Link to the commands used and some more in-depth info on PA tuning github.com/KevinOConnor/klipp... vorondesign.com/ for all your voron printer needs, dont forget to join our discord!
Excellent explaination. Instead of writing PA value in the printer config, I write it in my slicer(s) as filament-specific start G-Code. That way, each time something is sliced using that filament, the relevant PA is inserted (through Klipper's "SET_PRESSURE_ADVANCE ADVANCE=". Simple and efficient.
PSA!!!! Those of us that like prusaslicer over cura, you HAVE to disable auto cooling in the filament tab or you will NOT hit the 100m/s speed & will get inaccurate readings.
Tip: Watch your print and just cancel it when you see the corners degrading, it'll likely happen within the first 1-1.5", will save you some time and filament.
Helpful video, as usual! Unrelated question... What filament did you use for the blue accent color on your Voron? I've been looking for a nice blue ABS to match the light blue anodization on my AL extrusions and that color looks like it would be perfect!
New to the channel and I'm enjoying the klipper videos, just got my ender 3 8bit board setup with manual bed mesh. These videos help so much. Any custom config to get a manual mesh bed level option in the LCD screen?
Thank you for this video. I had a question, what should the pressure_advance_smooth_time be? I didn't see it mentioned in your video or at that link? or maybe I just missed it?
Hi! Very great video. Sorry, I`m a newbie with pressure advance (and klipper itself). Are these commands for setting a velocity and factor value temporary (until the next printer reboot) or permanent?
In general: your content is why I was comfortable converting to Klipper with incomplete Christmas gifts in Que! Unfortunately: I got a PA of ~1 for my bowden setup. Klipper documentation suggests that PA will have minimal impact, and I should consider disabling PA. This is a Tow Trees Sapphire Pro, with a stock hotend, bowdend tube, and upgraded LGX...OUCH! I'm designing a modular hotend kit now so I can upgrade to the Mosquito a.s.a.p!
I have Artillery genius with LGX + ACE copperhead hot end = so direct drive. got 0.0925 - huge improvements! printing at 100mm/s limit or 14.43mm3/s limit. Klipper does wonders
I get a result that I can't do anything with. It differs per corner for me and I can't get a good value from this to set. Have a bowden printer (cr10s pro v2) with dual extruder. Should I try 010 instead of 020?
this might save someone time, but the issue I was having wasnt this at all. It was wet filament going silly on the nozzle. try different filament and a higher temp if its just abysmal
I tried to do this a few days ago and the box had corners that appeared pretty uniform from top to bottom. I followed your instructions and will reprint overnight. Let's see what I get in the morning.
Ran it again and on my bowden setup the difference is not as great as your direct drive. I was however able to see where it changed on three of the corners and was able to get the measurement I needed. Thank you for putting the video together.
at 100mm/s it's practically hollow - but at 50% speed it's approximately normal Ender 3v2, bowden tube got heat-creep after a few months and I was unconfident so got the direct drive, along with a crtouch and the leveling screws worked great for a couple weeks, at up to 150mm/s then got heat creep - so got both a cheap bimetal heatstop and a redlizard since it was all ready assembled along with some capricorn installed the redlizzard to the capricorn and I think I had it working once, but then showed signs of heatcreep again so switched to the bimetal installed into my direct drive ... which works fine but only at factory speeds (not sure if the original bowden could have gotten faster, but faster was all I tried with the redlizard any ideas on what's happening? next I'll be trying the redlizard installed into my direct drive and see how that goes... I really don't want to buy another hotend, but I'll consider buying the recommended expensive heatstops when I next get some nozels (which I'm suspecting may be the problem anyway? I'm using the same nozel, the one in best condition outside the redlizard, as when I was going 150mm/s)
after i set the pressure advance in the config and start a print i shows pressure_advance: 0.000000 pressure_advance_smooth_time: 0.040000 should it be showing the value you input into your config instead? thanks
thanks for the guide, Do i have to go into my Config File and change this value every time i change filament types? if this is the case then i will sadly have to be doing this very often . Secondly, why dont i have to do this on my other printers?
I find its usually "good enough" with most same type filament, but if you run PA with say a .6 nozzle with abs, and then start running TPU with a .4 nozzle you may need to update it
As mentioned above you can add a line like SET_PRESSURE_ADVANCE ADVANCE= to filament-specific G-code in your slicer. You don't "have to" tune linear advance / pressure advance on any printer but it can improve the quality of printed parts, just like many other optional tuning steps.
Shouldn't you convert your caliper measurement to number of layers to use it in the formula? I thought the tuning_tower command increases the pressure advance every layer, which would require you to use the number of layers to get the value which was used for the best looking layer.
Why is this a firmware tweak if it depends on the filament and printing temperature ? Say I print two brand of abs, or abs and nylon for example. Would I see significantly worse corners ?
I've observed that, I don't even gets the same values between to copies of a printer, IMHO this should be something that slicers should have a value for in the filament profiles, I think prusaslicer now does, or something that can be controlled via M900 like in marlin gcode
Ive found that a small tweak MAY be necessary between filament types (pla vs abs vs petg) but so far between brands of the same plastic i havent noticed much difference
Ended up looking it up, seems like Klipper has a 'SET_PRESSURE_ADVANCE ADVANCE=' command, which you could use in Slic3r/PrusaSlicer/SuperSlicer in the filament specific gcode section, similar to how it's done with marlin via the M900 gcode.
I saw in another of your videos you said you'd have to redo pressure advance after doing an acceleration test. So after running an acceleration test would I put that speed in where you just did thr 500 I think was your value or after I get the values from my accelerometer I still just use this value?
I just have my pressure advance tune per filament, and then setup in that filaments start GCode, so that it sets it before it prints I do it as part of my filament calibration process. Dry the filament Then print a Temp Tower Flow vase Pressure advance. Retraction spikes And then boom. its ready to go.
hey i am new to klipper so i can tune every filament seperatly and can leave the printer.cfg stock so i only need the Start gcode where it sets dies value and after powering down it has no pressure advanced?! but what i dont understand why we make accelerations down. if we tune speed then pressure advanced needs to be tuned again or am i wrong?
I have tried to run this on my ender 3 v2 running klipper mainsail. In the console window in mainsail I constantly get the following error. Unknown command: "M900" . Any ideas?
I calibrate my esteps, then flow rate(extrusion multiplier), then PA. FR/EM is the factor that gives you more dimensionaly accurate parts as the width of the extrusion is calibrated, if the slicer thinks your line width is .48 but your FR is too high, than you line width might be .55 instead. PA helps with creating a ratio of extrusion of line width on speed changesthat varies your EM kinda...
@@CanuckCreator I mean, I haven't made the test with an smaller object, why does it need to be that big. A simple cal cube with extended Z would not achieve the same results?, Also, why the model is square and has internal shape? Won't be simpler a rectangle with single wall?
@@DARKEASC The extra geometry allows you to compare settings for different situations, and along different axis`s of the printer. sometimes inner and outer corners can be slightly different and also x vs y axis on some designs. The print itself uses less than 30g of plastic and can be stopped as soon as they corners start to get bad again.
if only klipper would talk to the adxl... :( tried 3 diff chips on both rpi3 and 4. klipper can only repeat id (got xx vs e5)with shielded cables, duponts, with R4 removed, 3v,5v. cheap adxl, adafruit fancyb1tch. must be getting close to 100h on this subject.... hopefully kevins current work on a adxl branch might bring some light...
I don’t seem to have "pressure_advance:" nor "pressure_advance_smooth_time:" on my printer.cfg, do I need to add it before or after the hollow cube print?
@@maxfuentes7541 You can use the default smooth time. It is possible to set it lower, but that will cause more and faster extruder movement (risk of skipping steps).
Is it necessary to add the line "pressure_advance = " in the printer.cfg file BEFORE executing the tuning test? I ask because my cube seems pretty uniform from top to bottom..
No, running TUNING_TOWER COMMAND=SET_PRESSURE_ADVANCE ... overrides the saved pressure_advance value in printer.cfg for that print. If the test cube looks fine and you can't feel a little bulge at the bottom of each corner, pressure_advance won't help your prints and you should set it to 0.
This has been driving me insane. Everytime I have tried tuning PA, after the first few layers I get insane extruder skipping, which only gets worse. Has anyone else had this problem? Any solutions? Cheers
you either have a clogged nozzle or a cracked extruder. if you have the stock plastic extruder, look for cracks. especially under the tension arm. also check the tension of the extruder. lastey, check and make sure your extruder gear (the one that "bites" into the filament) isn't worn down where the filament runs along it. then I would redo Esteps and try again.
Just read the manual: "The test print is designed to run with a high extruder flow rate, but otherwise "normal" slicer settings. A high flow rate is obtained by using a high printing speed (eg, 100mm/s) and a coarse layer height (typically around 75% of the nozzle diameter). Other slicer settings should be similar to their defaults (eg, perimeters of 2 or 3 lines, normal retraction amount). It can be useful to set the external perimeter speed to be the same speed as the rest of the print, but it is not a requirement."
If a high pressure advance value (eg, over 0.200) is used then one may find that the extruder skips when returning to the printer's normal acceleration. The pressure advance system accounts for pressure by pushing in extra filament during acceleration and retracting that filament during deceleration. With a high acceleration and high pressure advance the extruder may not have enough torque to push the required filament. If this occurs, either use a lower acceleration value or disable pressure advance.
honestly the marlin method's is just so much better. i wonder why you have to do this..... instead of the line method also my eyes aren't great at judging the this like that
