All good things › Forums › Morgan Builders forum › Triangular notches in the support arms – why?
- This topic has 16 replies, 4 voices, and was last updated 9 years, 6 months ago by Quentin Harley.
-
AuthorPosts
-
-
February 19, 2015 at 01:49 #5927JeffRodriguezParticipant
I noticed these notches cut into the bottom of the support arm and I was wondering why they exist. My printer’s a bit misaligned and the split is having problems with adhesion.
-
February 19, 2015 at 17:52 #5929JeffRodriguezParticipant
Update: I disabled those notches in my local copy and printing went much more smoothly.
Still curious why the notches are there π
-
February 21, 2015 at 17:23 #5931Quentin HarleyKeymaster
The notches are there to help you measure your exact arm length. Because the SCARA inverse kinematic transform relies heavily on the accuracy of the arm length, you have to get it as accurate as possible.
Problems here will cause scale issues when printing.
-
February 25, 2015 at 22:29 #5939MacMarty15221Participant
Speaking of those notches….
I was looking at the Configuration.h file today, and I noticed that the inner arm length and outer arm length are established right there at the top. It set me to wondering – If I wanted to have a larger build platform, would it be as simple as building appropriate arms (say, 200mm), changing the parameters, and recompiling the firmware? My hopeful self is encouraged, but my cynical brain, scorched by long experience, bets otherwise. (Optimistic self also wants to paint 3m by 3m pictures in the street using the SCARA mechanism.)
-
February 26, 2015 at 09:54 #5941Quentin HarleyKeymaster
Your optimistic self is right. You could make the arms as long as you want, but remember that longer arms will have lower resolution if you do not scale the drive wheels accordingly.
-
March 1, 2015 at 16:22 #5944EujeanParticipant
On this thought (Imagining a garage size Morgan Cheetah offspring) won’t the longer arms cause the load to be furthest away from the drive shaft causing sag?
-
March 1, 2015 at 17:06 #5945Quentin HarleyKeymaster
Yes, it could. This is not a problem though. The bed level code will compensate for any sag.
Of course you will have to make some material choices to limit sag as much as possible.
-
March 1, 2015 at 20:35 #5946JeffRodriguezParticipant
Quentin, I’ve been working on my own Morgan clone based on your design. First off, a lot of the engineering decisions you’ve made become obvious when you dig into the design.I can see there’s a lot of thought and work in the Morgan, so congrats and thank you for your efforts.
Q1: What made you decide to install the Psi arm on the threaded rod instead of the tube? It seems like the tube would better support the weight of the print head and reduce sagging.
Q2: I’m planning to replace the upper tube bearing with three 608 bearings to simplify the BOM. Thoughts?
Q3: I’m also working on replacing the print head bearing with a single 608 and moving the print head to the side of the outer Psi arm (perpendicular). I haven’t dug into the interpolation code yet, but I’d like to think 90deg and an offset wouldn’t be too big a deal. If it works, a second print head might not be too far off.
Q4: I’m also working on replacing the smooth rods with angle iron, which I intend to use as two legs of a 3-legged table (ditching the PVC and associated feet). Are there any problems with that approach that you can see?
Thanks again!
-
March 2, 2015 at 06:07 #5947Quentin HarleyKeymaster
Thanks!
A1: Mostly arm placement. I have been working on a new set of arms that will be anchored on the pipe for stability, will post photos of it by next week.
A2: Interesting… How?
A3: You can make it work by just making the opposing arm proportionally shorter as well. This will keep the code the same.
A4: The current triangular frame gives it the stability in that the frame is under stress. Removing the stress and triangles will introduce wobble. Wobble brings unwanted artifacts. Your job would be to combat those.
Best of luck on the build! Pictures mandatory…
Q
-
March 2, 2015 at 07:32 #5948JeffRodriguezParticipant
2: Like a lathe steady rest:
3: Thanks. I’m not quite there yet, hopefully this week. I’ve been dealing with printer issues and finally have a set of decent drive wheels.
4: That’s what I kinda figured. I’m gonna have to get creative and see if I can stabilize with less hardware. Have you considered guy-wires?
-
March 2, 2015 at 09:53 #5949Quentin HarleyKeymaster
2: Does the bearings ride on the pipe, or is the bracket attached to the pipe and the bearings run inside a hole?
If the former, you will just need to ensure your pipe is perfectly round to prevent inaccuracies.
4: I did try them initially, but they stretched, and I kept having to adjust. The triangular config followed, so I cannot say if I gave it enough time. Perhaps if you use pre stretched wire it should be good.
-
March 2, 2015 at 17:18 #5950JeffRodriguezParticipant
2: I was planning on a direct pipe/bearing interface so that’s good to know. I’ve got some 3/4″ round welded steel tubing. I wouldn’t say it’s “perfect” 19.04-19.25mm. So let’s see, that’s 0.21mm out of round on the diameter, or 0.105mm on the radius.
4. I figured you might have π Good to know. That does get me noodling on what materials might be good guy-wires and a good way to tension them. It would be *really* slick if I could use the same material for guy-wires and line-drive, e.g. spectra.
-
March 2, 2015 at 21:54 #5951Quentin HarleyKeymaster
You could stick the pipe in a lathe, and just make that section round…
-
March 14, 2015 at 04:45 #5960JeffRodriguezParticipant
Here’s an update Quentin!
-
March 15, 2015 at 14:44 #5971Quentin HarleyKeymaster
Looking good so far!
-
March 19, 2015 at 03:24 #5979JeffRodriguezParticipant
Hi Quentin, brain pickings again π
Here’s a new idea for the frame and table:
Did you ever explore this corner-style format?
One more with 225mm arms and a 300mm build plate:
-
April 5, 2015 at 12:05 #6033Quentin HarleyKeymaster
Nothing wrong, except that you need to cut a notch in the cornet for the wheels to stick through. Your wheels are too small to have a decent resolution on head movement. Other than that it seems good.
-
-
AuthorPosts
- You must be logged in to reply to this topic.