Having a lot of trouble  

RE: Having a lot of trouble

The issue with the nozzle crashing and scratching the bed with specific patterns during the bed levelling happened to me a few times. Contacted chat and got the usual advice to recalibrate and reflash. Even after i told them that i already done it 3 times. I send them pics with the scratched beds and they wanted a video. Next day that happened again i opened another chat uploaded a short video and now im still waiting back from prusa to respond

Posted by: @nerdofepic

So far my experience with the semi-assembled 5 tool XL is very similar to the series of utter failures that made me uninstall my MMU2.  I got it a few days ago and built it over a couple days.  The actual build process was quite a bit shorter than the MK4 and aside from a few things that needed a bit of clarity in the instructions it went smoothly.  Since then however, there has been a large collection of problems with it.  I've attempted 4 prints now, only one actually worked.  Here's some of what I've bumped into so far...

• Initial flash process (to firmware 4.7.2) was misleading, appeared to work, but then on restart asked to do it again.  I let it do it all again and it seemed to work the second time too, but didn't ask a third time.
• 2 of my 5 tools reported fan issues.  There was nothing in them, they were in fact spinning as expected.  Trying again had the same result.  Third time too.  Restarted the printer, tried again, reported everything was fine with all 5.
• Went through the entire remainder of the calibration and self test steps for every tool and every part it offered.  All green checkmarks.  (Took almost 1.5hrs.)
• Updated PrusaSlicer to 2.6.1, accepted latest profiles, sliced a 5 color Benchy set up for Prusament PLA and put it on the thumb drive.
• Started the Benchy print from the thumb drive (using 5 different colors of Prusament PLA), wipe tower repeatedly got in the way of the tools, tools left epic strings between Benchy and wipe tower, and between print area and tool docks.  Eventually crashed into the wipe tower and broke it loose.  I wasn't still watching it and came back to rainbow spaghetti and cancelled the remainder of the print.
• All the stringing inspired me to better dial in the nozzle drip stopper positions.  When I got to tool 2, I couldn't adjust it and it was sticking out farther than the other ones.  With the help of support chat, I realized the square nut in the dock-cable-router was spinning in its slot.  (This part was assembled by Prusa and arrived this way.)
• At first I thought the issue was a failure of the printed part, but as it turns out it was actually the nut being cross threaded in the nut. 
• With tool 2 out of commission and the nozzle drip stoppers dialed in on the other tools, I resliced a 4 color Benchy that didn't use tool 2.  I figured I was having issues because of hardened strings going all over the place and this would demonstrate the poorly aligned nozzle drip stoppers would have been the cause of previous issues. 
• Roughly 30 minutes into the 2.5hr print the wipe tower got smacked into, broke loose, and I decided to try to pause and tape it down using the brim.  It failed with a nozzle crash about 5 layers later and I stopped the print instead of trying to resume.
• I printed a replacement dock-cable-router on a different printer and eventually managed to disassemble and reassemble the thing.  (A bit tricky since it came assembled and there aren't publicly available instructions for how to do it.)
• I recalibrated everything for tool 2, including a complete 5 tool calibration pin, and all axis calibrations.
• Then I printed a single color Benchy (using tool 1) and it went reasonably well.  A bit more stringing than I'm used to made me thing the Prusament PLA built in profile might be a bit too hot, but it otherwise seemed fine.
• Since I kept having issues with the wipe tower breaking loose, I decided to give it a wider brim (20mm instead of the default 3mm) which revealed a bug in PrusaSlicer 2.6.1.  When I sliced it with the wider brimmed wipe tower, the path left the area of the print on layer 1 about half way out on the brim and went all the way to the front left of the bed and went back 2 or 3 times.  You can actually see this in the sliced preview, but I have no idea why it's happening.
• With the new 5 color Benchy with the bigger brim wipe tower, starting the print caused the nozzle to crash into the bed during probing, including dragging it on the bed between probe points (not to be petty, but perfect 1st layer is the first selling point of Nextruder and that seems like it should be impossible).  Retried 3 times, triple checked the bed and bed sheet (which are absolutely pristine since I've only tried to print 3 things so far) and redid Z-axis calibration again, even hit the restart button and it kept crashing.  4th time, it probed the front of the bed, no where near where it was going to be printing (I have no idea why, same gcode) and that passed and it started printing.  About 18% in, the wipe tower caused a nozzle crash again and the print failed.

At this point I'm out of ideas.  Parts were assembled incorrectly with a cross threaded screw by Prusa and shipped anyway.  Simple and small PLA print attempts with multiple tools always fail.  The firmware is behaving oddly and differently with the same gcode.  PrusaSlicer is slicing things with paths that run all over the bed for no reason.  Nozzles are crashing into calibrated and level beds with a clean sheet.  Wipe towers never survive even short prints.  I'm out a lot of money for a printer I was very excited about for almost 2 years but that can't things it was advertised to be able to do and I'm quite sad about it. 

I'm willing to help Prusa folks debug my situation or their firmware or slicer software or whatever is needed.

RE: Having a lot of trouble

Thank you for your reply.  I'm not a "misery loves company sort of guy" but it is sort of nice to know it's not just me with some weird anomaly in this case.  I figured printing a small armada of multi-color Benchy to practice with the new printer was a pretty basic test and would probably go better than diving into a 3 day print of a giant dragon or something.  Alas...

To add to the original list:

• I just noticed that the little "dump some filament off the front edge" thing it does on multi-color prints ripped the paint off my textured bed sheet (the surface is fine).  It's most obvious on the tab that says "Original Prusa" it now appears like it someone applied strike through to it, and this is after only 3 prints that did the purge thing on that side of the printer.
• After another nozzle crash (trying the 5 color keychain that came on the thumb drive to remove the PrusaSlicer variable for now) I redid the bed level adjustment step assuming something must have not gone correctly the first time.  Instead of doing cross tightening patterns with the screws when the bed was at the bottom, I decided to focus on the outer screws first (the ones that attach the bed to the A-Frames on the Z-axis) and only once those were retightened did I retighten the ones around the trapezoidal screws.  I think last time I did a sort of all over the place cross tightening pattern that mixed bed and trapezoidal screw tightening and I'm hoping that could have been an issue that I've now fixed.

RE: Having a lot of trouble

The strike through on Prusas logo i thing is something that everyone has (me too). The problem with scratching the bed for me is random. It happens once every 3 or 4 prints. I dont think is a bed thing since when it goes haywire it scratches in a specific pattern (lines and then on a small spiral pattern) which means there is some crazy logic behind it. At the start i though it was the load cell or something but it happened to another extruder and those things are preassmbled. I'm not so sure that it is a hardware issue since everytime it happens i park the tool restart the print and its succesful.

On a happy note i like this printer when it actually prints. Quick in comparison with mmu on multitools and excellent first layer. The profiles need some work. There is a bit of oozing and stringing with the default ones.

I didnt print any of the presliced model that came with the printer (i had only 2 plas -brand new-) . I usually use PETG so i tested it with a 4 color PETG model. And a 2 color sliced Prusa keychain.

Posted by: @nerdofepic

Thank you for your reply.  I'm not a "misery loves company sort of guy" but it is sort of nice to know it's not just me with some weird anomaly in this case.  I figured printing a small armada of multi-color Benchy to practice with the new printer was a pretty basic test and would probably go better than diving into a 3 day print of a giant dragon or something.  Alas...

To add to the original list:

• I just noticed that the little "dump some filament off the front edge" thing it does on multi-color prints ripped the paint off my textured bed sheet (the surface is fine).  It's most obvious on the tab that says "Original Prusa" it now appears like it someone applied strike through to it, and this is after only 3 prints that did the purge thing on that side of the printer.
• After another nozzle crash (trying the 5 color keychain that came on the thumb drive to remove the PrusaSlicer variable for now) I redid the bed level adjustment step assuming something must have not gone correctly the first time.  Instead of doing cross tightening patterns with the screws when the bed was at the bottom, I decided to focus on the outer screws first (the ones that attach the bed to the A-Frames on the Z-axis) and only once those were retightened did I retighten the ones around the trapezoidal screws.  I think last time I did a sort of all over the place cross tightening pattern that mixed bed and trapezoidal screw tightening and I'm hoping that could have been an issue that I've now fixed.

RE: Having a lot of trouble

Been trying things and having prints fail for a bit more time.  For the sake of simplicity, I dropped to a 2 color Benchy (and may drop to a Marvin to save time on my attempts).  I'm up to another 3 or 4 tries with various settings and none have finished. 

Things my research has helped me realize:

• First I realized the wipe tower was being badly configured.  For reasons I don't understand, the "wipe tower purge lines spacing" setting has a default of 100%, but on a fresh reinstall of PrusaSlicer where I deleted all user profile things mine was 150%.  I can't have changed it because I didn't even realize it was a setting that existed until earlier today.  I thought that setting was also affecting the spacing of the lines in the brim, but further research indicates that might actually be true.
• Every single print I did today resulted in something breaking loose from the bed (either the wipe tower or the Benchy depending on which attempt) and being covered in blobs of filament (and lots of strings), which led me to a couple more theories.  I lowered the temperature from 215C to 205C and increased the retraction from 0.7mm to 1.0mm and a lot of the stringing went away (not all).  I went down this hole for a while until I realized that the bottom of the Benchy was wildly inconsistent.
• That's when I realized that I was probably having serious bed leveling issues.  Things not sticking and lines in the first layer being variable width over the course of the object is a pretty good sign of bed issues.  With a lot of experience on the MK3 family of printers, my mind first went to tweaking Live-Z, but with the Nextruder that's not supposed to be a thing.  After a bunch of research on the internet (assisted by a friend of mine) it came to light that filament oozing out of the nozzle prior to bed probing throws off the probed height AND does so inconsistently because each probe can push some extra ooze out of the way or flatten it (and read lower) or have more ooze out (and read higher).
• The only way I can think of to actually fix this bed level / ooze issue is to rewrite the custom gcode with stuff to better clean the nozzle and make sure that happens immediately before the bed probing.  I haven't taken on this task yet, but I also kind of hope future profiles from Prusa include something better, or their firmware does something extra when it encounters the gcode that tells it to probe the bed, or something like that.  If I come up with something useful I'll try to remember to post it here. 
• If anyone has a good guide of some of the custom gcode bits you can use on an XL I'd like a link.  For example, I know basic gcode things (G0 X50 sorts of things), but I haven't ever seen a good list of available things like:  is_extruder_used[0],  nozzle_diameter[0], and first_layer_temperature[initial_tool] with definitions of what they mean (obviously these examples are pretty clear, but without a list I'd just be guessing and pissing off the printer as it has to skip my garbage commands).

RE:

bed leveling issues:

- check that the bed is (mostly) level with the frame (for example drive the extruder around manually and check the nozzle distane to the bed in various places)

- check that the aluminum frame of the heat bed isnt bent (check the print bed flatness with a ruler or whatever straight edged thing you have)

- also make sure the nozzle is clean

- semi-assembled...check that your vertical frame extrusions are fully seated. there may be no gaps between the aluminum end plate and the extrusions the vertical posts are attached to. check the comments on the online manual for these steps. you wouldnt be the first (or 10th) to have issues from not fully seated vertical posts.

the start gcode commands you refer to are slicer internal commands and variables to generate gcode depending on filaments, tools used, and print area. the final gcode file is standard gcode. takes a while to chew through if you want to understand what they do.

i edited mine a bit to add way more retraction when heating the initial tool (which is used for probing), added a wipe cycle before mbl, and some other small changes. working fine for single-tool applications so far, oozing (and those annoying little dots of plastic from mbl) are minimal with prusament petg.

havent tested with more than one used extruder but it should be ok.

;
; printer start gcode start
;
M862.3 P "XL" ; printer model check
M17 ; enable steppers
G90 ; use absolute coordinates
M83 ; extruder relative mode
M221 S100 ; reset flow percentage

; set first layer print area
M555 X{first_layer_print_min[0]} Y{first_layer_print_min[1]} W{(first_layer_print_max[0]) - (first_layer_print_min[0])} H{(first_layer_print_max[1]) - (first_layer_print_min[1])}

; inform about nozzle diameter
{if (is_extruder_used[0])} M862.1 T0 P{nozzle_diameter[0]} {endif}
{if (is_extruder_used[1])} M862.1 T1 P{nozzle_diameter[1]} {endif}
{if (is_extruder_used[2])} M862.1 T2 P{nozzle_diameter[2]} {endif}
{if (is_extruder_used[3])} M862.1 T3 P{nozzle_diameter[3]} {endif}
{if (is_extruder_used[4])} M862.1 T4 P{nozzle_diameter[4]} {endif}

; turn off unused heaters
{if ! is_extruder_used[0]} M104 T0 S0 {endif}
{if ! is_extruder_used[1]} M104 T1 S0 {endif}
{if ! is_extruder_used[2]} M104 T2 S0 {endif}
{if ! is_extruder_used[3]} M104 T3 S0 {endif}
{if ! is_extruder_used[4]} M104 T4 S0 {endif}

M217 Z{max(zhop, 2.0)} ; set toolchange z hop to 2mm, or zhop variable from slicer if higher

; set bed and extruder temp for MBL
M140 S[first_layer_bed_temperature] ; set bed temp
M104 T{initial_tool} S{((filament_type[initial_tool] == "PC" or filament_type[initial_tool] == "PA") ? (first_layer_temperature[initial_tool] - 25) : filament_type[initial_tool] == "FLEX" ? 210 : filament_type[initial_tool] == "PETG" ? (first_layer_temperature[initial_tool] - 65) : filament_type[initial_tool] == "PLA" ? (first_layer_temperature[initial_tool] - 55) : 170)} ; set initial extruder pretemp

; M109 T{initial_tool} S{((filament_type[initial_tool] == "PC" or filament_type[initial_tool] == "PA") ? (first_layer_temperature[initial_tool] - 25) : (filament_type[initial_tool] == "FLEX" ? 210 : 170))} original

G28 XY ; Home XY

; try picking tools used in print. initial tool is skipped
{if (is_extruder_used[0]) and (initial_tool != 0)} T0 S1 L0 {endif}
{if (is_extruder_used[1]) and (initial_tool != 1)} T1 S1 L0 {endif}
{if (is_extruder_used[2]) and (initial_tool != 2)} T2 S1 L0 {endif}
{if (is_extruder_used[3]) and (initial_tool != 3)} T3 S1 L0 {endif}
{if (is_extruder_used[4]) and (initial_tool != 4)} T4 S1 L0 {endif}

; select tool that will be used to home & MBL
T{initial_tool} S1 L0
M84 E ; turn off E motor
G28 Z ; home Z with MBL tool
G0 Y-8 Z40 ; add Z clearance and present nozzle

; extruder pretemp and retraction before cleanup and MBL
M109 T{initial_tool} S{((filament_type[initial_tool] == "PC" or filament_type[initial_tool] == "PA") ? (first_layer_temperature[initial_tool] - 25) : filament_type[initial_tool] == "FLEX" ? 210 : filament_type[initial_tool] == "PETG" ? (first_layer_temperature[initial_tool] - 65) : filament_type[initial_tool] == "PLA" ? (first_layer_temperature[initial_tool] - 55) : 170)} ; set initial extruder pretemp and wait
M302 S150 ; lower cold extrusion limit to 150C
G1 E-60 F2400 ; retract from hotend for nozzle cleanup

M190 S[first_layer_bed_temperature] ; wait for bed temp
G29 G ; absorb heat

; set extruder temps except initial tool
{if (first_layer_temperature[0] > 0) and (is_extruder_used[0]) and (initial_tool != 0)} M104 T0 S{first_layer_temperature[0]} {endif}
{if (first_layer_temperature[1] > 0) and (is_extruder_used[1]) and (initial_tool != 1)} M104 T1 S{first_layer_temperature[1]} {endif}
{if (first_layer_temperature[2] > 0) and (is_extruder_used[2]) and (initial_tool != 2)} M104 T2 S{first_layer_temperature[2]} {endif}
{if (first_layer_temperature[3] > 0) and (is_extruder_used[3]) and (initial_tool != 3)} M104 T3 S{first_layer_temperature[3]} {endif}
{if (first_layer_temperature[4] > 0) and (is_extruder_used[4]) and (initial_tool != 4)} M104 T4 S{first_layer_temperature[4]} {endif}

; nozzle cleaning for MBL
; wipe nozzle
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330)))} Y{(initial_tool < 4 ? -7 : -4.5)} Z8 F{(travel_speed * 60)} ; move close to the sheet's edge
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 10)} Z0.1 F500 ; purge while moving towards the sheet
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40)} F800 ; continue purging and wipe the nozzle
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 3)} Z0.03 F8000 ; wipe, move close to the bed
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 3 * 2)} Z0.2 F8000 ; wipe, move quickly away from the bed
; move to the nozzle touch cleanup area
G1 X{(min(((((first_layer_print_min[0] + first_layer_print_max[0]) / 2) < ((print_bed_min[0] + print_bed_max[0]) / 2)) ? (((first_layer_print_min[1] - 7) < -2) ? 70 : (min(print_bed_max[0], first_layer_print_min[0] + 32) - 32)) : (((first_layer_print_min[1] - 7) < -2) ? 260 : (min(print_bed_max[0], first_layer_print_min[0] + 32) - 32))), first_layer_print_min[0])) + 32} Y{(min((first_layer_print_min[1] - 7), first_layer_print_min[1]))} Z5 F{(travel_speed * 60)}
; nozzle touch cleanup
M84 E ; turn off E motor
G29 P9 X{((((first_layer_print_min[0] + first_layer_print_max[0]) / 2) < ((print_bed_min[0] + print_bed_max[0]) / 2)) ? (((first_layer_print_min[1] - 7) < -2) ? 70 : (min(print_bed_max[0], first_layer_print_min[0] + 32) - 32)) : (((first_layer_print_min[1] - 7) < -2) ? 260 : (min(print_bed_max[0], first_layer_print_min[0] + 32) - 32)))} Y{(first_layer_print_min[1] - 7)} W32 H7
G0 Z10 F480 ; move away in Z
M107 ; turn off the fan

; MBL
M84 E ; turn off E motor
G29 P1 ; invalidate mbl & probe print area
G29 P1 X30 Y0 W{(((is_extruder_used[4]) or ((is_extruder_used[3]) or (is_extruder_used[2]))) ? "300" : ((is_extruder_used[1]) ? "130" : "50"))} H20 C ; probe near purge place
G29 P3.2 ; interpolate mbl probes
G29 P3.13 ; extrapolate mbl outside probe area
G29 A ; activate mbl
G1 Z20 F720 ; move away in Z

; set extruder temp for initial tool
M104 T0 S{first_layer_temperature[initial_tool]}

; P0 S1 L1 D0 ; park the tool


; purge first tool if it is not the initial tool
{if (is_extruder_used[0]) and initial_tool != 0}
P0 S1 L1 D0 ; park current tool
M109 T0 S{first_layer_temperature[0]} ; set and wait for extruder temperature
T0 S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(0 == 0 ? 30 : (0 == 1 ? 150 : (0 == 2 ? 210 : 330)))} Y{(0 < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E10 X40 Z0.2 F500 ; purge while moving towards the sheet
G0 X70 E9 F800 ; continue purging and wipe the nozzle
G0 X{70 + 3} Z0.05 F8000 ; wipe, move close to the bed
G0 X{70 + 3 * 2} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[0]} F2400 ; retract
{e_retracted[0] = 1.5 * retract_length[0]} ; update slicer internal retract variable
G92 E0 ; reset extruder position
G1 Z10 F720 ; move away
M104 S{(is_nil(idle_temperature[0]) ? (first_layer_temperature[0] + standby_temperature_delta) : (idle_temperature[0]))} T0
P0 S1 L1 D0 ; park the tool
{endif}


; purge second tool if it is not the initial tool
{if (is_extruder_used[1]) and initial_tool != 1}
P0 S1 L1 D0 ; park current tool
M109 T1 S{first_layer_temperature[1]} ; set and wait for extruder temperature
T1 S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(1 == 0 ? 30 : (1 == 1 ? 150 : (1 == 2 ? 210 : 330)))} Y{(1 < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E10 X140 Z0.2 F500 ; purge while moving towards the sheet
G0 X110 E9 F800 ; continue purging and wipe the nozzle
G0 X{110 - 3} Z0.05 F8000 ; wipe, move close to the bed
G0 X{110 - 3 * 2} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[1]} F2400 ; retract
{e_retracted[1] = 1.5 * retract_length[1]} ; update slicer internal retract variable
G92 E0 ; reset extruder position
G1 Z10 F720 ; move away
M104 S{(is_nil(idle_temperature[1]) ? (first_layer_temperature[1] + standby_temperature_delta) : (idle_temperature[1]))} T1
P0 S1 L1 D0 ; park the tool
{endif}


; purge third tool if it is not the initial tool
{if (is_extruder_used[2]) and initial_tool != 2}
P0 S1 L1 D0 ; park current tool
M109 T2 S{first_layer_temperature[2]} ; set and wait for extruder temperature
T2 S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(2 == 0 ? 30 : (2 == 1 ? 150 : (2 == 2 ? 210 : 330)))} Y{(2 < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E10 X220 Z0.2 F500 ; purge while moving towards the sheet
G0 X250 E9 F800 ; continue purging and wipe the nozzle
G0 X{250 + 3} Z0.05 F8000 ; wipe, move close to the bed
G0 X{250 + 3 * 2} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[2]} F2400 ; retract
{e_retracted[2] = 1.5 * retract_length[2]} ; update slicer internal retract variable
G92 E0 ; reset extruder position
G1 Z10 F720 ; move away
M104 S{(is_nil(idle_temperature[2]) ? (first_layer_temperature[2] + standby_temperature_delta) : (idle_temperature[2]))} T2
P0 S1 L1 D0 ; park the tool
{endif}


; purge fourth tool if it is not the initial tool
{if (is_extruder_used[3]) and initial_tool != 3}
P0 S1 L1 D0 ; park current tool
M109 T3 S{first_layer_temperature[3]} ; set and wait for extruder temperature
T3 S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(3 == 0 ? 30 : (3 == 1 ? 150 : (3 == 2 ? 210 : 330)))} Y{(3 < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E10 X320 Z0.2 F500 ; purge while moving towards the sheet
G0 X290 E9 F800 ; continue purging and wipe the nozzle
G0 X{290 - 3} Z0.05 F8000 ; wipe, move close to the bed
G0 X{290 - 3 * 2} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[3]} F2400 ; retract
{e_retracted[3] = 1.5 * retract_length[3]} ; update slicer internal retract variable
G92 E0 ; reset extruder position
G1 Z10 F720 ; move away
M104 S{(is_nil(idle_temperature[3]) ? (first_layer_temperature[3] + standby_temperature_delta) : (idle_temperature[3]))} T3
P0 S1 L1 D0 ; park the tool
{endif}


; purge fifth tool if it is not the initial tool
{if (is_extruder_used[4]) and initial_tool != 4}
P0 S1 L1 D0 ; park current tool
M109 T4 S{first_layer_temperature[4]} ; set and wait for extruder temperature
T4 S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(4 == 0 ? 30 : (4 == 1 ? 150 : (4 == 2 ? 210 : 330)))} Y{(4 < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E10 X320 Z0.2 F500 ; purge while moving towards the sheet
G0 X290 E9 F800 ; continue purging and wipe the nozzle
G0 X{290 - 3} Z0.05 F8000 ; wipe, move close to the bed
G0 X{290 - 3 * 2} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[4]} F2400 ; retract
{e_retracted[4] = 1.5 * retract_length[4]} ; update slicer internal retract variable
G92 E0 ; reset extruder position
G1 Z10 F720 ; move away
M104 S{(is_nil(idle_temperature[4]) ? (first_layer_temperature[4] + standby_temperature_delta) : (idle_temperature[4]))} T4
P0 S1 L1 D0 ; park the tool
{endif}


; purge initial tool
M109 T{initial_tool} S{first_layer_temperature[initial_tool]} ; set and wait for extruder temperature
T{initial_tool} S1 L0 D0 ; pick the tool
G92 E0 ; reset extruder position
G0 Y-8 F8000 ; move over the edge of the sheet
G1 Z10 F720 ; move to the Z ready for purge
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330)))} Y{(initial_tool < 4 ? -7 : -4.5)} F{(travel_speed * 60)} ; move close to the sheet's edge
G0 E61 F400 ; de-retract into hotend after MBL
G0 E10 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 10)} Z0.2 F500 ; purge while moving towards the sheet
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40)} E9 F800 ; continue purging and wipe the nozzle
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 3)} Z0.05 F8000 ; wipe, move close to the bed
G0 X{(initial_tool == 0 ? 30 : (initial_tool == 1 ? 150 : (initial_tool == 2 ? 210 : 330))) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 40) + ((initial_tool == 0 or initial_tool == 2 ? 1 : -1) * 3 * 2)} Z0.2 F8000 ; wipe, move quickly away from the bed
G1 E{- 1.5 * retract_length[initial_tool]} F2400 ; retract
{e_retracted[initial_tool] = 1.5 * retract_length[initial_tool]}
G92 E0 ; reset extruder position
;
; printer start gcode end
;

RE: Having a lot of trouble

I don't know if this will help.

On my 2 head I had some similar issues. Like the head crashing into the wipe tower and the printer homing during long prints. I found two things which seemed to contribute to this.

1. The grub screws on the gears on the x/y steppers were loose. The left one was loose enough that I heard clicking when the head would change direction. I believe fixing this resolved most of my issues. Check yours.
2. When I first set my printer up it was on a cheap folding table, this table was not the most level thing. It was so warped that one of the 4 feet on the printer would either float in the air, or cause the system to wobble. I suspect that when the head moved back and forth that this would cause the frame to warp very sightly and with the tolerances involved it would throw the self leveling out of wack, which resulted into the head crashing into the wipe tower. Moving it to a sturdy level table also seemed to resolve this.

I hope my experiences help.

RE: Having a lot of trouble

I've tried several more things since my last post, and have a few things to report.

• Production 4.7.2 firmware is installed and was used for this entire process.  No alpha or RC firmware was used or harmed in this story.
• First, with the theory that I might have goofed up my vertical extrusion mounting, I want back after those.  All tight as a drum and I tried to squeeze a post-it into the line between the vertical and horizontal extrusions, and I couldn't get any of the paper to even remotely enter the crack.  As far as I can tell, I have correctly positioned and tightened extrusions.  (I really wanted this to be it.  There are even comments to this effect in the instruction manual however, and reading those made me extra careful during assembly apparently.)
• I redid the ENTIRE self check and calibration process (again, maybe 3rd time's the charm?) and double checked the tightness of the dock mount screws before I started.  Nothing was obviously incorrect.  During this process I selected the option to heat up all tools prior to the calibration pin step and made sure every nozzle was spotless.  (I've done so little printing with the XL that I couldn't imagine them being anything but clean, but I even made sure there wasn't even slight discolorations to be extra sure.)
• I took the advice to check the XY motors' grub screws and found all 4 to be fully tightened.  (Also wanted this to be the issue because who doesn't love an easy fix.)
• Because of what I describe below, I have NOT tried your custom start gcode from above yet.
• I reinstalled PrusaSlicer 2.6.1 (again) because I was having a weird situation where everything was behaving super slowly while doing almost anything.  I could watch the filament drop boxes (for the 5 tools) flash as they redrew over and over.  I also couldn't reasonably scroll the whole right properties panel after a slice.  I was trying to view the time and volume stats at the bottom and it was going so slowly with the mouse wheel that I grabbed the scroll bar and yoinked that to the bottom, but that redraw took a weirdly long time (multiple seconds) too.  After ripping out PrusaSlicer, and manually deleting the directory it used to live in to make sure even custom profiles were gone, the issue went away.
• It should be noted that the printer ONLY appears to support FAT32 file systems, and thus only smaller thumb drives that will become harder and harder to find as time goes by.  (Consider support for exFAT?)  This means larger thumb drives that require exFAT will not work.
• I would like to point out though that auto-selecting a print to do is not as good a user experience as simply putting me on the directory listing because essentially 100% of the time the printer guesses wrong which one I want to do (presumably using most recent file timestamp), but it doesn't appear to search in directories.  That means that even on the Prusa supplied thumb drive, putting a new file in the 5-tool directory will default to the keychain in the root which means I have to back out of that view to the directory list every time anyway.
• At this point it was time to start printing things...
• First, I printed the single color Benchy which worked pretty much exactly like the first time.  It grabbed tool 1 like it was supposed to, printed a Benchy which looked fine to me (slightly less smooth because I'm used to 0.4mm instead of 0.6mm nozzles, but that was expected).
• Next, I attempted a 2 color Benchy which finished, but was a horribly stringy blobby mess.  Not even close to a useful print.  Cleary all the double checking, tightening things, fine tuning, recalibrating, and self checking didn't result in any improvement in print quality.  (But it did at least finish.)
• Attacking the stringy blobby angle, I tried the custom gcode on tool change from that video with the dogs.  I printed a 2 color Benchy and it was only a minor improvement over not having that gcode, but it did finish.
• Because I'm a glutton for punishment I think, I tried a 200% 5 color Benchy with the same dog video stuff and it failed in a new and interesting way.  It was a roughly 9hr print that stopped at around the 4hr point.  Note I said stopped, not failed.  The printer happily reported "Printing..." but it clearly was not.
• Multiple attempts to resume would come over tap the wipe tower and return to the corner.  It would go immediately back to the "Resume" option in the interface.  With no clue about what was happening I pulled the print off the bed and figured I'd try again after grabbing a crash dump.
• Dumping the crash log failed with an all orange screen saying it needed a thumb drive to put the dump on.  This was strange because the thumb drive was still in the printer from the now failed print, and the icon for the thumb drive was still visible in the corner of the screen.  I pulled the thumb drive out, put it back in, and tried again.  The dump saved successfully.
• This means to me that there is a rather critical firmware error that loses the ability to communicate with USB storage, but that has no idea it has done so.  That the printer gave me no clues about what was going on (and in fact gave me clues like the USB icon to the contrary) is proof that something needs fixing.   To be clear, there is no WiFi, PrusaLink, PrusaConnect, or anything else involved.  Purely local USB thumbdrive.
• Instead of giving up when no USB data is available, perhaps reset the USB subsystem and try again in software since it is highly unlikely that the physical connection to the thumb drive has failed, and/or put up a message explaining that USB isn't working anymore, and/or take down the USB icon?
• As a follow up to this, I tried the same 5 color Benchy print again and the same thumb drive stopping thing happened on this try too (but this time at roughly 6hrs in instead of 4hrs, which indicates it was not an issue with the drive or the gcode file I was printing).  Before I started I turned the printer completely off to make sure it was in the cleanest default state.  This time because I had a good guess at what was happening, I pulled the USB and stuck it back in.  I hacked off the goo ball that resulted in this delay with flush cutters and it's still printing now.  It is however still a stringy blobby mess.

Sorry for the response lag, it's been a long few days of trying to figure things out and trying print after print.

RE:

What USB thumb drive are you using? It might not like it. I am using SanDisk Ultra thumb drives and the printer seems to like them fine. Was it a new drive or a older drive? Just a thought.

RE: Having a lot of trouble

I have a rather embarrassing number of thumb drives, almost all of them are SanDisk (a fair few are ultras).  The one I'm using right now is not an ultra, but was brand new from its package.  Had to delete that random SanDisk software off it before I started loading it up with gcode.

I also found that it's pretty intermittent.  I was strongly considering searching GitHub issues for similar complaints and making a bug ticket if I couldn't find one but I have no reasonable reproduction steps.  "Just keep printing a lot trying to make things that look like they came off a $4000 printer.  Eventually it might crap out a few times."  Not super helpful to the firmware devs unfortunately.  Then again, if no one has good steps and a bunch of people are seeing issues like this and no one makes the bug, Prusa doesn't even reasonably get a shot at fixing it.

RE: Having a lot of trouble

A bit of googling shows that someone else saw the USB issue too it seems.  https://www.reddit.com/r/prusa3d/comments/17a61yq/prusaxl_print_stopped_midprint_with_no_errors/

I laughed when I realized that person ALSO didn't appear to make a ticket.

RE: Having a lot of trouble

I've printed the same gcode twice and had the issue happen both times, but at very different times during the print.  I changed thumb drives, tried a different gcode and it still happened.  After searching github, I didn't find anything mentioning the thumb drive problem so I made a new bug ticket. 

https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3378

RE: Having a lot of trouble

So I sliced up a 6 color model with a single color change (M600) on tool 2.  PrusaSlicer supported this well by asking which tool was being changed and even asked for a new color for that tool so it could render the sliced model in the correct colors.  Very happy about that.  Then I loaded up the gcode and started the print. 

But...

• The color swap came command came and went and the printer never even slowed down for it.  
• Fortunately, it was a large print (14hrs) and I could swap the roll on tool 2 manually before it needed the new color for that tool.
• Except the tune menu option for "Change Filament" only works for the current tool without allowing you to switch tools.
• So I had to wait and watch the print until it grabbed tool 2 on its own to start using it in the new color (which was over 5 hours after the last time it used tool 2).  Only once it finally got back to using tool 2 again could I manually swap the rolls to try to save the print.
• I managed to pause it as it was heading to the wipe tower with tool 2, but it definitely shouldn't have been that dramatic because it should have just stopped for a color change on its own as the gcode requested.

This is definitely a firmware bug based on the slicer doing all the right things.  Sorry I can't attach the gcode for this print here because it was ~66.5 MB.

RE:

Learned more things...

• USB issues blow up prints in dangerous ways.  I woke up and went to the printer to discover it had been at full hot print temperature with the nozzle parked over the print for the prior 8 hours.
• The symptoms of USB failure and color change issues show similar behavior sometimes, so it's hard to tell them apart.  In fairness I have only tried a couple prints with a color change compared to the dozens of other prints now, but both a M601 and a USB failure will present a resume option on the LCD, and the USB version will try to resume and then immediately fail again without indication of why.  Pulling out the USB and putting it back in and then resuming again will fix the USB issue if that was the actual cause.
• I also chased down a bug with a friend.  We found that in PrusaSlicer 2.6.1 there is an issue that also demonstrates a bug in firmware 4.7.2.  We've discovered that PrusaSlicer does sometimes put out an M600 for a color change, and other times an M601/M117 pair.
• With multiple tools, it's reasonable to want to change colors for a single tool in an area of the print that is between two heights that use that tool.  For example, tool 4 is used for the first 25 layers, and again starting at layer 50, it makes sense to request a color change at something like layer 37 so it's between the two places it is used.  When this is done, PrusaSlicer puts out an M601/M117 pair (in that order).  It seems logical to me that those be put out in the other order so the message can be shown and then the pause can happen without needing to read ahead for a message or something similar. 
• The firmware pauses when it reaches the M601 (before it gets to the message, but the firmware doesn't appear to show messages anyway).  When paused, you can't select a different tool to unload filament and reload it with a new color, you're stuck with whatever tool it happened to be on (which can be somewhat random because even though tool 4 is not used, tools 1, 2, and 5 could be in use on the layer you set up the color change for).
• When the tool you want to change colors for IS in use on the layer you selected, the slicer puts out a M600 instead, which behaves correctly in the firmware as far as I can tell. 
• The situation of an unused tool having its color changed requires a longer gcode process something like:  1. switch to the tool that is going to be swapped, 2. unload the filament, 3. reload the new filament, 4. switch back to whatever tool was needed next.

On the note of USB failures, I'm starting to think I may have a defective LCD or mainboard at this point.  I've now used 4 different thumb drives (including the Prusa branded one that came with the printer) and all of them work perfectly in my laptop and SEEM to work properly to start a print, but then all of them have lost the ability to be used during a print by the printer.  Since I've pretty solidly ruled out actual thumb drive issues now, and the laptop is fine with all of them, the only source of problem is now on the printer.  That could be a fault in the firmware, but it is also possibly a hardware issue for something that talks to the thumb drive within the printer.

RE: Having a lot of trouble

And now it can't grab the tool from dock 1...  You can't make this stuff up.  I'm so frustrated at this point that I looked up the Prusa Return Policy and might use it if things don't rapidly improve.

• I've triple checked dock 1.  Nothing has moved, nothing is loose, nothing has filament deposits.  The dock is still firmly mounted and doesn't wiggle and hasn't shifted.  The dock pins are still solidly where they go.  The side bolts (with the 5.5mm hex heads) are still tight and haven't moved.
• An attempt to grab the tool fails to engage the locking pins, claims to crash into something, and shows an angry orange screen.  I have watched it several times and can't see it doing anything weird.  Restarting the printer doesn't correct the situation.

RE: Having a lot of trouble

I took the time and scanned all of your posts. I notice( please correct me if I’m wrong on this), that you never mentioned selecting a Preset “Print Setting” from the pulldown menu in the Plate view. I suspect there’s a bug in PS which messes up all the printer and print settings unless you reset them all manually by selecting a preset. If I am correct, this explains a lot of the weird behavior you are seeing.

Try this… open a small stl file such as a 20mm cal cube in PS. Select the unmodified Xl System Preset for the printer, then go to the print setting pull down menu in the plate view and select the unmodified setting for “Speed” and tell us what it does.

RE: Having a lot of trouble

I'm not exactly sure what you're referring to here because I've reinstalled things from absolute scratch several times (including both 2.6.1 as a proper install, and 2.7.0-alpha1 as a portable zip install) and nuked their various left over directories where profile changes can live from orbit in the uninstall process.  In all cases I've at least started with whatever profiles are pre-canned and latest from Prusa, and since switching to alpha at the recommendation of a Prusa chat person, I haven't modified anything.  All just trying to get a print that goes more than half way without some issue. 

Attempting to follow along with your suggestion, here's what I did:  With 2.7.0-alpha1, the printer drop box in plate view is on "Original Prusa XL - 5T 0.6 nozzle" (the normal one, not the IS one, no modifications).  All 5 filaments are on "Prusament PLA" with no modifications.  Selecting "0.25mm SPEED" from print settings only caused the wipe tower to move closer to the middle of the bed than it was after I loaded the cube and it moved it to the back of the bed.

For full disclosure, I also spent over 12 hours printing a set of 5 full sheet single layer prints yesterday (once each for each of the 5 tools).  The printer probed the bed with a spotless clean nozzle 144 times to build up its mesh data and then printed the first (and only) layer with blob shaped variances all over.  To be clear there were NOT oozy filament blobs involved!  I just mean the thin/high parts of the sheet were sorta blobby shaped.  Some parts of each print were what I would call perfect, other parts were what I would call too high.  The sales pitch of "The XL achieves a perfect first layer across the entire surface every time you start a print." is absolutely inaccurate at best.  And since the vast majority of my print attempts have either the wipe tower or the object break loose from the bed if I put them in the areas I can now describe as the bed blobs (such as along the back edge) now that I've done this experiment, I would argue it's worse than my description is calling it.  I imagine sometimes this is warping off the bed due to bad adhesion in the high spots, but also there are definitely stringy blobs that the nozzle will whack into and break something loose when it does.  (Sometimes realizing that is a crash, other times it doesn't even know something bad happened.)

I have been keeping a bag of my print attempts with the XL.  It now has 27 prints in it (excluding the 5 full sheet prints) that I've done over the last couple weeks.  Literally two have worked properly all the way (both were 1 tool prints).  Some were reasonably close, some required literal painters tape mid-print and babysitting to get anywhere near done, a lot of them had the USB system fail mid-print (one time it happened twice in the same print), but far and away the biggest issues are obvious failures at bed levelling.  When you flip over a failed print and look at the bottom, half of the first layer looks great and the other half is high enough up off the bed that it practically looks like it was done on a different printer entirely.  What it's doing when it's probing the bed with a clean nozzle is beyond me.  (Please ask me how much I wish "heat all nozzles to some temp" was an option in the temperature menu to make cleaning them easier.)  I don't know if I have faulty load cells or what but it's clear that the mesh data is not doing what it should be.  I have triple checked my bed, busted out a level to make sure it actually is, and the variances I'm seeing are way too small for normal tightening of screws to cause.  Micro-adjustments in the z-motors not small enough?  Not accurate enough?  Firmware applying the mesh data incorrectly?  There are tons of theories and I've tried everything I can think of to fix it.  Such as cleaning the steel sheet on both sides, redoing the Z calibration, making sure all of the nozzles are spotless before starting a print, redoing the "auto-home" thing, upgrading the firmware, not sure what's left.

I really thought it was reasonable to expect a printer delayed for way more than a year to be able to select a model, slice it up, and just have it print successfully.  Perhaps I've been spoiled by my MK3 and MK4, but the XL has proven at best to be an essentially non-stop source of frustration and at worst maybe a conspiracy to sell filament since it's only presently capable of wasting it.  And I get it, the MK3 and MK4 have a trillion hours of use in the Prusa print farm to iron out all the kinks and the XL simply doesn't, but for the price I really was hoping for some 5 color prints that "simply worked."

RE:

Ok…

next…

edit the pla preset and put the temp at 200c. Prusa’s presets are too hot.

edit the print setting pulldown so its 0.32 Speed not the .25.  I assume you have the .6mm nozzle.

Disable the wipe tower.

then go to the single extruder you’ll be printing with in Printer Settings and set the “Retract when tool disabled” to 11mm to retract out of the melt zone. PLA behaves like ice. When it hits its glass point it goes completely liquid right away and vice versa when cooling. There’s no soft transition. As such if it’s in the melt zone it will drip down and touch the nozzle seal from which it will draw a string when moved away. It could also drip out a blob. However all of this really only matters when using more than one extruder. But note this anyway.

Before you print and before the bed heats up wipe it with some alcohol to ensure a good stick. I gave up and started using Aquanet on it.

RE: Having a lot of trouble

In my various attempts I've lowered the temp by 5C at a time all the way down to 195C.  It barely helped the stringing.  I increased the retraction (normal retraction that is) from 0.7mm to 2.0mm.  That also barely helped.  As mentioned above, I used the custom tool gcode from the dog video which also slightly helped.  None of these things should affect the first layer though since it's going to prep the wipe tower and then dive right into the first layer.  Not much chance to have the PLA drip out between those times.  I have gallons of 99% IPA that I use before literally every print.  The bed is spotless.

You mentioned disabling the wipe tower, but this one leaves me a little confused.  I understand the uber retraction approach, but is that really enough to skip the wipe tower entirely?  The real question at the moment though is why is mesh bed levelling with a 12x12 grid not allowing the printer to accommodate the tiniest imperfections?

RE: Having a lot of trouble

I’m not referring to regular retraction I’m talking about retracting out of the melt zone while the extruder is parked. Please look for “Retraction when tool disabled” in each extruder setting in the Printer Settings view tab.

but let’s slam on the breaks and start all over. There’s a lot of variables that complicate it if you do anything different.

Please do a 20mm cal cube using a single extruder using preset Print settings. Select .32 Speed please.. Let’s see what the printer can do using a single extruder. When you go multiple things get very complicated. Keep it simple until you know each extruder is working optimally. If you get a good clean print then do it again on the next tool and so on. First verify that each tool prints ok.