Thanks, its definetly a work in progress. I'm thinking about naming it the Armstrong A2 to pay homage to the groundwork layed by ttsalo.
I increase the speed from the last video, which was originally set to 30 mm/s. Slicer set to 100 mm/s
A2, I have not noticed the arms move vertically yet, but will keep a look out for it.
Theothermike, next time I take the arms apart I will take some pictures, but the elbows each have two 608 bearings. The extruder is on a bushing. That's a 5/16-18 stainless steel bolt that is drilled for the ptfe tube to pass through. The tube is threaded into the nozzle inside the heater block.
Edited 1 time(s). Last edit at 08/19/2014 09:03PM by Evil Monkey.
Also need to clarify some firmware variables. The scars_axis_y variable refers to the position the hot end is in when the print begins. It is measured from the end of the build platform. If you look in the video, you will see 200 written on the platform, This point is 200 mm from the axis of the stepper motors. In the firmware, I am currently set to -80, so measuring from the 200 mm mark inward by 80 mm results in the point at which the hot end is placed prior to beginning the print. My assumptions on how the arm steps are calculated proved to be incorrect. By using ttsalo's calculation and substituting 51 for his 25, the 1 inch cube was coming out as 0.5 inches or half the correct size. I multiplied my calculation by 1.5, which got the cube closer to the correct size.
I needed a away to level the bed for the time being, but did not want to drill any more holes in the aluminum bed right now. This is my solution. The 8-32 bolt is threaded inward which in turn pushes a small lever up to raise the corner. Little plastic feet keep the glass in place.
I took the printer a part to paint the sheet metal and give the arms a bit of a polish. Took some pictures of the arm bearings in a exploded diagram sort of way. Hope that clears up some confusion for you, theothermike. They are held together with a 5/16'' bolt.
Ah...perfect. Thank you so much for the detailed Images. So it seems that you´ve milled the bearing seats in the upper arm from two sides...wasn´t it somehow tricky to align the seats on both sides coaxially?
I am still messing around with the firmware. Currently trying to disable this startup sequence where as soon as start print is clicked in Pronterface, the y axis swings counterclockwise until the hotend heats up. The problem is this drives the y arm into the x axis arm. Does anyone know what part of the firmware governs this and how to disable it? Startup Sequence
Since I am at an impasse with the firmware, I am considering a mechanical work around. Redesigning the arm configuration allows me to use Morgans firmware yet not need to redo any of the sheet metal or rebuild the entire machine.
The initial movement of the Y axis appears to occur independently from the gcode. I say this because I edited some gcode to make the x arm move first yet when the start button was pressed in pronterface, the y axis still moved first. This is a problem because when the arms are manually set in the correct start position, this initial movement causes them to collide. You can see in the youtube video I made that the y arm swings over as soon as the program begins. It did not collide because it wasn't in the correct start position. Suarabh Lanje commented that adding a homing function got rid of this problem, but I have not been able to figure out how.
I haven't played with pronterface before but from what I'm reading the problem is probably something in the start.gcode file that gets sent to initialize the printer before it starts printing. Perhaps you can post the contents of that here?
I read through your blog on the printer and this thread but it's still not quite clear to me if you are using ttsalo Marlin-SCARA or ErikZalm Marlin. I'm guessing you are using ttsalo. The M109 code calls plan_buffer_line in the #ifdef EXTRUDER_RUNOUT_PREVENT section which activates the stepper code. It looks like all the axes except the extruder are supposed to be disabled at that point int he code but that obviously is not happening. I would try undefining that section and see if that helps.
I'm a little behind doing the full write up on that blog, but yes I'm using ttsalo's firmware. The printer is currently in pieces, but I will reassemble it, undefining this section section and report back what occurs.
Yeah, kind of wish scara's would catch on like delta's did. There would be a lot more development then. Spring loading the geared stepper motors works really well and simplifies things. The morgan configuration adds a bit more complexity to this design, but I just want this printer running so I can move on to other projects.
In time perhaps. I'm better with the machines than with coding, though I am learning Python at the moment. Lyman released the fifth or so iteration of his desk top filament maker. Very interesting design that I want to build upon and make more compact so that it doesn't take up quite so much of the desk top. So I'm trying to finish this project up, but I will definitely keep the original arms in case someone else figures it out. This arm subassembly is held on by 4 bolts and is very easy to swap out.
Any idea how to calculate the steps per millimeter with a total gear reduction of 49:1, 1.8 degree stepping, and 1/16 microstepping? Was looking at the Morgan M92 setup and I'm not sure where Quentin Harley arrived at his numbers. He's using a 11.78:1 gear ratio. Morgan Startup and Calibration