I cannot change the M109 code made by Cura.
so the PID settings inside printer.cfg are the point?
Of course you can:
You can put anything you like as start or end gcode. In the Klipper world we typically call own macros there that do what you think should be done. Example here Dual Loop PID, a accurate bed surface temperature - #13 by koconnor
1 Like
I can ? so if I define M109 in my Cura start code, Cura itself does not fill an extra M109 in the printer file? I have to test.
your link is for pid_calibrate (the description is a bit short). does Klipper finde out the correct PID settings by performing a calibration test?
now it takes 53s to start printing (start g-code)
it overshoots on 3°C and needs time to come back
could it be a faster start by these several settings in cura with start print tmp and so on?
tuning?
I am getting closer to my individual fully changed and tuned Tronxy! THX a lot to all helping hands!!
but still the cancel procedure needs to be changed.
I have a display on the Raspi, if I press cancel there, he does nothing!
If I press cancel on my pc (frontend of mainsailos), it moves the print head to maximums
I do not see the solution here.
This PID graph does look strange.
Put the nozzle only 1-2mm above the bed and turn on the part cooling fan and then tune for the temperature you want to print with.
And perform PID with filament in the hotend.
I will! the graph seams not to be abnormal for me, he pushes to tmp, turns off until a tmp below target, then turns on 100%⌠this is the PID Testprogramm.
The PID graph should have I think 7 or 8 peaks and not such a plateau.
This plateau might maybe result from a network transmission error.
As everybody is saying, the PID graph looks weird.
I just did a couple on one of my printers and this is what it looks like for both the extruder and heater bed, The heater_bed graph hasnât completed because I wanted to get both on the same screen shot. Note that there was a reset after doing the extruder to execute the SAVE_CONFIG command.
Your PID values donât look out of line, but that temperature waveform is very strange. You should see a nice regular wave like what I got.
1 Like
just did it again with your instructions:
looks better
tuning? how do I get fester into target window to start printing?
1 Like
Youâve got a pretty good slope on the temperature rise - I donât think you can do much better than that.
If you look at my curve, it takes around 90 seconds to get to 195 and in yours, it takes around 30. What wattage heater are you using in your hot end? Iâm using a 40 Watt.
It takes him about a minute to reach the temp not 30 seconds.
But that looks fine now. Did the new values differ much from the previous ones?
old:
#control: pid
#pid_kp: 20.427 # BQ
#pid_ki: 0.721 # BQ
#pid_kd: 144.775 # BQ
new:
#*# [extruder]
#*# control = pid
#*# pid_kp = 21.207
#*# pid_ki = 0.889
#*# pid_kd = 126.445
small or big difference?
I really do not know, bought year ago, thinking of 50W
One of the characteristics of PID values is that seemingly small differences can make huge changes in how it works to the point where the system becomes unstable. It really is a black art unless you understand exactly what you are doing - control theory for me was almost 40 years ago, so while I understand the basics, I wouldnât trust myself to come up with the values on my own without a lot of research. Automatic PID tuning works well and produces good values.
Have you tried a print since doing the tuning?
actual I am working on a double fan duct to optimize cooloing of extruder and print part
after build of the twin print part fan:
#*# [extruder]
#*# control = pid
#*# pid_kp = 20.535
#*# pid_ki = 0.801
#*# pid_kd = 131.678
I will do it again, to see if the values are stable
second time:
PID parameters: pid_Kp=20.873 pid_Ki=0.809 pid_Kd=134.632
seams good to me
just started a new print, he stops (pause) mid in 1. layer ??
and moves head to max endstopsâŚ
moonraker.log (51.8 KB)
klippy.log (1.5 MB)
and stops again after a view layersâŚ
moonraker.log (51.8 KB)
klippy.log (1.6 MB)
what is the problem?
Filament Sensor filament_sensor: runout event detected, Time 2577.28
Exiting SD card print (position 18258)
Your filament sensor is misbehaving
1 Like