Tri-Polar Bot Concept -- Meet Wally

@see3d

I have unfounded concerns scaling up Wally with longer overhanging arms.
I think you will create a singularity (wobble, sway)?
Can you envision 1 meter long arms working?

“What is your idea of the ideal print envelope for a general purpose desktop 3D printer?”

Look at the print envelope of the professional machines for guidance (i.e. companies that have stock).

Stratasys smallest printers:
5X5X5
8X6X6
8X8X6
This is your start point.

Does Wally bolt directly to the wall?
Clever idea for saving desktop space if it does.
I have not read this thread, I've only caught bit's and pieces.
A wall mounted printer sounds attractive for a cluttered office.

What is the current print envelope of Wally?

@ Annirak
“And what's wrong with Wally's print area?”
Sounds like Wally is not scalable?

“Why do you want to create a 3-DOF bot that requires 5 actuators? “
You get a larger print envelope without using linear rails and bearings!

A2

Edited 1 time(s). Last edit at 10/25/2013 03:45AM by A2.

Darn, I can't wait when Nicolas starts a project at kickstarter. If plastic parts kit for Simpson or Wally will be cost around $60 it will be a bomb. I have not possibility to print the parts, so I wait. But from now I start to cut a chassis (back, bottom and side panel) for Wally.

Me as well. I can get everything except the plastic parts, so I, too, have to wait.

Here is the state of affairs.

1) I am getting electronics made that meet my requirements. That is a big undertaking so it will be pushing back any KickStarters for my printers.

2) Selling printed parts will be hard when it takes 48 hours to print a printer. The margin would be pretty small to be competitive. I doubt I could justify the volume for injection molds. That leaves laser cut machines.

So the moral of the story is, I will probably KickStart with a laser cut machine no earlier than April. I am working on a 1 arm SCARA that will be designung specifically for a laser cutter.

Don't worry. I am dedicated to RepRap and am building up a bot farm to that ends. It would just be dumb to commit to a KickStarter for a time intensive printed bot without doing test production runs and crunching the numbers.

You could check at [www.reddit.com] and see if someone would print them for you. Heck, I could probably print one for you but I'd want a bit of compensation for my time and plastic.

I was hoping you'd be able to make a medium sized bot farm work. It would be nice to see 3d printing demonstrated as viable for small run production. Is QuadMorgan not looking suitable for this?

How about resin casting?

Edited 1 time(s). Last edit at 10/25/2013 11:33AM by asb.

Hmmm... some of those parts could be cast in resin, but others probably would be more difficult.

@sgraber
How much do you want for all Wally's plastic parts?

Honestly I'd have to check to see how much plastic it uses. Was the approximation 1kg?

1kg+/- depending on infill.

It would cost $300 to print at MakeXYZ.

Whenever I've printed for people, I normally charge plastic cost times 3 or 4 depending on machine time and how much I need to baby sit the print. 1kg of PLA runs me right around $20. Does that put it out of the ballpark for people? If not, I'll print for you guys. I've checked with Nicholas and he's OK with this as it gets some Wallys in the wild.

Edited 2 time(s). Last edit at 10/25/2013 12:16PM by sgraber.

If I not find a printer, I've a thought to make the arms and bracket from MDF 22mm. This is not easiest way, but I try.

@nicholas.seward
How much distance between axles in the arm? Or is it doesn't matter and I always can alt this values in firmware?

@sgraber
So if you want around $80 then I first

Edited 1 time(s). Last edit at 10/25/2013 12:24PM by Serenum.

nicholas.seward Wrote:
-------------------------------------------------------
> I do .3mm layers with 15% infill. That is
> probably just over 1 kg of plastic.

Some parts are not nessesary for me so total weight might be lower.

Edited 1 time(s). Last edit at 10/25/2013 12:29PM by Serenum.

Serenum: OK you're first in line. If anyone wants on the list, please PM me.

Looks like a hiccup 35 seconds into the video?
Why would the print head move diagonally across the part?
Simple vase so I don't think it was changing geometry.
I imagine the same issue would be in the Simpson slicer?
[www.youtube.com]

In this video you mention that the slicer was modified for Wally:
[www.youtube.com]

Has this been corrected or am I seeing things?

A2

Edited 1 time(s). Last edit at 10/25/2013 02:27PM by A2.

@A2: That is how Slic3r did it. I could put it in spiral vase mode but this was a good test of the retraction.

Nicholas: how many perimeters were you printing @ 0.3mm layer height and 15% infill? I normally do 3.

I had found someone nearby who can print a set of parts for about $0.25 per cubic centimetre, which is why I earlier asked about the volume of the parts. I don't know what he charges in terms of weight.

I would interpret that as cubic centimeter of filament used. When you slice a model with Slic3r, if you look at the very last line of the generated gcode file it will say something like:

; filament used = 4135.2mm (29.2cm3)

That is what I'd think that volume quote would mean if it were me.

At 1kg material (1000 grams) using a density of 1.1 g/cm^3 for PLA, 1 kg of material would be 1100 cm^3. The math:

1100 cm^3 * $0.25/cm^3 = $275

They're making some nice money @ $0.25/cm^3.

Oh and I just got done slicing all of the files and it does indeed take 48 solid hours of printing for Wally! :O

Shane



AbuMaia Wrote:
-------------------------------------------------------
> I had found someone nearby who can print a set of
> parts for about $0.25 per cubic centimetre, which
> is why I earlier asked about the volume of the
> parts. I don't know what he charges in terms of
> weight.

Edited 1 time(s). Last edit at 10/26/2013 07:23PM by sgraber.

What do you think if the inside of arms to make cylindrical chamber? It would save enough filament and time of printing without sacrificing strength and rigidity.

IMO after printing almost a complete Wally set over the last couple days for @Smooshed, I really think you're going to lose strength and rigidity going below 15% infill and going to a hollow part. I wouldn't do it. You'll save a bit of filament but sacrifice part rigidity, not to mention introducing problems when holes widen or start in the middle of a part, which then would require printing with support inside the part.

Shane


Serenum Wrote:
-------------------------------------------------------
> What do you think if the inside of arms to make
> cylindrical chamber? It would save enough filament
> and time of printing without sacrificing strength
> and rigidity.

@Serenum: I think the cylindrical chamber would work. However, you are replacing a low infill area with a chamber that will need its own perimeters. I suspect that the plastic savings will not be as great as you would hope. The ideal solution here would be to vary the infill as a function of the distance from the central axis. We can have near zero infill in the middle and 100% on the outside.

@Everyone: I finally got Wally retrofitted with all my changes and sat down with it last night to do a little calibration. The good news is that I fixed a bad sign error in my code. I can now print silly things like straight lines. I updated the preprocessor in GIT but there are still quite a few things that are going wrong.



The squirrel printed near perfect. However, it is hard to tell if he is skewed or scaled wrong. (My nieces and nephews have depleted my squirrel army so I had nothing to compare it with. When they visit they always ask if they can keep my test prints. What can I say to someone so cute?) I wanted to check the scaling so I printed the torture test. It is an obvious failure that is due to the combination of slipping pulleys, weird speed scaling issues, less than ideal retraction, less than ideal acceleration and jerk settings, etc. I will start going through the list and getting Wally dialed in. The torture test may look like crap to you guys but to me it looks like I am very close.

Known issues:
*The speed is scaling wrong. I am doing something majorly wrong. There is one spot that I landed on that scaled the the speed so wrong that it took 1 minute to move away from it. (What is concerning is this is the same code that I use for Simpson. It now seems like I may have never had Simpson's speed scaling fixed. I am sure it is an easy fix and I did something very stupid.)
*My set screws for the xy pulleys are too short but the next size up that I have is too long so I had some major shifting during the torture test.
*I need to come up with a test print that will allow me to dial in the exact calibration numbers. I have it 90% there but I have to be able to refine the numbers.
*I haven't bothered to fix the triangulation errors for the z axis. Now that I am fairly accurate on the first layer I will worry about it more so my prints won't get warped.
*The tension in the string pulls the arms down. This is a minor issue. It is very easy to software level your prints from a few quick measurements. (When I say measurements all you have to do is move the bed up until the hot end touches the bed at multiple places and you just record those numbers.) I think I will replace the tension block with a guitar tuner so getting the optimal tension is quick and easy. More importantly, you can remove the tension when you aren't printing.
*etc.

Edited 2 time(s). Last edit at 10/29/2013 01:19PM by nicholas.seward.

Nicholas,

Thanks for keeping us updated, I am sure you will sort out the issues. Do you have any other color but orange

Yes, thank you, your offer was too good to pass up.

I'm wondering, though, if you might not have completely realized what you were getting into :-)

We're going to go ahead and make a back board, sides and bottom out of half inch aluminum plate.

Battleship Wally :-)

sgraber Wrote:
-------------------------------------------------------
> IMO after printing almost a complete Wally set
> over the last couple days for @Smooshed, I really
> think you're going to lose strength and rigidity
> going below 15% infill and going to a hollow part.
> I wouldn't do it. You'll save a bit of filament
> but sacrifice part rigidity, not to mention
> introducing problems when holes widen or start in
> the middle of a part, which then would require
> printing with support inside the part.
>
> Shane
>
>
> Serenum Wrote:
> --------------------------------------------------
> -----
> > What do you think if the inside of arms to make
> > cylindrical chamber? It would save enough
> filament
> > and time of printing without sacrificing
> strength
> > and rigidity.

asb Wrote:
-------------------------------------------------------
> I was hoping you'd be able to make a medium sized
> bot farm work. It would be nice to see 3d printing
> demonstrated as viable for small run production.
> Is QuadMorgan not looking suitable for this?
>
> How about resin casting?


I am positive that I could make it work and I could make it work now. The problem is just that the margin is so small. Why would I print the parts for less than $0.25/cm^3 when I can make that much at MakeXYZ. I figure the market will calm down and the cost will come down to <$0.10/cm^3 and it will start to make sense to use 3D printing for production. In the meantime, I am going to focus on less labor intensive production items for my own sanity. (I will have RepRap versions of the laser cut printers.) Printing a printer for your friend is a labor of love but print printers for 100's of people at a small margin would be a crusade.

I am still planning on making a QuadMorgan. I am not sure if it is going to be my farm bot or not. BTW, I might not be the first person to make a printer in the QuadMorgan family.

Really I just wanted an excuse to print one. I showed the SeeMeCNC deltabot guys (RostockMAX guys) the parts on Monday at their shop and they were really impressed with the size of the parts! The more I'm thinking about this, the more I'm tempted to print one of my own now...

Shane


smooshed Wrote:
-------------------------------------------------------
> Yes, thank you, your offer was too good to pass
> up.
>
> I'm wondering, though, if you might not have
> completely realized what you were getting into
> :-)
>
> We're going to go ahead and make a back board,
> sides and bottom out of half inch aluminum plate.
>
> Battleship Wally :-)
>

What size printing area is required to print a Wally (or Simpson)? As it was mentioned it may not be suitable for a first printer, what size printer should I look for to be able to print out these parts?

I can print everything on a 120x140mm platform but it is tight.

Well here Weely is already printed. The mechanical part is already collected. I didn't find Bearing 6702, the next replacement 68012 - here outer diameter too 21мм, but internal 12мм therefore I changed hub screw to 12 mm (at me while it doesn't turn out to print a small carving and I turned hub screw on the lathe from polyamide).Начал to assemble electronics.
@Nikolas is necessary now to me your help. I while have no display. I will be able to print on Weely directly from the computer? explain as in more detail to use software.
where it is possible to look at design of a wheel at filament drive mount?
where it is convenient to put end sensors?

Edited 2 time(s). Last edit at 11/01/2013 06:02AM by Vladimir_Lukashuk.

No end stops are needed. I use TrinityLabs hobbed pulley. I will get back to you on how the software works.