Delta without parallelogram?

Hello everyone,
I am planning a printer delta, and during research for a bit of inspiration for something new, I found this:

torx robot

The most interesting aspect of this robot is that it does not use the parallelogram system and the effector moves without rotation through the placement of only 3 legs with a pair of universal joints at both ends.

This system greatly simplifies the construction of a printer delta and industrial universal joints, although a bit 'expensive, do not have the limitations of the rod ends in terms of rotation angles, thus allowing to expand the printing surface.


I tried to implement this system in the design of my printer, but I doubt that without parallelogram, a small misalignment can create rotations of the effector.

What do you think about?

Hmm,

by the looks of it, it could work :-)
However, I maunder the impression that the parallelograms of each tower have a great impact on keeping the platform horizontal.
Not sure whether that will be easily achieved with single arm and joints.

Try and if not... multiply the arms by 2 :-)


Thomas
Having both Cartesian (Orca) and Delta (Rostock Mini)

---

www.3daybreaker.blogspot.com

Orca V4.4 rebuild to Ramps with Mk8 and E3D, as well as a Rostock Delta Mini and an OLO in backorder :-)

That design allows for a lot of effector freedom which would be great if you can control it, bad if you can't.

If you read the last line of the pdf, they state that the joints were a problem, so while they could control it with some degree of precision, it wasn't enough for say, 3d printing. Granted, this was 14 years ago and you can say things have progressed, but one could also say what we use, is what evolved from those experiments.

Looking at older designs can be good for ideas, but you have to be careful that you aren't working backwards.

As for the pneumatic actuation, that could be interesting (and fast!), but probably not quiet.

Thanks guys for your answers,

as you can imagine I'm not a magician with the inverse kinematics ... and before I start the construction of my first 3d printer and burn a bit of money with a wrong project, I need some certainty.

@replace

I also think that the parallelogram is a sort of redundant kinematic system which ensures that the effector remains parallel to the bed

in fact for the purposes of kinematics is treated as a single link
but in action could be critical

your optimism pushes me to try, but if I am wrong I have to redo a lot of stuff ...

@ sheepdog43

yes, this design allow great print area and great freedom of effector (single cardan joint allow 45 ° + and -), but I'm still not sure of being able to control him :-(

However I do not want to make a Chinese copy of the project and use pneumatic actuators, but I'm just very fascinated by its minimalist design, and I'm looking for some kind of confirmation on the function of the single link

Control would need custom firmware and slicing.

I'm still looking for information to clarify my doubts and how the intuition suggested to me ..

I found a study that describes the error in parallelism effector relative to the plane, which can occur due to joint backslash, difference in length of the arms, various mismatches, etc.


In the delta system with parallelogram inaccuracies constructive reveal themselves as positioning errors, while in the delta system without parellogramma, the 'effector rotates as you can see in the attached file.

The image is inverted because it refers to another robot but is very clear ...


@sheepdog 43

This is my first project of a 3d printer so I do not have the certainty of being right, but from the information I've found it seems to me that the firmware running on traditional delta could work without too many changes even on delta with single link, i just have to insert correctly the geometrical parameters of my printer.

Regarding the slincing I have not yet informed of ......

I would HIGHLY recommend a more traditional build than this for your first. Your first design will cost almost double what it would had you had some experience, and that is just building a normal Reprap. Trying to design it yourself at the same time as something original is likely going to double the price again, if not more.

It's very easy to get in over your head just building an already proven design and the changes you need to firmware and slicing would only make that even more likely. I'm not saying you can't do it, or it can't be done, just that for this to have any chance of succeeding at a reasonable price in the (semi) near future, start with a smaller mountain to climb and get some experience first. You aren't just diving in, you're heading straight up to the high dive.

I really appreciate your warnings about the difficulties of developing an original project.

Even if I already have some experience in other fields, mechanical design was not my area of study,

and all that concerns the 3d print is totally new to me, although I'm trying to document myself as much as possible.

So I guess to start with a conventional delta, and once I can print my custom parts in 3d, I can begin to experience something new without wasting too much money.

Quote
Fede3D
So I guess to start with a conventional delta, and once I can print my custom parts in 3d, I can begin to experience something new without wasting too much money.

It's handy having a printer handy to design a printer, but also starting from a known place and the ability to quickly try new things will really speed your design along for little expense.

I priced a few parts out when designing the Griffin OS and what cost $2 and an hour on my Rostock, would have taken weeks and several hundred dollars from a place like Shapeways. By the time I was happy with the printer I had burned through A LOT of plastic. If I was outsourcing it, it would take years and tens of thousands of dollars. They wanted $300 per frame connector in plastic and $800 in aluminum for what I was doing for $2.

It's a lot easier to tolerate a mistake at $2 and an hour of your time than it is at $300 and a few weeks.

Your first printer ? Hmmmm.

I have built 2 now.

A full cartesian Orca v04x and
A delta Mini

I bought the Delta for fun, while the Orca suffered from a blown electronic part, but it is not fun. It looks amazing (one reason to build it ) but quality stays behind the cartesian.
Of course original parts quality has to do with that, as well al calibrating experience, but also those are better adapted on cartesian. For my delta to become better I will have to reprint parts myself, which I will do on the Orca :-)

So without being the expert: for your first build I advice cartesian :-)

---

www.3daybreaker.blogspot.com

Orca V4.4 rebuild to Ramps with Mk8 and E3D, as well as a Rostock Delta Mini and an OLO in backorder :-)

Quote
Replace
I bought the Delta for fun, while the Orca suffered from a blown electronic part, but it is not fun. It looks amazing (one reason to build it ) but quality stays behind the cartesian.
Of course original parts quality has to do with that, as well al calibrating experience, but also those are better adapted on cartesian.

Actually calibrating delta is easy [groups.google.com]
It may require some math skills now. But that is only because nobody bothered to add it to firmware and pc sw.

@ Replace

I would start with a delta, because they are simple to design,

3 symmetries, once you've developed 1-axis ...
you're done

as regards the precision of the delta should theoretically be very accurate if:

- The frame is very stiff
- there are no inaccuracies in the construction and assembly
- Very low backlash in the joints

of course it is not easy to get all this

The reason why I want to realize it and design it by me is that I have not yet seen a delta that I really like and want to make.

Many of delta frames that I see around with high aluminum pillars and connection beams only on top and bottom are very few stiff

Also those delta with parallelogram made by carbon fiber with a diameter of a few millimeters are very undersized in my opinion, it is true that they are very light and have little inertia, but they are also very flexible and will never be fast and accurate at same time.

I prefer to have a precise machine, even if slow.


The parallelograms made of tubes aluminum or carbon fiber are a good choice, but I hate rod ends because they have working angles very limited.

The magnets seem to work, but do not give me confidence.


Some time ago, I started to design my delta, but I'm still very uncertain on how to make the parallelogram and if I have the right idea I could even try to do it.

Otherwise reconsidered the idea of ​​a Cartesian printer, most likely a core xy.


@hercek

I read a lot about the configuration of the firmware of the delta, and having done the project with a CAD , I should not have a big problem to obtain the geometric data


I have not yet figured out how to configure the slicing ...

but I decided to do it later


Do you have some quick info for me? :-)



Bye

Slicing is not an issue .... Only the positioning of the object on the bed can surprise you if you don't take care.
For the rest, speed issues need tuning by trial and error only

---

www.3daybreaker.blogspot.com

Orca V4.4 rebuild to Ramps with Mk8 and E3D, as well as a Rostock Delta Mini and an OLO in backorder :-)

The slicing problems were in regard to a 5 axis machine where the nozzle tilts.
It's not going to be the same as normal slicing when your nozzle can print on the side of an object.

Ahh,
I missed that one.

This is funny, potential head crashes on the side :-)

---

www.3daybreaker.blogspot.com

Orca V4.4 rebuild to Ramps with Mk8 and E3D, as well as a Rostock Delta Mini and an OLO in backorder :-)