Anycubic Mega S: Difference between revisions
m describing more filaments |
|||
| Line 16: | Line 16: | ||
Travel speed at 100 mm/s seemed to help with stringing. | Travel speed at 100 mm/s seemed to help with stringing. | ||
=== General Notes === | ===General Notes=== | ||
==== Unnecessary Travels along Inner Wall ==== | ====Unnecessary Travels along Inner Wall==== | ||
When trying to print a curved wall that is 3 layers thick (1.2mm thick with 0.4mm nozzle), the middle inner wall is broken up into multiple sections by Cura and the nozzle prints each section backwards, causing the nozzle to jutter backwards instead of printing the inner wall as a single smooth line. This seems to be an issue when a curved section is exactly a multiple of the nozzle width and increasing the wall thickness from 1.2mm to 1.3mm seems to fix this issue. | When trying to print a curved wall that is 3 layers thick (1.2mm thick with 0.4mm nozzle), the middle inner wall is broken up into multiple sections by Cura and the nozzle prints each section backwards, causing the nozzle to jutter backwards instead of printing the inner wall as a single smooth line. This seems to be an issue when a curved section is exactly a multiple of the nozzle width and increasing the wall thickness from 1.2mm to 1.3mm seems to fix this issue. | ||
| Line 88: | Line 88: | ||
====Print Issues and Observations==== | ====Print Issues and Observations==== | ||
===== Lung Dragon ===== | |||
Dragon head printed with a raft lifted part way through the print and detached from the bed. Raft was 5mm. Bed was 70C. Tried again with the above profile but with the brim speed set to 15mm/s (it was 40mm/s, possibly causing weak brims?). This sort of worked, but the horns still shifted. This might have been rectified if additional supports were added to join all the parts together since each horn was separate until they're joined a couple millimeters higher. The shifting resulted in some horns being misaligned. The brim appeared to have adhered strongly, but the part still shifted somehow. | Dragon head printed with a raft lifted part way through the print and detached from the bed. Raft was 5mm. Bed was 70C. Tried again with the above profile but with the brim speed set to 15mm/s (it was 40mm/s, possibly causing weak brims?). This sort of worked, but the horns still shifted. This might have been rectified if additional supports were added to join all the parts together since each horn was separate until they're joined a couple millimeters higher. The shifting resulted in some horns being misaligned. The brim appeared to have adhered strongly, but the part still shifted somehow. | ||
| Line 95: | Line 97: | ||
Printing the dragon claw segments is problematic since the small pieces are stuck to the included supports and are printed warped causing the clips to not snap properly. The round joints also turn out to be more of an oval shape and overhangs appear stringy. I tried printing at 30mm/s and it's still looking like crap. | Printing the dragon claw segments is problematic since the small pieces are stuck to the included supports and are printed warped causing the clips to not snap properly. The round joints also turn out to be more of an oval shape and overhangs appear stringy. I tried printing at 30mm/s and it's still looking like crap. | ||
===== Hollow Case (for ozone generator) ===== | |||
I printed a hollow case with the Grey PLA (Mech Solutions) that is 2mm thick with 0.2mm layer height, 205C/65C, 95% flow rate, 120% initial layer flow rate, 50mm/s, 40mm/s outer wall, 55mm/s inner wall, 50mm/s support, 100mm/s travel speed, 6mm retraction, 40mm/s retraction speed, 100% cooling, 45% initial cooling. Without a brim, the hollow case corners lifted from the build plate. One of the corner support disconnected from the base and fell over. There was a lot of stringing occurring between the four corners. | |||
===== Small Container Boxes ===== | |||
I printed many 80x80x80mm boxes at 0.2mm layer height with a 0.4mm nozzle. Wall thickness was 3 walls deep at 1.2mm. I eventually started printing these boxes at 60mm/s, 70mm/s for walls, 50mm/s top/bottom layers, 30mm/s and 75% fan speed for the initial layer. Combing mode is 'Not in Skin' and 'Minimum Wall Flow' at 30% to reduce travel while printing. Prints all exhibited a weird issue where about half way up, a single layer appears to be under extruded. This occurred at the same layer across many prints, at different locations on the bed, and using different PLAs. The fix was to rotate the print 45 degrees while ensuring the seam is still at one of the corners. | |||
===Yellow PLA (Mech Solutions)=== | ===Yellow PLA (Mech Solutions)=== | ||
| Line 102: | Line 110: | ||
Prints similar to the white PLA. | Prints similar to the white PLA. | ||
100% flow rate. 215C / 65C. | |||
=== Red PLA (Mech Solutions) === | |||
This prints without issue, similar to the white and grey PLA. | |||
100% flow rate. 215C / 65C. | |||
=== Green PLA (Mech Solutions) === | |||
This prints without issue, similar to the white, grey, and green PLA. | |||
100% flow rate. 215C / 65C. | |||
Ironing required bumping up flow rate to about 16% before it looked OK | |||
=== Black PLA (Mech Solutions) === | |||
Printing at over 195C results in a rough finish on the walls. When I first printed a test 2.5cm hollow cube at 215C, the walls ended up looking uniformly very rough, almost matte like. Reducing the hot end to 200C almost resulted in a clean print. | |||
100% flow rate. 195C / 65C. | |||
==Firmware== | ==Firmware== | ||
Revision as of 06:25, 13 October 2020
The Anycubic Mega S is a cheap 3D printer. I bought this printer for about $280 CAD with shipping.
Configuration
Serial talks at 250,000 baud.
The print area is 210x210x205 (width, depth, height).
Anycubic suggests that the following slicer settings:
- bottom layer speed not exceed 20mm/s
- top/bottom speed 30mm/s
- outer shell speed 30mm/s
- retraction speed at 60mm/s at 6mm or for flexible filament at 91mm/s at 6.4mm
Cura Profiles
Travel speed at 100 mm/s seemed to help with stringing.
General Notes
Unnecessary Travels along Inner Wall
When trying to print a curved wall that is 3 layers thick (1.2mm thick with 0.4mm nozzle), the middle inner wall is broken up into multiple sections by Cura and the nozzle prints each section backwards, causing the nozzle to jutter backwards instead of printing the inner wall as a single smooth line. This seems to be an issue when a curved section is exactly a multiple of the nozzle width and increasing the wall thickness from 1.2mm to 1.3mm seems to fix this issue.
However, Cura still has unnecessary travel around the perimeter of the initial inner wall. This was rectified by setting the Minimum Wall Flow (under Shell) to 30%. This forces Cura to not travel through printed walls since it must be extruding at 30% flow rate. It still might travel for some sections as it tries to start the next seam in another corner which can be fixed by setting the Z Seam Alignment (Shell) to 'Shortest'.
White PLA (Mech Solutions)
With the enclosure, print profile:
| Setting | Value |
|---|---|
| Layer Height | 0.15 |
| Print Temperature | 200 |
| Printing Temperature Initial Layer | 210 |
| Final Printing Temperature | 185 |
| Build Plate Temperature | 70 |
| Flow | 98% |
| Initial Layer Flow | 125% |
| Print Speed | 50 mm/s |
| Outer Wall Speed | 30 mm/s |
| Top/Bottom Speed | 20 mm/s |
| Support Speed | 30 mm/s |
| Travel Speed | |
| Initial Layer Speed | 15 mm/s |
| Skirt/Brim Speed | 15 mm/s |
| Retraction Distance | 6 mm/s |
| Retraction Speed | 40 mm/s |
| Fan Speed | 100 % |
| Initial Fan Speed | 20 % |
| Regular Fan Speed at Height | 0.5 mm / 4 layer |
Print Issues and Observations
Lung Dragon
Dragon head printed with a raft lifted part way through the print and detached from the bed. Raft was 5mm. Bed was 70C. Tried again with the above profile but with the brim speed set to 15mm/s (it was 40mm/s, possibly causing weak brims?). This sort of worked, but the horns still shifted. This might have been rectified if additional supports were added to join all the parts together since each horn was separate until they're joined a couple millimeters higher. The shifting resulted in some horns being misaligned. The brim appeared to have adhered strongly, but the part still shifted somehow.
Printing dragon segments without support or raft is OK but the initial lip lifts up causing poor print quality on what is to be the underside of the dragon. Printing with custom support holding up the underside of the dragon segment seemed to help; doesn't seem to be such an issue with the larger segments.
The round joint segments don't appear to be completely round.
Printing the dragon claw segments is problematic since the small pieces are stuck to the included supports and are printed warped causing the clips to not snap properly. The round joints also turn out to be more of an oval shape and overhangs appear stringy. I tried printing at 30mm/s and it's still looking like crap.
Hollow Case (for ozone generator)
I printed a hollow case with the Grey PLA (Mech Solutions) that is 2mm thick with 0.2mm layer height, 205C/65C, 95% flow rate, 120% initial layer flow rate, 50mm/s, 40mm/s outer wall, 55mm/s inner wall, 50mm/s support, 100mm/s travel speed, 6mm retraction, 40mm/s retraction speed, 100% cooling, 45% initial cooling. Without a brim, the hollow case corners lifted from the build plate. One of the corner support disconnected from the base and fell over. There was a lot of stringing occurring between the four corners.
Small Container Boxes
I printed many 80x80x80mm boxes at 0.2mm layer height with a 0.4mm nozzle. Wall thickness was 3 walls deep at 1.2mm. I eventually started printing these boxes at 60mm/s, 70mm/s for walls, 50mm/s top/bottom layers, 30mm/s and 75% fan speed for the initial layer. Combing mode is 'Not in Skin' and 'Minimum Wall Flow' at 30% to reduce travel while printing. Prints all exhibited a weird issue where about half way up, a single layer appears to be under extruded. This occurred at the same layer across many prints, at different locations on the bed, and using different PLAs. The fix was to rotate the print 45 degrees while ensuring the seam is still at one of the corners.
Yellow PLA (Mech Solutions)
I had to crank up the flow rate to 210% in order to not get under extrusion. This was after checking the filament size is correct and that the extrusion is calibrated. Using a normal 100% flow rate lead to extreme under extrusion which was not observed with the white PLA. It doesn't appear to be an issue with a plugged nozzle, a skipping extruder, or too low of a temperature. I raised the temperature to 210C with the same result.
Grey PLA (Mech Solutions)
Prints similar to the white PLA.
100% flow rate. 215C / 65C.
Red PLA (Mech Solutions)
This prints without issue, similar to the white and grey PLA.
100% flow rate. 215C / 65C.
Green PLA (Mech Solutions)
This prints without issue, similar to the white, grey, and green PLA.
100% flow rate. 215C / 65C.
Ironing required bumping up flow rate to about 16% before it looked OK
Black PLA (Mech Solutions)
Printing at over 195C results in a rough finish on the walls. When I first printed a test 2.5cm hollow cube at 215C, the walls ended up looking uniformly very rough, almost matte like. Reducing the hot end to 200C almost resulted in a clean print.
100% flow rate. 195C / 65C.
Firmware
The stock firmware can be found on Anycubic's website, near the bottom of the page. https://www.anycubic.com/products/anycubic-i3-mega-s?variant=32094219305091. The stock firmware uses Marlin 1.1.0-RC8. UUID cede2a2f-41a2-4748-9b12-c55c62f367ff.
The TFT display is driven by a separate microprocessor (more information on the displays at https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S/wiki/Types-of-Anycubic-Touchscreens) which communicates via serial with the main board. Updating the firmware will not update the display GUI.
Custom Firmware
There is a custom firmware for the Mega S maintained by knutwurst at https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S. This firmware uses the latest version of Marlin. You can download the pre-compiled release .hex file or obtain the source and compile it in Arduino IDE manually.
Precompiled
If you use the precompiled version, note that there are the following issues/annoyances with knutwurst's precompiled version:
- The firmware still has startup sound. This can only be disabled by editing Marlin/Configuration.h and ensure STARTUP_CHIME is not defined.
- Filament sensor needs to be closed or else it shows an error on the display when the machine is powered on. The special menu to disable the filament sensor is only temporary and reverts on a reset. You will need to undefine the
ANYCUBIC_FILAMENT_RUNOUT_SENSORin Configuration.h to disable this completely. The screen nags you of the filament because this firmware will send J15 to it on startup to the display ifFILAMENT_RUNOUT_PINis high. - Bed size is changed from 210x210 to 225x220 (width, depth). I reverted this back to 210x210 by compiling the firmware manually (see section below) because the nozzle can crash to the clips on the right side of the bed.
Compile Yourself
If you choose to not use the pre-compiled version, compiling a modified version is very easy. I eventually went with this route to address the issues listed above. Clone the repository and open the Marlin/Marlin.ino file in Arduino IDE. Set the target for the ATmega2560 board. Get the compiled binary by going to Sketch -> Export compiled binary. The binary will be located in the same location as the Marlin.ino file. Upload the .hex file that does not contain the bootloader via OctoPrint to update the firmware.
Review the first 100 lines of Configuration.h and uncomment whatever applies. Ensure that the correct device is enabled with the proper configuration flags (eg. KNUTWURST_MEGA_S, KNUTWURST_TMC). You may want to uncomment line 287 so that #define MOTHERBOARD BOARD_TRIGORILLA_14 is defined.
Flashing Firmware
After obtaining the .hex file, to update the firmware via OctoPrint, install the OctoPrint Firmware Updater plugin and configure the plugin with the following settings:
- Flash method: avrdude
- AVR MCU: ATmega2560
- Path to avrdude: /usr/bin/avrdude
- AVR Programmer Type: wiring
- Baud Rate: auto
Mods and Upgrades
- Y Belt Tensioner - https://www.thingiverse.com/thing:2789837
- Motor driver duct - https://www.thingiverse.com/thing:3796022
Filament Holder
Filament Holder - https://www.thingiverse.com/thing:2841440
I installed this filament holder so that the entire printer fits inside a Creality enclosure which was designed for the Ender 3.
Issues
The holder is bending under the weight of the roll of filament inside the warm environment of the Creality enclosure. The enclosure interior might reach temperatures of 40-50C since the Raspberry Pi had a temperature warning at one time during operation.
This is remedied by moving the filament to the opposite side of the arm. The sagging when the filament roll was mounted on the right hand side (above the extruder) might have been a result of lack of leverage with the design. Mounting it on the left side of the arm seemed to have better balance, but the filament is no longer directly above the extruder and as a result rubs against the side of the roll.
TMC2208 v2 Motor Driver
Follow the guide at: https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S/wiki/Swap-stepper-motor-driver-for-TMC2xxx
To install the new TMC2208 drivers, disconnect power and open the unit to remove all the existing drivers. Plug one driver in one by one and apply power. Adjust the potentiometer so that the measured voltage on the Vref pad is 1V. You can do this all at the same time, but due to space constraints, it was easier to do it sequentially.
If you intend to update the firmware to use the TMC module, you can leave the wiring as-is. Otherwise, you will need to reverse the wirings so that the motor directions are reversed.
See Also
- https://www.thingiverse.com/thing:2901190 - Goes over mods for the Anycubic i3 Mega. Some of the slicer settings may be of interest.