Restart a print long after printing failure or stop.

https://groups.google.com/forum/#!searchin/deltabot/resume$20printing/deltabot/hHZ0AD5Ha1s/jTvc4-yH2T4J

Yes, so the gcodes that got missing (you want to print again) will gave to be preceded with I.e.

G28 ;home
M106 ;fan on
M109 S205 ;wait for temp 205 deg
G92 Ewhatever ; set extrusion to last position extruder
G1 Xstart Ystart Zstart ; set position to last position nozzle
( below is the starting point )
G1 first line that you want to print, including the “original” E position in that line

Naturally if your slicer inserts more magic before printing include that also.

> I understand deleting all the g code above the position I want to start, because It has been printed already, correct?
Yes

> Wouldn’t the G92 Ewhatever code be whats left to run?
G92 sets the extruder coordinate to the given position,
It’s what already has been run

Here is my story.

I’ve been testing Cura instead Slic3r because it produce better support and infill control.
One serious, dangerous bug of Cura is it generates X and Y coordinates that beyond physical dimension of the 3D Printer user sets up.

Bulbul Junior’s safe printable diameter is 150.00 mm.
X: -75.00 to 75.00
Y: -75.00 to 75.00

When I was printing a 3D model, the carriage of X tower banged so hard the lower endstop, it literally pushed down the endstop. For a while I thought it caused by loose bolt. Then second trial of the same 3D model, the hotend was moving to empty space.
“Oh, shit! It happens again.”

I inspected stl file of the model, remade new Gcode file in Cura. In third attempt, I noticed that dimension of the object is out of printable area.

190.00 mm was inserted by Pause at Height plugin which I had to turn on unknowingly.

The print goes:
- from -55.84 mm to 190.00 mm in X and is 245.84 mm wide
- from -43.40 mm to 190.00 mm in Y and is 233.40 mm deep
- from 0.00 mm to 101.11 mm in Z and is 101.11 mm high
Estimated duration: 1004 layers, 5:18:57

So I opened up half million lines of the model’s Gcode file, found the outstanding Gcode lines.

;TYPE:CUSTOM
M83
G1 E-5.000000 F6000
G1 X190.000000 Y190.000000 F9000
M84 E0
M0
G1 E5.000000 F6000
G1 E-5.000000 F6000
G1 X20.526000 Y5.598000 F9000
G1 E5.000000 F6000
G1 F9000
M82
G1 E-5.000000 F6000
G1 Z15 F300
G1 X190.000000 Y190.000000 F9000

Screen shot of 3D Gcode editor in Pronterface
dangerous_gcode_by_Cura

Above extreme coordinates occurred right after Z5.010, Z height 5.010 mm.
I wanted to save one hour of printing time and played with Gcodes. I deleted about 150,000 lines of Gcode which were before Z5.010, inserted three Gcode commands like below.


;Basic settings: Layer height: 0.1 Walls: 0.28 Fill: 20
;Print time: 6 hours 42 minutes
;Filament used: 8.36m 25.0g
;Filament cost: None
;M190 S70 ;Uncomment to add your own bed temperature line
G0 F12000 X-2.699 Y-16.558 Z5.010
G90 ; use absolute coordinates
G92 E0
M82 ; use absolute distances for extrusion
;TYPE:FILL
G1 Z5.010
G1 F2400 E2031.17371
G1 F4800 X-6.491 Y-20.351 E2031.22319

One mistake I make in the code is, G92 E0. I should set it close to E2031.22319 for preventing excessive filament feeding. Above codes will extrude 2031.22319 mm of filament at one spot.

I think correct codes are followed by.

G0 F12000 X-2.699 Y-16.558 Z5.010
G90 ; use absolute coordinates
G92 E2031.2000
M82 ; use absolute distances for extrusion
;TYPE:FILL
G1 Z5.010
G1 F2400 E2031.17371
G1 F4800 X-6.491 Y-20.351 E2031.22319

G92

G92: Set Position
Support FiveD Teacup Sprinter Marlin Repetier Smoothie RepRapFirmware BFB/RapMan MachineKit MakerBot grbl
Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
Parameters
This command can be used without any additional parameters.
Xnnn new X axis position
Ynnn new Y axis position
Znnn new Z axis position
Ennn new extruder position
Example
G92 X10 E90
Allows programming of absolute zero point, by reseting the current position to the values specified. This would set the machine’s X coordinate to 10, and the extrude coordinate to 90. No physical motion will occur.
A G92 without coordinates will reset all axes to zero.

Posted in Reprap 3D Printer | Tagged , , , , , , , , | Leave a comment

New effector design

I’ve been designing new effector for Bulbul Junior.
– Expose only heater and nozzle below the effector plate
– Use steel ball joints
– Use 40x40x10 fan instead stock 30x30x10 fan of E3D V6 Hotend
(The reason is 30x30x10 fan is weak.)
– Dual fan: one heat sink fan, one layer fan
(I learned that layer fan improves PLA printing quality.)

Previous design
round_effector_tilt2

The fan duct assembly is almost finished. I found couple flaws yesterday after I printed a sample.
– Increase diameter of the round effector for 40x40x10 fan about 24 mm
– Increase thickness of the hotend clamp and flanges up to 5 mm
– Remove the mount plate on the fan mount side
– Shorten the height of round duct by 4 mm

Retraction calibration got to be done. But I hasn’t found the right setting yet.

Slic3r removed part of the round shroud when it produced the Gcode file.
Was it because the wall thickness of the shroud is thinner than the default extrusion width (0.42 mm)?

Layer height: 0.3 mm
Default extrusion width: 0.42 mm

cDSC_0004

cDSC_0002

cDSC_0003

Improved fan duct assembly

fan_mount_beta1

fan_mount_beta1_top

fan_mount_beta1_side

Effector plate

round_effector3

Ball seat simulation by OpenSCAD


ball_diameter = 10;
ball_radius = ball_diameter/2;
theta = 90;
sunk_height = ball_radius * sin(theta);
seat_radius = ball_radius * sin(theta);
wall_thickness = 2;
seat_height = 4;

$fn =50;

module ball_seat()
{
difference()
{
cylinder(h=seat_height, r1=ball_radius+wall_thickness, r2=seat_radius+wall_thickness, center=true);
cylinder(h=seat_height+2, r=seat_radius+0.1, center=true);
}
}

translate([0, 0, (ball_radius-sunk_height)/2+ball_radius/2])
# sphere(r=ball_radius, center=true);
ball_seat();

translate([0, 20, 0]) rotate([0, 45, 0]) ball_seat();

90 degree isn’t practical.
ball_seat_demo_90degree

45 degree
ball_seat_demo_45degree

30 degree
ball_seat_demo_30degree

Posted in Reprap 3D Printer | Tagged , , , , | Leave a comment

Journal of 05/19/2015

I dislike concrete but I had no option to build the foundation where the dome stands on due to time and lack of construction knowledge. Thickness of the concrete foundation is 5 cm.

Mixture ratio of making concrete I used.
Crushed stone : sand : cement = 3 (shovels) : 3 : 1.5

The steel geodesic dome
—————————–
five centimeters of concrete – 5 cm
—————————–
one layer of crushed stones – 2 cm to 2.5 cm
—————————–
one layer of plastic – water protection
—————————–
soil

05/09/2015

Water drain pipe installation

cDSC_0081

cDSC_0082

cDSC_0088

Wire mash installation

cDSC_0273

05/10/2015

Putting first layer of concrete

DSC_0293

DSC_0496

DSC_0733

05/11/2015

DSC_0002

DSC_0180

Putting second layer of concrete

DSC_0113

DSC_0427

DSC_0654

05/16/2015

DSC_0661

DSC_0701

DSC_0910

DSC_1061

DSC_1077

05/17/2015

DSC_0704

DSC_1020

DSC_1231

DSC_1241

05/18/2015

DSC_0712

DSC_1155

DSC_1306

05/19/2015

DSC_0662

DSC_0918

DSC_0937

Posted in DIY | Tagged , , , , , , | Leave a comment

Slic3r vs Cura – honeycomb pattern

I use Slic3r v1.1.7 to generate a Gcode file for 3D printing.

round_grid2

I spent last Sunday to get rid of the hairy strings on printed object like this. I thought adjustment of retraction values in Slic3r would solve it – retraction calibration.
cDSC_0035

Well after printing a lot of round hoods, I realized that I was attempting the impossible solution. I had played with different retraction setting which all ended up producing hairy strings.

Retraction length: 5 mm
Lift Z: 1 mm
Speed: 20 m/s
Extra Length on restart: 0 mm
cDSC_0030

Retraction length: 2.8 mm
Lift Z: 1 mm
Speed: 10 m/s
Extra Length on restart: 0.01 mm
cDSC_0035

Retraction length: 2.8 mm
Lift Z: 1 mm
Speed: 50 m/s
Extra Length on restart: 0.02 mm
cDSC_0036

Retraction length: 3 mm
Lift Z: 1 mm
Speed: 100 m/s
Extra Length on restart: 0.05 mm
cDSC_0038

Cura saved the day!
Default retraction setting
cDSC_0040

cDSC_0029

cDSC_0042

After I noticed that Cura produces different pattern of first layer, I carefully watched how Slic3r builds layers of the grid object.

First sweep doesn’t have any honeycombs. cDSC_0020

In second sweep it firms the honeycombs by draws a lot of short vertical strokes between wavy channels. It looks fine but here the real problem of hairy strings. The nozzle does 50 short jumps. Each jump involves one retraction.
Jumps count:
3 (1st row)
5 (2nd row)
6 (3rd row)
7
8
7
6
5
3
cDSC_0022

cDSC_0026

cDSC_0020_eColor green represents the strokes.cDSC_0026_eCura draws a complete honeycomb one at a time which reduces the hairy string caused by short jumps.cDSC_0026_cura

Posted in Reprap 3D Printer | Tagged , , , , , , | Leave a comment

Back to ALSA in Debian 8.0 (jessie)

After few times of upgrade and downgrade of distribution lately, the system lost sound. Reinstallation of pulseaudio package didn’t work out. What would I do? Instead troublshooting pulseaudio, I decided to use ALSA. I barely any help from the official Debian Wiki nowadays. How many users can understand Debian Wiki’s ALSA (https://wiki.debian.org/ALSA) guide?

Source: http://linuxg.net/how-to-properly-replace-pulseaudio-with-alsa-on-crunchbag-linux-and-debian-squeeze

$ sudo killall pulseaudio

Remove all pulseaudio packages:
$ sudo apt-get purge pulseaudio pulseaudio-utils gstreamer0.10-pulseaudio libpulse-browse0 paman pavumeter pavucontrol

$ sudo mv /etc/asound.conf /etc/asound.conf.bak

Delete the pulseaudio directories from the user’s home:

$ rm ~/.pulse-cookie $ rm -r ~/.pulse

Install ALSA packages

$ sudo apt-get install alsa-base alsa-tools alsa-tools-gui alsa-utils alsa-oss alsamixergui libalsaplayer0

Check the sound by running multimedia player such as vlc. Playing flash video on Google Chrome had no sound. I followed the instruction here (http://dallarosa.tumblr.com/post/19626256742/fixing-no-sound-in-browser-problem-on-linux) to fix it.

Posted in GNU/Linux | Tagged , , , , | Leave a comment

Magnetic / spring-loaded ball system – 1

cDSC_0735

cDSC_0736

When the 70.00 mm length of spring is extended to 11.00 mm, it lifts 1,803.00 g of weight.

cDSC_0737

Printing fan hood in 0.05 mm layer height.

cDSC_0742

cDSC_0752

cDSC_0744

Calibration cubes (100.00 mm x 100.00 mm x [20.00 mm to 5.0 mm])

cDSC_0774

A pair of magnetic ball adapter, carriage of linear rail, effector.

cDSC_0772

cDSC_0771

cDSC_0768

cDSC_0766

Rod diameter: 6.00 mm

Magnet: 8.00 mm in diameter, 8.00 mm in height

Ball stud: 12.70 mm in diameter, M5x20 set screw.

How to secure magnet in the adapter? I think I need to modify the adapter shape.

cDSC_0764

Printing two cylinders for the magnetic ball adapter

cDSC_0762

cDSC_0761

cDSC_0759

Carriage of linear rail

cDSC_0754

Kossel Mini spring-string effector and carriage
https://www.youmagine.com/designs/kossel-mini-spring-string-effector-and-carriage

Magnetic Ball Joint Rod Ends
https://www.youmagine.com/designs/magnetic-ball-joint-rod-ends

Posted in Reprap 3D Printer | Tagged , , | Leave a comment

Magnetic / spring loaded ball bearing

Diameter of the ball is 12.70 mm. Diameter of the thread is 5.00 mm. cDSC_0003

cDSC_0001

Steel tube, Outer diameter 6.00 mm, length 230.00 mm
M5x20 set screw x12
Neodymium magnet, diameter 8.00 mm, height 8.00 mm x12
cDSC_0006

cDSC_0004

Posted in Reprap 3D Printer | Leave a comment