Hi, I'm having the same problem, except I'm stuck in a reconnecting / /connected / disconnected loop. Refreshing or closing firefox , and deleting all sites data , doesn't work.
It happens after a somehow long operating time (~ 2 weeks)
Irecently upgraded everything to 3.5.3
Hello everyone
In resurrect.g generated after a power failure something looks weird with the way z height is handled.
Example from a few hours back (but happened a few times before): powerloss during a print at Z22.380 while babystepping is -0.10.
Resurrect.g has G92 Z22.380 in the beginning , then sets babystepping back to -.10, then goes to last print move at 22.280 .
Last time i had a hint about this happening, different print, babystepping was set to .30 when power failure occured, I didn't take a look at the details in resurrect this time but the resuming layer was exactly .3 too high and i had to set babystepping back to 0 to compensate.
Let me know if you need more info
Best
@chrishamm Thank you very much I'll stick with the G4 in the macro then. Will report if the event occurs
Hi everyone,
I have a question : for my specific porject, I have a macro on every layer change that calculates a dwell timing for the print to take a specific time to complete. I was thinking of a scenario: what happens if the printer is shut down (either accidentally or intentionaly) during the G4 command executed from a macro, called from the print ? Will the frimware be able to generate a proper resurrect.g with the correct parameters ?
If no: I was thinking of three things to mitigate this:
-add a dwell right after the macro call in the print file. Tricky, I have a file currently being printed that will go on for one year, i.e. I can't modifiy the original file. Do we have a way to "dynamically" add gcode to the file being printed ?
-call the dwell from deamon.g using a global, we would be out of a macro, would that pause the print ? Or just the deamon? (can't test right now, printer has been shut off and I won't be able to run such tests confortably as i'll be handling it remotely)
-call M27 and M114 from the macro so I can make a custom reusrrect.g . Will their output get written in the eventlog ? If not can I write the output to a file ?
Thank you all in advance
Best
@JayT In most slicer you can choose where the seam is located. The seam is where an external perimeter starts and end, it usually creates an irregularity at that point
Prusaslicer
Hi eveyone here is an update on this. So I'm stil lnot able to have Prusa's macro to work, in the meantime I did two things , one is working the second remains to be tested. I needed to implement the first solution in deamon.g to accomodate an ongoing long print (2weeks) that didn't have a single macro called on tool changes.
This assumes that succesive layer times have little variations (successive tool work times are approximately similar) which is the case for me with variations from layer to layer being a couple tens of seceonds at max . The thing is, every tool is going to be preheated regardless of their future usage, which is OK in my case as both tools are used until the last layers
Here is what i have in deamon.g
if state.status == "processing"
if exists(global.previousTool)
if state.currentTool != global.previousTool ; executes right after tool change
G10 P{abs(state.currentTool - 1)} R160
if exists(global.newToolchangeTime) ; intitialize previous tool change time with last know tool change time
if !exists(global.previousToolchangeTime)
global previousToolchangeTime = global.newToolchangeTime
else
set global.previousToolchangeTime = global.newToolchangeTime
if !exists(global.newToolchangeTime) ; update tool change time
global newToolchangeTime = state.time
else
set global.newToolchangeTime = state.time
echo "Last Toolchange time was :", global.previousToolchangeTime ; see if it works as intended
echo "New Toolchange time is :", global.newToolchangeTime
if exists(global.newToolchangeTime) && exists(global.previousToolchangeTime)
echo "Calculating time between last tool changes..."
if state.currentTool == 0
if !exists(global.lasttoolworkduration0)
global lasttoolworkduration1 = global.newToolchangeTime - global.previousToolchangeTime
else
set global.lasttoolworkduration1 = global.newToolchangeTime - global.previousToolchangeTime
if state.currentTool == 1
if !exists(global.lasttoolworkduration1)
global lasttoolworkduration0 = global.newToolchangeTime - global.previousToolchangeTime
else
set global.lasttoolworkduration0 = global.newToolchangeTime - global.previousToolchangeTime
if exists(global.lasttoolworkduration0)
echo "It took", global.lasttoolworkduration1, "s between two tool changes from 1 to 0"
else
echo "Not enough tool changes were sampled"
if exists(global.lasttoolworkduration1)
echo "It took", global.lasttoolworkduration0, "s between two tool changes from 0 to 1"
else
echo "Not enough tool changes were sampled"
if !exists(global.previousTool)
global previousTool = state.currentTool
else
set global.previousTool = state.currentTool
if !exists(global.preheatingoffset)
global preheatingoffset = 40
else
set global.preheatingoffset = 40 ; start heating up 60 seconds before next layer
if exists(global.previousLayerTime) && exists(global.lasttoolworkduration0) && exists(global.lasttoolworkduration1)
if state.currentTool == 1 && state.time >= global.newToolchangeTime + global.lasttoolworkduration1 - global.preheatingoffset
echo "Preheating extruder ", abs(state.currentTool - 1), "..."
G10 P{abs(state.currentTool - 1)} R{heat.heaters[abs(state.currentTool -1) + 1].active}
;G10 P{state.currentTool} R150
elif state.currentTool == 0 && state.time >= global.newToolchangeTime + global.lasttoolworkduration0 - global.preheatingoffset
echo "Preheating extruder ", abs(state.currentTool - 1), "..."
G10 P{abs(state.currentTool - 1)} R{heat.heaters[abs(state.currentTool -1) + 1].active}
;G10 P{state.currentTool} R150
Now the goal, and second solution will be to execute a macro called preheatextruders.g in prusaslicer tool change custom gcode that will handle the Time globals
and let deamon.g do the regular checks
It's DIY and has a rather narrow scope i know but I hope it can help others in similar situations
So I should have looked closer at the gcode before but it's actually quite messy so there could be this
Anyway I setup a wipe tower for both extruders. Created a 40mm wide, .4mm thick pipe, sliced it in the place i want the wiping to happen with the two extruders separately, extracted the gcode for 1 layer for each, deleting the G1 Zx.xx instructions, and created a macro called wipetower.g executed on every tool changes. In the macro depending on the tool selected it will extrude the corresponding circles (2 lines wide for .2mm nozzle and 1 for .4mm) at the designated place.
Will test and report
I also redesigned the cooling system that i was not really happy with, should be much better
I'll definitely look into having smoother walls for the organic support but I don't think we have much control over this...
@o_lampe said in Troubleshooting print quality issue - frosty effect:
I don't see a reason to add extra PA, because you have PTF tube before the hotend. Unless you also have an inline extruder pushing filament into this tube...
My advice would be to purge after toolchanges. Even when you retract the parked tool 10mm, some filament remains in the nozzle and overcooks.
It's either
- active cooling of the unused hotend (watercooled heaterblocks are rare) or
- setup a purge tower in Prusa-slicer (which isn't easy with different nozzle diameters)
- define your own purge corner in the toolchange macros
My thought process was that the added resistance of the long ptfe tubing was adding springiness to the whole extrudion process, a lag. I'll still test it anyway and not just rely on predefined values to be 100% sure.
Active cooling of the hot end: can't do
Purge tower in prusa with two different nozzles is not possible in prusa slicer options indeed. I could set up something in the tool change gcode section (i like the idea that it is embedded in the gcode and doesn't interfere with the purging and tool changes of the resurrect , start, etc... like printing 2 concentric circles in some place of the bed at every tool change. Shouldn't be to hard. Is it what you had in mind ?
Purge corner seems doable too, I'm not familiar with the technical solutions for this but will look into it.
Thank you for your help!
@T3P3Tony Actually i didn't test the PA setting, just used the recommended value for direct drive extruders but that is a mistake as I have this long tubing setup and specific nozzles for each extruder.
So will try testing this after everything else seems correct (iirc PA is the last thing to be set regarding extrusion?)
For now retraction distance is 1mm at 30mm/s , i have set deretraction speed to 0 in prusa, maybe i should correct that actually. The retraction on tool change for the parked tool is 10mm. I don't have any "extra length on restart" or "deretraction extra length". Maybe that could help but the print is not stringy so it doesn't seem like some material is dripping during travel moves or when an extruder is on standby.
You also mentionned extrusion factor.
Should i just do single wall small diameter towers to reproduce said small features and test varying settings for all those factors ? Or are there more elaborate benchys you might think of ?
@T3P3Tony thank you , i have tested many temperature settings and did a few temp towers, I think this at least is on point. What other basic tests do you suggest i should run ?
Hi everyone,
Sorry i didn't really know how to phrase the topic.
I'm having an issue with a long print where small circular portions of the tree support printed with transparent pla out of a .4mm nozzle look frosty. It looks like the movement and/or the extrusion are jagged. Maybe with some underextrusion.
Machine is a custom build with bmgx2 extruder and chimera+ dual hotend system. I tried setting different PA values, not sure it made a huge difference. For now they sit at .025 for both extruders. Other extruder has .2mm nozzle. Although both are direct drive, there is a very long ptfe tube routing the filament from the dryer to the filament monitors attached to the frame, then to the extruders. Nozzles are both hardened steel from brozzle (i like the reliability of those). I suspect PA could be the issue and the solution but am all ears to any other possibilities.
Attached are pictures, actually even the larger parts don't look that smooth, but as they don't fall apart I didn't notice that much.
Thaznk you all in advance , best
@sebkritikel Thank you I tried with the nightly build. Still not able to slice but if i understand correctly, a report was sent to cura devs ? Hopefully this helps
@sebkritikel I don't have large flat overhangs, but i do have a lot of complex overhangs (printing a "realistic" model of a plant) using, indeed, tree supports.
I found the setting to add M73 commands but it still would not output the preheating commands, I'll try contacting the creator of the script, see if he is willing to share his config file
@sebkritikel so Cura fails to slice my model, I'll leave that aside for now
@droftarts As for Prusaslicer's script i'm having trouble getting it to function . It is my understanding that it is looking for every M109 commands in the gcode to replace them with other commands a bit before the tool change. I tried editing my tool change gcode so it includes an M109 instead of the M116 I had set, but it still doesn't work. I need to dig
; tool change gcode
G91
G1 Z1
G90
T[next_extruder]; select tool
M104 S160 T[previous_extruder] ; set previous extruder to standby temp
M109 S[temperature] ; wait for temperature
Edit : Maybe i need M73s in the gcode too?
@sebkritikel Thank you very much for the info I'll give it a try too. Last time I used Cura (maybe one year ago?) it also had issues handling the large files I'm dealing with. Will report !
@droftarts thank you very much i will give this a try.
Unfortunately IdeaMaker doesn't handle large files very well, Prusa seems to have much better optimisation
Hello everyone,
I'm using a dual extruder setup on my custom machine with a Mainboard 6HC, slicer used is PrusaSlicer. I also have a Raise3D pro3 at work and their slicer (IdeaMaker) has an option to start heating up the inactive extruder before tool change so the machine doesn't stay idle too long.
Posting this here because I don't think the setting exists in PrusaSlicer, and I can't think of a way to detect an imminent tool change with a macro.
Maybe I could use the last known layer time, but i will have to assume each subsequent layer has approximately the same completion time.
Does someone have an idea ?
Thank you in advance
Best
@Phaedrux Thank you for your reply, I haven't tested that it's a good idea, will do after the current print is finished (about 12days...)
Hi everyone
When resuming after a power failure i notice a slight layer offset in z (too high generally. At the start of a print I will home Z with the probe then set the Z max limit thanks to a max endstop
;update Zmax endstop position
echo "Looking for Z max..."
G91
G1 H3 Z400 F600
G1 H2 Z-5
G1 H3 Z10 F{global.probingSpeed}
G90
M500
So after a power failure the routine is to home x and y normally then home Z with the Z max endstop . My question is:
If i probe the bed at X155 Y500 (middle of the bed), does it make a difference regarding the mesh bed compensation if I home Z max at a specific X Y position ? i.e. should I also home Z max at X155 Y500 ?
; homezmax.g
; homing Z using the max endstop
G4 P10
echo "Homing Z max..."
G4 P10
G29 S2
M290 R0 S0 ; cancel baby stepping
G1 H1 Z5
G1 X10 Y10 F5000 ; move slightly away from the 0s
G91
G1 H1 Z400 F600
G1 H2 Z-5
G1 H1 Z10 F{global.probingSpeed}
G1 H2 Z-10 F600
G29 S1
G90
Thank you in advance