TH Morgan is now out and working. What to do…

Thomas needs a wife… Lilian Vaughan Morgan was the real life spouse of TH Morgan, and as it turns out, was a geneticist in her own right.

What would you like in Lilian? I have my own ideas… like to hear yours!

Lilian needs to be an ever cleaner, simpler version of Morgan – with more than one print head. Real life Lilian was a geneticist as well as a mother, and managed her famous husband’s affairs so that he could focus only on his work. Our Lilian is a real multitasker!

The big things for me are two print head (one for support structures) and autoeject/continuous printing. If we can get this done for the Gada prize in a cheap package we will be doing good.

100% autocalibration would be nice. Turn on the machine right after assembly and magic happens.

3D scanning is all the rage now days. You can use the extra DOF I talk about later to help with the scanning.

On the material side, it would be nice to get cheap conductive filament. If they get that figured out, it would be nice to have a vacuum gripper with an additional rotational degree of freedom. This will let you print circuit cubes with embedded components. You can even leave tunnels in the design for water cooling. 3 dimensional circuitry.

If you want any more crazy goals I am here all week. :-)

So that's it?

Some crazy ideas coming through on the Morgan forums as well.

Most want more extruders, quite a couple wants to get rid of the copper pipes. Enclosures... No belts / pulleys... Worm gear drives?

optimise morgan,incorporate biger pulleys on arm,use smaller ammount of printed plastic etc...morgan could be a lot cheaper,easier,faster. belts or fishing lineworm gear could not be so precise...

Hi Woo,

Yes. The comcept I have been playing with will be all of that.

It already has

• Some trimmed down parts - more work here though
• Line drive arms with a much higher ratio
• The beginnings of a dual extruder toolhead
• A new arm arrangement concept is brewing, that will allow the machine to appear cleaner without sacrificing

how software works on actual morgan, any glitches, problems?


one another idea, to minimise printing time, also ammount of material, maybe using 3 part bearings for joints
[www.trgo-agencija.hr] (only showing type of bearing)

one of these in the middle, arm on each side, and 2 smaller on outsides, everything bolted together...those bearings are very cheap, i can get them in local store for about 1 usd.


with these type of bearings arms could be printed or lasercutted and bolted together(2 or 3 pcs of plywood, on one cutted flange, one just with hole for bolt etc..)

i could write those ideas whole day


i have developed my own concept of par. scara arm, but my most important criteria is that everything must be vertically very rigid, with not more play then "normal" cartesian bot with linear rails, didnt managed to get there yet.. but those details up lead to that...

Edited 1 time(s). Last edit at 09/16/2013 01:08AM by woo.

I am down to $0.25 USD per 608 bearing. They are overengineered for a 3D printer but they are the cheapest bearings. If someone has a line on cheaper bearings let me know.

Also, bearingless design is coming along. There is the double lamina compliant joint (DLCJ) which is awesome for passive joints. There is also the proportion gear drive joint (PGDJ) which is awesome for active joints. I have put a spring in the PGDJ and it keep 3 lbf on tension in the string. This equates to 20 lbf pressing the sides of the joint together. This removes all backlash and side play. This is sufficient to be used sideways for a Morgan arm or a Wally arm. I didn't believe it until I got to play with it today.

Proof:


My vote is for the 1 arm SCARA. If you use the PGDJ you can get away with no bearings for the XY. I am trying to come up with some bearingless z stages.

Edited 3 time(s). Last edit at 09/16/2013 01:26AM by nicholas.seward.

yes, those joints are looking good.

ok, we are talking in different directions, you are looking cheaper, diy solution, its ok, i know how stuff works when working series production.

i am looking for simplicity, and possibility to make better machine with not so good machine...so for me those 2-3 bearings or 10-20usd are not so important...
so if printer is only good leveled, and if z axis is leaned on one side with upper bearings new printer will be fine

just one question, those PGDJ are looking good, what do you think, how will this joint work when printed on not so good calibrated printer, we are assuming lets say some oozing , not 100% rectangular all axis?

I think the PGDJ's will be slightly self-correcting. Time will tell. Simpson will have to try to print its own arms. That will be the real test.

I am so excited about this project! I built a Pursa but my best friend took it with him when he ran away with his GF. I am axnious to build a 3d printer again. (and it think this one is totally amazing!!)

As I have no access to 3d printing I have been working on a design that takes no printed parts. Like the arms. I was thinking of making them out of copper pipe. For the lower drive train I was thinking bicycle gearing.

Becuase I have no access to a printer I am using what I have and trying to limit my tools as I know most people who build these do not have a metal lathe or welder.

Pardon my total ignorance but is there a better place to collaborate on this build? I have some parts list (not necessarly the cheapest but what I am looking at for parts).

Oh, and on an engineering question, why not have the legs straight?

Thanks!

Andrew

The legs are not straight to stabilize the platform. The legs are two sets of triangles (triangles are stable) and they are placed in such a way as to want to twist the platform away from the center, each to its own side. This creates stress in the PVC pipe, which in turn takes the flex out of the frame.

Morgan had straight legs initially, and it was swaying like a tree in the wind...

on the argentinian RUG we have made something similar to what you are searching

its a multiple extruders, scara mechanism but without arms, its movements are composed by two kind of platens, one with extruders and one with bed rotating, it can also add an scanner

it has 3 extruders but can reach up to 6, the only problem would be the cantilevered load ,that increasing costs could be solved

another problem is the calibration needed to reach the center, I solved it increasing the diameter and giving posibility to adjust bed manually

It is in a phase to do first machine but I personally dont have the capital necessary

its fully spanish spoken, but I can answer any question about it

[forums.reprap.org]

Another silly question. Why does the "prefered envelope" seem to be 8 x 8 x something? Why not 12x12x12? I feel like 12 makes for a more useful size (I know you can glue stuff together, etc.)

Thanks!

Andrew

Going from 8" to 12" cost way more than 50% more. The material costs will scale somewhere between the 2nd and 3rd power. Also, if you are using a heated bed it will require 225% more power.

That is my nay saying. Now for the good news. Designs like Morgan, Wally, and Simpson can scale up for less than anything else.

The main reason I stick with 8" for now is to make it easy to move around and fit on a desk while we are in the experimental phase. It also makes the R&D easier to handle on a teachers salary.

All that said, I am planing on making a Mega Simpson that can print 6'x6'x6' for a proof of concept of how scaling is easier when you don't go cartesian.

---

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

The main reason why I stick to 200x200 (I live in metric world...) is that the printer can fit comfortably on a desk, hotbeds are readily available, and "only" consumes 100 to 120W.
Larger printers will not be a problem if you resort to printing PLA only on an Acrylic bed,

Mega Simpson... That thing would need a blast alarm to go off at least 10 seconds before printing starts!
and it would need a room ;-)

@qharley: LOL. Mega Simpson will have to break down into a flat pack and will have to get setup outside when needed.

---

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

Hi

My thoughts about a mega Simpson was more in the direction of selfinflating parts. A mix of gogogadgeteer and an inflatable castle. Okay, now i am off wandering around with the next crazy goal for Simpson, but Lilly should do the trick. Apropos crazy goals. What about repstrapping the Morgan family out of paper?

I better should delete my thoughts....

Magic worx the way around.

Next silly question. What print head / assembly are you all using? The idea of removing the drive mechanism (and associated weight) is a fantastic idea!

Thanks!

Andrew

Machines like this needs Bowden cable filament feeds. Bowdens have matured a lot in the last couple of years, and has been running reliably on my machines.