Progress of my own Morgan
This bill of materials is based on the status of my own build of a RepMorgan. You can see, how far I've got until now over here:
The actual Inventor-files (including the assembled Morgan from the picture above) can be found here: File:RepRap Morgan Inventor Files.zip
- Update: Wed. 08/14/2013
My struggle with Quentins pipe-supports and pvc-pipes lead me to redesign them, because I was unable to use his original files in a way they fit together acceptably (neither in Inventor nor in reality). I'll update the extended BOM.
deprecated: The bracket from Quentin's repository doesn't seem to fit. Here's a new version for the z-mount-bracket with an axis distance of 175mm instead of 190mm. This is since resolved in the repository (qharley)
- Update:Fri. 08/16/2013
The CAD-model now contains screws and gets more complete. Parts have realistic movement-parameters, so the completed model should be able to act like my real RepRap Morgan. I will update the extended BOM soon.
- Update: Sat. 08/17/2013
It's alive. The virtual Morgan has it's arms and toolhead and they can be moved by turn the steppers, just like in reality. Now I'm only a few parts from completing the model. I'll try to make the z-axis behave realistic too, but don't know how to do that - we'll see. Completing the building platform is the next task.
- Update: Sun. 08/18/2013
I've worked on the real Morgan prior to complete the virtual one. At this point I have to recommend a tool that kept my project going, while some of my parts broke or didn't fit. My pvc-pipe-supports, that were printed by a member of the community, didn't fit and two of them broke apart in my experiments. As if that wouldn't have been enough the z-mount-bottom broke too, when one of the smooth rods didn't fit as expected. Additionally the bed-z-mount-bracket had the wrong size (it's rod-distance was 190mm instead of the required 175mm). Instead of waiting for new parts I decided to repair them, using a special industrial glue. It consists of two components, a very fine granulate and a cyanoacrylate adhesive. In the very moment these two components come together, they build a very strong compound that is extremely hard and like welded to the surfaces it has contact with.
It can even be used to glue broken car bumpers. The plastic will break on any other position than the glued region. It can be found over here: Hosch-Kleber
Sorry for advertising this here, but it's a unique product with its abilities and one of my most important and frequently used tools for repairing and building. It's quite expensive though, but worth every single (Euro-)Cent.
With this I finished my real Morgan to the point of it's virtual brother this afternoon, without the need to reorder some of the printed parts. But reality hits me anyway: I only had 1m of T2,5 6mm drive-belt. Not enough for my Morgan, that requires app. 1,2m with its actual setup, and I had to order some more. And I need to buy some grub screws (replaced them with standard M3 screws for the moment).
My new designs for the top and bottom platform do fit as expected and I even don't need reprinted pvc-pipe-supports (but the newly designed ones would be ways better, of course).
I'm using a threaded rod for z-axis for the moment, because the bed-z-axis-bracket only fits with M8 nuts at the moment (or with Quentin's Alpen SDS solution, that isn't an alternative for me right now). So one of my Morgans first printjobs will be a new bed-z-mount-bracket that fits with my trapezoid thread and nuts. (Addendum: Or maybe it will be a threadless z-axis. Look at the forums for more on that topic: ).
- Update: Fri. 08/23/2013
I've designed a new building platform and integrated it with my new PCB for the heated bed (RK2). This PCB has a very homogeneous distribution of the copper traces.
I've tried a three-point-adjustment for the platform. It is not tested yet, but there are several spare montage holes if three isn't stable enough.
The copper traces of the heating-PCB are covered with isolating lacquer (Polyurethane) and for thermal contact to the upper aluminum building platform thermal heatsink paste is used. The aluminum building platform distributes the heat from below, so that there is a very homogeneous thermal distribution on the building surface.
The fan for the V3-toolhead is mounted already, but I didn't draw the printhead with Inventor yet.
- Update: Mon. 08/26/2013
Damn. My (60€) laser printer doesn't work anymore, just when I was printing the new PCB heated bed layout. I think I'll have it repaired until tomorrow. The real brother of my virtual Morgan lacks his heated bed though. For a few days.
I'll have to use M4 instead of M3 screws for the building platform adjustment of my real Morgan, because I didn't get knurled nuts in M3 in our DYI-store. The PCB-layout works with both (there's enough room for M4).
- Update: Sat. 08/31/2013
Didn't manage to repair my printer until now. I'll try further, but buy a new one too,because I need it for the progress of my projects. The new heated bed still isn't made though but I think this weekend will bring more progress to the project.
Meanwhile I worked on the virtual Morgan, making the model more accurate and completed the movement-dependencies. The z-axsis now works realistic. The whole model is made up of groups of components. This prepares it to ease exchanging complete function-groups with new designs (e.g. air-bearings).
A very simple and effective air-bearing concept for the z-axis is almost ready to be published. It will replace the smooth rods by two very inexpensive aluminum-rails (L-shaped, 10mm each side 1,45€/m). The LM8uu-bearings are no longer needed. The z-mount-bracket and the bed-arms are redesigned instead and a simple aquarium-air-pump works as a compressor (will be replaced by a printed one later, of course). This is experimental of course. I'll finish the classical version to be able to printed the needed parts.
Here's a preview of the new z-axis-design, prepared for the air-bearings:
- Update: Sat. 09/07/2013
As you can see, my real Morgan needs some correction on the PSI- and Theta arm. I think this weekend will be the one to start with the electronics. I have to use my prototype heated bed, because I still don't have a functional laser-printer to realize my new pcb-design.
The pipe-supports don't fit, because I had to use them unmodified, but wanted no tension in the pvc-pipes. Nevertheless it's absolutely stable and rigid.
On the right photo you can see a detail of my z-axis-mount-bracket. I had to cut it, because it had the wrong dimensions. Using my favorite adhesive for that (Hosch Industriekleber), the glued section will never break, because it's more rigid than the rest of the bracket.
I used that adhesive to glue the pipe-supports to the pvc-pipes too and to repair a pipe-support that broke during the montage.
- Update: Tue. 10/15/2013
After reorganization and expansion of my workplace, iow: after my victory over chaos, I've etched my new heated bed (2,4Ohm) and cut a new aluminum building platform. Now these parts of my real Morgan look like in my drawings. The PCB of the heated bed will be placed under the building platform with thermal conductive (while electrically insulating) epoxy. Today I'll also rearrange the PVC-pipes, so that they get their tension, that's intended by Quentin. New photos will follow and you will see the very limited space, where my Morgan grows up.
- Update II: Tue. 10/15/2013
Here are the pictures I've promised:
- Update: Sun. 11/03/2013
Recovering from influenza I've just configured my Morgan hardware (Taurino Mega2560 R3 Power with RAMPS 1.4 & RepRapDiscount LCD controller) and software successfully. The modified Configuration.h file can be downloaded from here: File:RepRap Morgan Marlin armlevel firmware Configuration.h
In the days before I already modified an ATX-power supply (700W should be enough) for my needs.
My next tasks for today: - Connect my Morgan to the Taurino
- Smoke test
- preparation of the new big heated bed (printspace 620mm x 260mm app. halfround)
- Update: Mon. 11/11/2013
My Morgan is alive now. All electronics are connected and installed and have passed the smoketest successfully.
Now I'm struggling with calibration and Marlin-configuration. Especially the endstops didn't work as expected, and correct homing was impossible. After reading about that in several forum entries I think I can fix that problem this evening.
The new big heated bed still isn't finished (was working in the garden due to unexpected very good weather - very unusual for November over here). But the bed of nails that will form the copper-wire-loops is almost half ready. Believe it or not, I ran out of nails on my first attempt.
- Update: Mon. 12/03/2013
It took more days than expected, but that was not due to technical problems n my project, but other themes needed my attention. The bed of nails is ready and the heating-wire for the big platform is formed:
I only had wire with 0.3mm at hand. This gives a too big resistance of 9.6Ohm (measured). I will unwind the wire from the bed of nails, fold it into 3 segments of equal length and mark the resulting segments. Attached to the bed again I can cut the segments and connect them in parallel, giving a total of 1.0667Ohm resistance (three times 3.2Ohm in parallel). This should heat up the platform quickly, with 11A and 135Watt.
- Update: Thu. 12/05/2013
The platform is ready. The platform is of 3mm thick aluminum, 625mm x 268mm. The heating wire, prepared on a bed of nails to form the wire-loops, is glued to the platform with hot glue that withstands 160°C (hot enough for the max.100°C of a heated bed).
New ideas and notes
More stability for the frame
The stability of the frame could be improved by using steel cables or sth. similar to pull the supports together with some tension. Therefor there could be some wing nuts on the top platform, to tense the cables. (drawing will follow)
I'm planning to extend my morgan (once it is completed) to three print arms. With this configuration, it should be possible to use an extreme fine nozzle for the (slow) detailed work, a big nozzle for fast infills and a third one for support structures or e.g. to combine two materials with different properties into one print. Using the three printhaeds one after each other (they would collide otherwise) could make it possible to use the same electronic for all three printheads by completely switching the connections between the three sets of steppers, printheads and RAMP.
It should be possible to extend the print height to almost any size, because there's no horizontal movement of the printbed. Therefore it should be possible to extend the print-height easily by placing the Z-Axis behind or in front of the drive-wheels, instead of above. The count of necessary modifications for that sounds not too big. Even the firmware would no be affected. The problem, that the print material on the top could become too cold (because of a greater distance to the heated printbed), could be avoided by using simple infrared lamps to keep the top portion of the print warm.
Already discussing air bearings in the forum I'd like to place this idea here too. I came across this idea, because air bearings are common in high precision applications and they don't wear out, reducing the TCO and preventing losses in print quality over the time.
The following is only a theory at the moment and not tested in praxis. Just what I've learned from a simple air-rail experiment I saw as a young boy, it can be done with very little effort.
Lacking the access to a laser printer for a few days I've continued my work on the air bearings. Here'a detail shot of an air bearing for the z-axis:
One short cut of the aluminum L-beam is used as a counterpart of the bearing. The air comes through 4 small holes (0.5mm) on each side. Very small holes result in a high pressure in their direct surrounding and keep the distance between the aluminum surfaces of bearing and rail constant.
A little chamber behind the aluminum-plate in the bearing distributes the air to the small holes. They are connected to a tube-fitting that can be connected to a small compressor.
Calibration of the bearing-position on the bracket is quite easy. Peaces of paper between bearing and rail on both sides will keep a minimum distance while the bearings get fixed to bracket by two screws for each bearing (the bracket has longholes to make the position adjustable).
After removing the paper and starting the compressor, the bearing should work without any abrasion and almost no mechanical resistance.
Update 09/23/2013: Thinking about the air-bearing idea I've com to the conclusion, that one L-beam may not be enough for the platform weight and may bend in action, what would stop the air-bearing-effect, because the tolerances (between beam and platform-holder) would become too big. The following construction seems more stable:
It needs two rails per side, but at a price of app. 1,50€/pc. the extra printed parts needed have a greater effect than that. Using recycled materials as filament I don't care about extra printed parts though.
The next step to a printed air-compressor for my air-bearings is done. Here is geometry of the two screws of the screw-aircompressor.
Update 10/13/2013: The actual Compressor-Design with standard parts (app. 2€ for gears and 6€ for bearings):
With more experience in plastic based air bearings (something nobody seems to have done until today) it would be very smart to use air bearings in the compressor too.