LISA Simpson

@cozmicray: LISA started as a design challenge after I felt bad for telling someone on this forum that lead screws are a horrible idea for column deltas. It was mostly a joke. I never intended to build it. My students happened to need a project and everyone seemed to like the concept so I threw caution to the wind and let them build it. For me, that is the end of it. I have shared the idea and the files and it is up to the community to carry LISA into the future.

LISA is inherently bad because any error in the straightness of the screws will be doubled because of the rotation and could be even more than doubled if you go for high speed rotations. It was a fun concept but it doesn't fit in my value engineering vocabulary.

Edited 1 time(s). Last edit at 07/16/2014 05:19PM by nicholas.seward.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Quote
Dejay
I guess an alternative might be to create a really big 3D printed bearing, similar to these, using delrin balls

... for a better DIY-able 'recirculating bearing' concept cylinders instead of balls are even better for linear bearings ... simple cut a smooth rod in short pieces and fill them into the tray

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
cozmicray
did you get some of these ball screws?

Hi, sorry no I didn't order them. Although I am tempted, I'm not sure if I want to build a LISA. I like the concept and would like a delta printer that is strong and sturdy enough for a multi direct extruder and light milling, I'm not sure if this is the way to go. I guess the main problem is that any forces acting on the effector (acceleration / deceleration but for milling also buffets) directly act on the precious ball screws. Like Nicholas said, as soon as they bend it's all over. So it's not really something that helps me with what I ultimately want (print large scale simultaneously with 3-4 different materials).

I thought about stabilizing the leadscrew but it's difficult because of the rotating carriage. I guess the best way would be to go back to the delta design and have a wheeled carriage on an extrusion and use the ball screw instead of a belt. But that is just much more expensive.

I'm still trying to get my first printer (Kossel Mini) working but I try to build a "wheel screw", a rubber wheel on a precision steel rod.

Quote
VDX
... for a better DIY-able 'recirculating bearing' concept cylinders instead of balls are even better for linear bearings ... simple cut a smooth rod in short pieces and fill them into the tray

I thought about this too in the thread. I wonder if you could actually use "rubbery" recirculating rods for a threadless lead screw? The rubbery mini rods would circulate in a spiral around the rod, having friction in one direction but rotating freely. Probably not feasible because when the mini barrels touch each other they would create high friction and also have friction recirculating back. Just another crazy idea In any case, the delrin balls are really cheap and precise. I dream about being able to print something like this completely assembled with 3 materials, low friction hard material, rubbery flexible material and easily dissolvable support material.

Quote
cozmicray
What are "these"
link is not right?
Please explain these recirculating ball screws?

Not sure why the link doesn't work, but FYI below are some more links. The link is simply to make a "DIY" printed linear bearing that runs directly on extrusions (like lm8uu bearings but for extrusions instead of rods). This saves you wheels and mounting them etc. Theoretically this could save time and money and space, practically it's probably a bit more hassle. I want to try to build this though because it's cool.

My idea was probably better explained as "lazy-suzan rings" (didn't know the name) to create "typical" delta effector platform that can carry multiple extruders. Basically a big bearing with effector in the middle the arms attached to the outside. This concept will likely not be bearing fruits (sorry for my unbearable puns!) since Nicholas multi extruder concept looks far easier to build. Also the problem that shoulder joint offsets and effector joint offsets need to be the same if I understand correctly, so the bigger the effector, the bigger your shoulder and you loose even more print space.

[github.com]
[github.com] (updated, easier to print version)
[www.flickr.com]
[youtu.be]

Sorry for the long post and noise!

@Dejay: I am going to experiment with my own acetal/derlin lazy susans for the GUS hub. I will also break the convention that the shoulder offset and the hub offset are the same. It will make the math harder but I have some Azteeg X5s coming.

The print volume actually goes up as the size of the hub increases to a point.

LISA is not a super rigid arrangement. A traditional arrangement is what you need for a mill.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Here is a printed slew bearing that you might be interested in Nicholas.... [www.marginallyclever.com]

Hello,

If the problem of precision is due to the rotating screws, why not turn the carriage nuts, each equipped with a PAP motor.
It allows to have screws of very big section, and more rigidity.

Quote
Pascal68
Hello,

If the problem of precision is due to the rotating screws, why not turn the carriage nuts, each equipped with a PAP motor.
It allows to have screws of very big section, and more rigidity.

I thought about using something like this.

The main problem is the screw wobble so this is a great solution. The secondary problem of having horrendous kinematics isn't really an issue these days. I initially shied away from this approach because one of my main guiding design principles is to have a static stepper if at all possible. I will have to think about this some more. Maybe LISA has some practical potential.

Problems I see with this approach... (none of them are bad but I felt I should delineate them)
*Large enough screw that will double as the support will be heavy.
*Nuts for those screw will be quite large requiring very large bearings. (weight and cost concerns)
*Friction will be higher for this design especially if we do some sort of anti-backlash.
*Wire fatigue.
*Large dynamic mass.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Hello,

Thank you for answering so fast.

For the nut, it will have less inertia in rotation than a screw that rotates.
For the standard pitch may be sufficient by changing the PAP motor to screw ratio.
For the bearing it can be replaced by 3 bearings placed on the periphery of the nut.
For the size of the screw I was thinking about a Ø 30 mm, ensuring the rigidity for 1 m 50 of height.

The advantage of the LISA is to have a heavy effector without problems, that's what interests me.
I currently have a Kossel 1 m 50 and I can not implement complex effector :
[forums.reprap.org]

For the cables of the motors I thought of the cable in tablecloth.

But for python mathematics, can we use them with Smoothieware firmware ?
I think we need to put the mathematics program in Repetier, but I'm not sure.

Excuse me if my translation is not very fair, I am French.

cordially

For the screw I expect a price for 1m50 at IGUS, with a nut anti backslash at 34,24 €, in Ø 24x5

The IGUS nut anti backslash in 24x5:
[www.igus.fr]

On the other hand, the IGUS screws are expensive, the quality having a price, 154,08 € piece, better to try this other reference of screw:
[www.ecmu-csr.com]
To 44,78 € inc. VAT piece.

Even on a length of 1m50 one should not have bending.

I don't see how you can use only 3 bearings. To kill wiggle in the z direction I think you would need 6 bearings at least. That said...I found many large enough bearings that can go around the screw for less than $5 which keeps the total bearing cost to $30 which is nothing.

Agreed that the rotation of the nut is low inertia but now the carriages have larger inertia. I don't think this will be a problem. In fact, a nut with backlash will perform much better when the carriages get heavier by having gravity peg the carriages to one side of the thread.

I personally would go huge with the screws. I found this.

I am not exactly sure what a PAP motor is? I think I know what you are talking about but can you provide a link.

Now for the bad news, rostock style deltas have a ton more rigidity than LISA. (Let me say that again...LISA is not very rigid.) There are things you could do to make LISA more rigid but the geometry is just inherently less constrained than the standard configuration.

Can smoothie do this? Sure. It won't be too hard. I already know the math. (I have solved the inverse kinematics. I don't remember if I solved the forward kinematics. That might be a trick.)

One thing you might have never thought of is homing. The interdependence of the axes cause this to be a tough computational problem. I might suggest a simple solution. Some controllers like the DUET (I think) can turn down the current and just run the carriages into a wall. Otherwise, you are stuck manually homing or an iterative approach with a decreasing backoff.

Sorry to say that LISA just isn't super practical. However, it has style so I would be glad to help if you want to keep going down this path.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Sorry for the word PAP, I mean the Nema 17.
For the rigidity of the LISA, one can apply what I did on my kossel quoted in the link.
The tube is 2020, and the rods are diagonally threaded in M10.
And it's very rigid.

Remains mathematics not easy for me ...

I am saying that the arm geometry is inherently less rigid than other column deltas. There is very little we can do about that.

One solution is to have two screw on each axis with two or three spinning nuts. This will allow you to use the standard 6 arms. However, that is just a more complicated way to do the original.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Hello,
There may be solutions for the geometry of the arms, to reduce the deviations.
I write them as soon as possible.

Hello,

If the arms go from the axis of the screws to the axis of the effector, the deviations will be eliminated.

Mathematics will surely be simplified.

For the arms I thought of the carbon tube big diameter, with at each end a fork fixed to the big bearing.

It takes 3 rolls superimposed in the effector
The fork of the arms must be twisted before attachment to the effector So that the arms do not interfere.
The XYZ limit switches will have an offset equal to the superimposition of the bearings of the effector.

Mathematics should no longer need offset since it works from axis to axis.

My hypothesis is perhaps not good, and perhaps you have already thought of it?

Edited 2 time(s). Last edit at 04/04/2017 04:06AM by Pascal68.

Can you clarify what you mean by axis of effector?

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

The line passing vertically through the center of the effector.

If this solution can work, the arms must be very rigid in torsion.

For the trapezoidal nut, printed IGLIDUR plastic, it would be better.
I do tests in PLA but it's not easy.

Another option for the nut, a teflon molding, I am looking for ...

A trapezoidal bronze nut 40x7 costs 90 €, the idea is to mold it in epoxy resin with teflon charge powder.
The trapezoidal screw 40x7 serves for the inner shape, the release is wax, to create the game.
The outer mold may be the part equipped with a toothed wheel, ready to be used after pouring the resin.

[kQAAOSwA29Y3WS3" target="_blank" rel="nofollow">www.ebay.fr]

Edited 1 time(s). Last edit at 04/14/2017 04:30AM by Pascal68.

Hello,

Here's how I see the arms, with forks.



The forks do not touch.

The trapezoidal nut in resin: 10% epoxy, 45% powder bronze, 45% graphite powder.

L'effecteur n'a pas de roulement, remplacé par le mélange résine + bronze + graphite.

La vis trapézoidale n'aura pas besoin d'être lubrifiée, grâce au graphite.

Plus qu'à dessiner les pièces qui reçoivent la résine.

Edited 2 time(s). Last edit at 04/17/2017 09:43AM by Pascal68.

Now that's interesting! I had no idea that is what you where talking about Also every time you talked about forks I got hungry and distracted

I'm sure Nicholas can tell you better, but I think the mathematics are still the same. Any horizontal movement of the effector / arm screws each carriage up and down the lead screws. I don't think this is a big issue either way except for homing and probably homing and auto-calibration.

I too was interested in Lisa for a sturdier delta style assembly, and it looked to me like that too. But sadly it's not, except with a giant massive lead screws. Hmm well maybe it's time someone made a lisa with 3D printed lead screws? [i.imgur.com]

What Dejay said.

Edited 1 time(s). Last edit at 04/17/2017 01:10PM by nicholas.seward.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Sorry Pascal

And now all say it with me. DeJAY Actually the Y is silent, it's like Dejah

Hello,

Motorized trapezoidal nut with Nema 17.
What gear?
How RPM?
How many volts?

I think 24 volts, 600 RPM, chevron gear, what do you think?

Hello,

One problem is the speed of rotation of the nut.
The best is a toothed belt between the motor and the nut, and bearings between the nut and the motorized trolley.
For the effector the bearings Ø 60 mm are not expensive, 6 € piece.
I begin the drawings ...