01:13PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

The print area of a delta is not a circle. It's just convenient to make the bed round. In your diagram, it's not possible for the arms of the lower left tower to reach the effector position shown.

Quote
etfrench
The print area of a delta is not a circle. It's just convenient to make the bed round.

Agreed.

Quote
etfrench
In your diagram, it's not possible for the arms of the lower left tower to reach the effector position shown.

I disagree - it's entirely possible, unless the arms are too short. Bear in mind that the diagram is a 2D projection of a 3D structure.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote

I disagree - it's entirely possible, unless the arms are too short. Bear in mind that the diagram is a 2D projection of a 3D structure.

youre right, its doesnt have to be, but if you take the rule of thumb that the arms are 3/4 of the triangle side, then the distace from one tower to the opposite triangle side is roughly the arm length. as here in my case. I have 50 cm triangle side and 40cm arms.
In the future Ill look into making the arms as long as the triangle side, that should enable me to use the space more effectively.

Quote
sungod3k
... but if you take the rule of thumb that the arms are 3/4 of the triangle side, then the distace from one tower to the opposite triangle side is roughly the arm length. as here in my case

I don't claim to be an expert in delta mechanics, however I'm certain that's not a good rule of thumb. The diagonal rod length should be related to delta radius, not the triangle side. It's not equivalent, because the size of the effector does not need to scale with the size of the printer. So larger printers generally have a proportionally larger delta radius, and therefore need proportionally longer rods for optimum geometry. The rods from a tower should not be horizontal when printing on the opposite side of the bed, they should be at an angle of at least 20 degrees.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Yes I also heard about the 20 degrees rule as well. I guess after taking all observations from this thread into account the the arms should at least be as long as the sides and be so long that the with a maximum of 20deg angle you can make use of the biggest circle inside your towers as possible.

I have another rule for diagonal arm length:
[forums.reprap.org]
And it approximately corresponds to your 20° rule.

I have build the arms now and I have an effective reach of 30cm.

On the other hand, setting your maximum angle to 30 to 40 degrees instead of 20 will allow you to run at higher speeds at the edges because the opposing carriage isn't trying so hard to convert vertical motion into horizontal. Of course that means using longer arms, which will allow a little more flex, which means using slightly heavier rods, which increases the load on the motors. YMMV.

Yes I agree, my next version will have at least arms as long as the triangle sides.

btw is the increased length of movement in the extreme position when the arms are near 20° the reason more lower precision that is always talked about?

actually, at an extreme angle, say 10 degrees and an arm length of 200 mm, just for clarity, then one stepper motor step results in only 18% movement laterally, so just based on that the precision should be at max near the edges.

But..other factors can come into play: trying to go too fast near the edges may require the stage to move faster than it can. Also, there might be some rounding errors in the firmware, which is an issue I keep running into, with auto bed-leveling. There might also be a magnification effect of tower misalignments and other errors when the arms at at the extreme angles, compared to the center of the bed.

Is the firmware accounting for the increased speed needed by slowing down proportional to the distance to the center?

Marlin slows down when arms are almost horizontal. That is because Marlin limits speed based on carriage speed and not effector speed.
Repetier does not slow down; it is controlling speed based on the effector movement and not carriage movement.

If you want to calibrate your printer precisely you do not want to have arms much longer than the diameter of the bed. That is because the calibration errors are most easy to see at the edges of the board when arms are almost horizontal. If the arms are too long you will have hard time to detect calibration errors.
But you can look at this problem also just the opposite way too: "I want to have the diagonal rods (arms) too long so that I can quickly make the printer working (level the bed), even with wrong calibration (i.e. I do not care the dimensions may not be right)."

retracted, found Herceks link to calibration errors.

Edited 1 time(s). Last edit at 02/27/2015 01:42PM by shadowphile.