Custom Delta Size/Arms Advice

Over a year ago i builded my first 3DPrinter (a prusa i3 boxed), the information from this forum were incredibly helpful , now i'm thinking about building a delta and again I find myself in need of some advice, hope you guys can clarify something for me!

I've all the information i need about the electronics, hotend, structure part, my "only" doubt is about the ratio between the base triangle side lenght, the printed area and the arms.
Is understandable that > is the triangle side > is the circle i can fit in it, obviously this means that i require a > arms lenght, what i'm not sure about is the correct ratio between these 3 dimensions..

Let's say i want to build a printer with a base triangle side of 35cm, this mean i can use a circe of ≈ 20cm of diameter staying inside the triangle, now, how should i calcolate the ideal lenght of the arms?
Right now the most useful information i could find were:
-this calculator (for the kossel mini), [docs.google.com]
-the notion that when in homing position the arms should rest at 30°(angle of base to arm).

If it's true that the rod should rest at 30° in homing position then all is solved, i can decide the print area i want, then calculate the arms lenght, otherwise i'm lost.


I also wonder another thing, is there a size limit over which the printer start to lose precision/speed due to excessive arms lenght (let's assume that the structure is solid)? for example if i want a 20x20 print area this means a 285mm radius circle and something like 40-41cm base triangle side, is this doable without losing precision/print speed or not?


Sry for the long post and the poor english, hope you guys can give me some info.

Edited 1 time(s). Last edit at 01/15/2015 03:41PM by Nova.

I think your calculation is mostly right, Im sourcing arms as well now and the rules I follow are "side lenthg * 80% = arm length" (the google doc I think uses the same rule) and "angle of base to arm in maximum position should be bigger than 20°" I will by my arms from trid printing and then have 4-5 sizes there you could check for which machines they sell arms and how the dimensions are there. I am aiming for a base length of 50cm with 40cm arm, and this fulfils my rules.

Quote
sungod3k
I think your calculation is mostly right, Im sourcing arms as well now and the rules I follow are "side lenthg * 80% = arm length" (the google doc I think uses the same rule) and "angle of base to arm in maximum position should be bigger than 20°" I will by my arms from trid printing and then have 4-5 sizes there you could check for which machines they sell arms and how the dimensions are there. I am aiming for a base length of 50cm with 40cm arm, and this fulfils my rules.

Right, the 4:5 ratio, i've read about this, it seems like a good general rule but i've read of people saying that it's more a Kossel specific rule than a general Delta rule... Still, since i plan on using the kossel mini effector i guess that it could apply.
As for the arms angle, sry if i was not clear, when i wrote "60°" i meaned a 30° "angle of base to arm" as you say, (i'm gonna change the original post).

Well, thanks for the information sungod3k! if anyone have more knowledge i'd love to hear it

The 80% rule is not really great to use. Its a rule of thumb. You want to know exactly how large you can print. Do some trig and calculate the minum angle to be 20 degrees. Add in your effector offset and you'll get something much closer to reality

Quote
3DRapidClone
The 80% rule is not really great to use. Its a rule of thumb. You want to know exactly how large you can print. Do some trig and calculate the minum angle to be 20 degrees. Add in your effector offset and you'll get something much closer to reality

Ok so, let me understand correctly, you're saying that the arms should be long enough so this ( [drive.google.com] ) angle it's never less then 20° while printing?

I may be a novice in delta printer mechanics, however I think that having that angle go below 20 degrees is not of itself a problem. It could even go down to zero. There will be some dilution of XY position accuracy when the effector is close to one of the towers, but that is caused by the angle between the parallel rods from the other 2 towers becoming much lower than the ideal 90 degrees.

I am willing to be persuaded that the 20 degrees does matter, but only by reasoned argument.

Edited 2 time(s). Last edit at 01/16/2015 05:14AM 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].

Hi

Maybe you can have a look to the following link and more precisely the pdf file if you're a math fan

Paul

---

Site perso en construction / personal blog under construction: [histoires-de-constructions.net]

Quote
paul18
Hi

Maybe you can have a look to the following link and more precisely the pdf file if you're a math fan

Paul

That seems rly interesting, even if it's gonna take some time to actually study it
Thank you

Edited 2 time(s). Last edit at 01/16/2015 12:11PM by Nova.

Paul, thanks for that link, it's good to see a mathematical treatment of the geometry. I wish I had found that PDF link a couple of weeks ago. I ended up working out the forward kinematics myself when I implemented delta support in RepRapFirmware.

One thing I have wondered (and reading that PDF makes me think I am on the right track) is that if we lengthen the delta rods a little compared to the traditional value, at the cost of a small loss of precision at the centre of the print area, we may gain a larger improvement near the extremities.

---

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].