scalability on dibond version

[3D-Darryl]

I scaled the Mendel90 to have a build area of 24x28x22cm (x,y,z) and there were some problems with the Dibond drawings. (1)The vertical edge of the right stay has a bracket attaching to the air. You can see it in the rendered SCAD. (2)The square tube on the right goes exactly over the holes for the wire ties, and over the holes where the wire exits (through the base) to the power supply. (3) The left stay has a square-bottomed slot which besides the fact that it is not millable with a round router bit, appears to have no purpose. (4) The left stay attaches on top of (not between) the brackets for the main gantry piece. It was necessary to trim a triangle out of the corner of the left stay in order to attach it.

[nophead]

Odd, I just tried it and can't see any of the issues you describe.

[3D-Darryl]

I can send you the config.scad file I used if it would help. We discovered the issues when assembling the Dibond. My friend rendered under OSX, and was able to see the issues (My i5 never finishes rendering).

[nophead]

I just changed the three dimensions you mentioned. In general making things bigger doesn't cause problems. Making it smaller can cause things to collide.

Did you change something else, like the PSU size, for example? I suggest you start with the original config and change one thing at a time until you find what breaks it.

[3D-Darryl]

I did this back in October, the version is from October 9, 2013. To my recollection I only changed the dimensions of the build area, but the config is attached.

The issues are not visible in an SCAD compile view, and the SCAD render does not ever complete on my i5 so I am unable to further troubleshoot it. We made the (four) Dibond sets from files generated on my Win7 laptop, but my friend independently downloaded SCAD on his Mac, compiled and rendered with this config file, and he was able to see the issues that we encountered.

On Sat, Feb 1, 2014 at 10:06 AM, Chris notifications@github.com wrote:

I just changed the three dimensions you mentioned. In general making things bigger doesn't cause problems. Making it smaller can cause things to collide.

Did you change something else, like the PSU size, for example? I suggest you start with the original config and change one thing at a time until you find what breaks it.

Reply to this email directly or view it on GitHubhttps://github.com/nophead/Mendel90/issues/40#issuecomment-33875337 .

[nophead]

I don't think you can attach files other than images to Github. I don't wish to spend any more time on this as as far as I am concerned it works with those dimensions. You can compare your config with the latest on Github yourself and see what you changed.

Nothing has changed since September except the manual and BOM.

If F5 doesn't ever complete then there is something seriously wrong and I am not too surprised it doesn't generate the correct files. I would never build anything without looking at the render first.

[3D-Darryl]

I used Gmail to attach but apparently GitHub blocks a 2K file. F6 never completes but F5 works. The instructions say to edit the config and run the Python scripts, so the significance of being able to view first was lost on me. Likely to annoy somebody but the config file follows: // // Mendel90 // // GNU GPL v2 // nop.head@gmail.com // hydraraptor.blogspot.com // // Configuration file // echo("DiBond:");

Z_bearings = LM8UU; Y_bearings = LM8UU; X_bearings = LM8UU;

X_motor = NEMA17; Y_motor = NEMA17; Z_motor = NEMA17;

hot_end = JHeadMk5;

X_travel = 240; Y_travel = 280; Z_travel = 220;

bed_depth = 254; bed_width = 294; bed_pillars = M3x20_pillar; bed_glass = glass2; bed_thickness = pcb_thickness + sheet_thickness(bed_glass); // PCB heater plus glass sheet bed_holes = [249, 289];

base = DiBond; // Sheet material used for the base. Needs to be thick enough to screw into. base_corners = 25; base_nuts = true;

frame = DiBond; frame_corners = 25; frame_nuts = true;

case_fan = fan80x38; psu = ALPINE500; controller = Melzi; spool = spool_300x85; bottom_limit_switch = false; top_limit_switch = true;

single_piece_frame = true; stays_from_window = false; cnc_sheets = true; // If sheets are cut by CNC we can use slots, etc instead of just round holes pulley_type = T2p5x16_metal_pulley;

Y_carriage = DiBond;

X_belt = T2p5x6; Y_belt = T2p5x6;

motor_shaft = 5; Z_screw_dia = 6; // Studding for Z axis

Y_carriage_depth = bed_holes[1] + 7; Y_carriage_width = bed_holes[0] + 7;

Z_nut_radius = M6_nut_radius; Z_nut_depth = M6_nut_depth; Z_nut = M6_nut;

// // Default screw use where size doesn't matter // cap_screw = M3_cap_screw; hex_screw = M3_hex_screw; // // Screw for the frame and base // frame_soft_screw = No6_screw; // Used when sheet material is soft, e.g. wood frame_thin_screw = M4_cap_screw; // Used with nuts when sheets are thin frame_thick_screw = M4_pan_screw; // Used with tapped holes when sheets are thick and hard, e.g. plastic or metal // // Feature sizes // default_wall = 3; thick_wall = 4;

On Sat, Feb 1, 2014 at 11:06 AM, Chris notifications@github.com wrote:

I don't think you can attach files other than images to Github. I don't wish to spend any more time on this as as far as I am concerned it works with those dimensions. You can compare your config with the latest on Github yourself and see what you changed.

Nothing has changed since September except the manual and BOM.

If F5 doesn't ever complete then there is something seriously wrong and I am not too surprised it doesn't generate the correct files. I would never build anything without looking at the render first.

Reply to this email directly or view it on GitHubhttps://github.com/nophead/Mendel90/issues/40#issuecomment-33877060 .

[nophead]

I don't think F6 will ever complete on any machine because it would try to make a single object out of the entire machine. F5 is used for assembly views. F6 is only used for making STL and DXF files of individual objects.

The reason it goes wrong with your config is you have the bed width and depth the wrong way round. So it makes the cut out in the gantry too wide compared to the length of the X axis, which then causes all the problems you saw.

[3D-Darryl]

Remaining questions, what is purpose of square-bottom slot (nothing to route through there) and why square-bottom is considered CNC millable with a round bit?

On Sat, Feb 1, 2014 at 12:04 PM, Chris notifications@github.com wrote:

Closed #40 https://github.com/nophead/Mendel90/issues/40.

Reply to this email directly or view it on GitHubhttps://github.com/nophead/Mendel90/issues/40 .

[nophead]

The long thin square bottom slot disappears when the bed size is set correctly. It is caused by the gantry fixing block overlapping the stay sheet. One of the screw holes for the block drilled through the gantry hits the stay and makes a long thin slot in it.

No it would not be millable but it shouldn't be there and isn't when the config is correct.

I use a cylinder 100mm long to drill the holes so that I can see them easily when I make them visible to check they line up exactly with the screws. Normally they would only affect the target sheet but the fixing blocks are processed as a list so all the holes are subtracted from all the sheets. As the sheets are orthogonal there are no drill collisions provided they don't overlap the plastic parts.