Question about first test with Pressure Advance
-
Hi,
Acting on the suggestion of someone here I decided to see what Pressure Advance (PA) had to offer.
I use Zesty Nimble devices on all my printers so I didn't know if PA would make a difference.
The nozzle is 0.4. The printing speed is 15mm/s for the first layer and 90mm/s after that. The printing temps are 200 and 60.
I couldn't figure out how to make the seam appear in the middle of a face of a cube (the suggested shape for testing) so I went with a cylinder which solved that problem.
Using the conditional code feature of firmware v3.3 and the layer change code hook of Simplify3D I created the code to change the PA
setting by a fixed increment of 0.05 every 5mm of height starting with a PA setting of 0.The result is shown:
The bottom 2mm is solid to create adhesion after that there are just two walls.
I cannot see a lot of difference in the first 5mm where PA was 0 and the next 25mm where PA went from 0.5 to 2.5.
That seam in the last 5mm band (25 to 30) seems just a bit smoother than the seam in the first 5mm band but it is not drastic.
I'm guessing the gap that started to appear is to be expected.
Any feedback would be welcome.
Thanks.
Printers: A FT-5 with the 713 upgrade bits, a custom MarkForged style, a small Utilmaker style, a small CoreXY from kit and a E3D MS/TC setup. Various hotends. Using Duets (2 and 3) running 3.4.6
-
@fcwilt Yeah my opinion of it is... unclear. Several test pieces were made and since it's a direct drive setup, the range was kept between (0 - 0.06 or so). There was so little difference it was left at 0.04, but honestly, I could probably not see a difference between a part with zero and another with 0.04. Hoping that this is a good thing?
-
@fcwilt said in Question about first test with Pressure Advance:
That seam in the last 5mm band (25 to 30) seems just a bit smoother than the seam in the first 5mm band but it is not drastic.
Now that you've found the useless values, tighten in on that useful range.
The nimble extruder isn't really designed with PA in mind. And being direct drive the needed PA values would be low anyway. I would focus on tuning between 0.01 and 0.06.
-
Having followed the advice given here I tested the range 0.01 to 0.06.
Using 0.05 seems to produce the best result but the difference is slight and I doubt I could tell if I didn't have them side by side for comparison.
Thanks for the feedback.
Frederick
-
-
@phaedrux said in Question about first test with Pressure Advance:
@fcwilt said in Question about first test with Pressure Advance:
0.5
0.5 or 0.05?
My bad - a typo - yes 0.05 - I tested as you suggested from 0.01 to 0.06.
Having never used anything other than a direct drive extruder of one kind or another I have no experience with the workings of Bowden tube setups but I gather they can truly benefit from PA.
Frederick
-
@fcwilt To evaluate the effect of just one parameter, one needs to devise a test method which will isolate that single parameter from all others. The test should also magnify the effect of that single parameter so that changes are easy to see. I devised such a test several years ago, wrote it up on my blog and have been banging on about it for years but nobody takes any notice.
Printing a cylinder which will be made up of multiple short segmented moves is probably the worse test method one could devise for evaluating pressure advance (IMO).
-
@fcwilt which slicer?
-
@pcr Simplify3D
-
@deckingman said in Question about first test with Pressure Advance:
Printing a cylinder which will be made up of multiple short segmented moves is probably the worse test method one could devise for evaluating pressure advance (IMO).
Well it got the seam away from the corner of the cube I tried first and it made the effect of changing PA easy to see.
If I can figure out how to move the seam to the face of the cube I can try a cube again.
But the goal of any such setting is to improve the quality of the objects being printed and they don't all fall in the category of nice simple cubes.
-
@fcwilt said in Question about first test with Pressure Advance:
@deckingman said in Question about first test with Pressure Advance:
Printing a cylinder which will be made up of multiple short segmented moves is probably the worse test method one could devise for evaluating pressure advance (IMO).
Well it got the seam away from the corner of the cube I tried first and it made the effect of changing PA easy to see.
If I can figure out how to move the seam to the face of the cube I can try a cube again.
But the goal of any such setting is to improve the quality of the objects being printed and they don't all fall in the category of nice simple cubes.
A simple cube won't work either. Neither will anything that focuses on the seam because that only happens on layer change and will be affected by other factors such as retraction settings or tool dwell time.
-
@deckingman said in Question about first test with Pressure Advance:
A simple cube won't work either. Neither will anything that focuses on the seam because that only happens on layer change and will be affected by other factors such as retraction settings or tool dwell time.
I was just following the docs on PA testing and a cube was suggested.
Feel free to add your own suggestions to the docs.
Frederick
-
@deckingman said in Question about first test with Pressure Advance:
...I devised such a test several years ago, wrote it up on my blog and have been banging on about it for years but nobody takes any notice... Printing a cylinder which will be made up of multiple short segmented moves is probably the worse test method one could devise for evaluating pressure advance (IMO).
I went to your site and was surprised that the test object file isn't being offered up. How about posting up the test object G file and help others out instead of saying that their test objects are poor.
Scratch-built 350mmx350mm coreXY, linear rails, ballscrews, 3 Z axis, Duet3 6HC, v3.3, Tool Board v1.1, BondTech LGX + Mosquito hot end
-
@fcwilt i would try using the marlin K-factor Calibration Pattern. Need only some find/replace on the generated gcode because it uses the marlin language.
-
@mikes said in Question about first test with Pressure Advance:
@fcwilt i would try using the marlin K-factor Calibration Pattern. Need only some find/replace on the generated gcode because it uses the marlin language.
Thanks.
I went to the site but am at a loss for what sort of thing is being printed.
Just have to try it.
Frederick
-
@fcwilt here's something to read to understand how to proceed with the test and determine the correct PA value based on it.
-
@kb58 said in Question about first test with Pressure Advance:
............ I went to your site and was surprised that the test object file isn't being offered up. How about posting up the test object G file and help others out instead of saying that their test objects are poor.
How about you look again my friend. The gcode file IS listed.
I didn't make a downloadable file because:
a) It's only about 20 lines in total and a simple cut and paste will work.
b) nothing could be simpler than a series of parallel straight lines
c) the lines need to be as long as possible but that would depend entirely on the size of the build plate. So any "downloadable" gcode file would have to be edited by the user to suit their particular build plate.
That post and the associated videos took me several hours to put together. In fact, if you look at my blog post and YouTube channel, you'll see hundreds, if not thousands of hours worth of content, all of which I have published for absolutely no financial gain but with the sole purpose of trying to give something to the community. So how bloody dare you accuse me of not being helpful. !!
