Calibration calculation alternatives?  

RE: Calibration calculation alternatives?

The procedure described at Mattshub is great for printers with a lot of variability. Most Prusa's are very close without adjustment, so you can just use the default. This gives you a baseline for all filaments.

For each filament type, you do want to tweak the extruder multiplier. Print once at 1.0 then adjust based on those results, print again and verify. Don't forget to measure, average and enter the actual filament diameter into your slicer as well. 

RE: Calibration calculation alternatives?

Posted by: bobstro

The procedure described at Mattshub is great for printers with a lot of variability. Most Prusa's are very close without adjustment, so you can just use the default. This gives you a baseline for all filaments.

For each filament type, you do want to tweak the extruder multiplier. Print once at 1.0 then adjust based on those results, print again and verify. Don't forget to measure, average and enter the actual filament diameter into your slicer as well. 

gotcha , the issues i was having was when doing the 100mm test using calipers, i got 20.10 mm left over when using gray/silver PLA filiment which i got with my kit but when i used my 3rd party PETG and did the 100mm test i got little over 20.50mm left over, since 20.10 is close to 20mm, i think like you said prusa default settings come close to the values i need to ideally get. The PETG on the other hand i think is hydroscopic and  i need to build a dryer. also the PETG varies in diameter when measuring with calipers, not always 1.75 after measuring a few spots(seem to average off to 1.68mm), maby the inconsistency in the diameter is causing me to get 20.50mm and not 20mm.

For now i think i dont need to change the the M92 E steps in the firmware since the prusa gray/silver PLA is close to 20mm.

The cube test on the other hand i am trying to get the right extrusion multiplier for the the 3rd party PETG, but the issues is when i leave the extrusion multiplier(EM) at 1 for the PETG i get a lot of ozzing and stringing and prints fail some times and its hard to print at EM 1, when i set it to 0.9 it seem to print fine and i can continue to do the cube test with out to much fails. The 0.9 number i chose after i did some non-test print which had failed and i just chose a number that i thought was good(no test prints just picked a number). 

so that is why i was wanting to know if i could use 0.9EM to print since i just randomly chose that number seem to work ok, and seem like you said this test can be done with lot of variability assuming its ok to use 0.9 to get the correct EM value, weird part is when i calc the wall width my default is 0.45 and when i measure the wall on the cube(0.15 layer high print) with a caliper i get 0.38(average after measuring 4 walls) that is .45/.38 = 1.18EM

i am thinking i will just have to try to dry my filament when i build a dryer and redo the cube test with EM of 1 and see what happens.

what are your thoughts?

RE: Calibration calculation alternatives?

I look at it this way: The setting in firmware (Mattshub) is a baseline. Unless you really intend to set it for every filament type, you just want it to be close enough to typical to be a baseline that remains constant to adjust other filament types against. This is your "0" setting. You can try the technique on Mattshub, but you'll probably wind up within a  percentagepoint or two for one type of filament. The main thing is that it remain constant.

Regardless of what you do in firmware, every filament type will vary. Unless you really want to do the full firmware calibration, calibrating your extrusion multiplier is a good idea when trying new filament types. Measure and enter the averaged filament diameter into  your slicer (important!) then do the hollow cube measurements to determine your extrusion multiplier for each filament type. 

RE: Calibration calculation alternatives?

Posted by: bobstro

I look at it this way: The setting in firmware (Mattshub) is a baseline. Unless you really intend to set it for every filament type, you just want it to be close enough to typical to be a baseline that remains constant to adjust other filament types against. This is your "0" setting. You can try the technique on Mattshub, but you'll probably wind up within a  percentagepoint or two for one type of filament. The main thing is that it remain constant.

Regardless of what you do in firmware, every filament type will vary. Unless you really want to do the full firmware calibration, calibrating your extrusion multiplier is a good idea when trying new filament types. Measure and enter the averaged filament diameter into  your slicer (important!) then do the hollow cube measurements to determine your extrusion multiplier for each filament type. 

ah..so the 100mm test i did with the silver PLA should serve as a baseline for other filament types but i just have to modify extrude multiplier to compensate for the difference i might see in the 100mm test for other filament types. i see so theoretically it should be ok to leave extrude multiplier at 0.9 for now for the 3th party PETG but then i should do an average of the filament diameter on the PETG and put it in the slicer then do the cube test and then change the extrude multiplier(from 0.9 for the PETG) to what ever the cube test says is best. 

RE: Calibration calculation alternatives?

Posted by: mkio9

ah..so the 100mm test i did with the silver PLA should serve as a baseline for other filament types but i just have to modify extrude multiplier to compensate for the difference i might see in the 100mm test for other filament types.

That's how I view the world, yes. The "extrude 100mm" test on Mattshub is excellent, and doing it certainly won't hurt anything, but all you're doing is shifting your baseline for that specific filament. From the reading and testing I've done, on a (relatively) high-precision machine with good repeatability between printers like the Mk3, there will be very little difference between machines. You'll be making adjustments of < 1-2% most times. And having done that, you'll still want to make final adjustments for each filament type, so obsessing over the firmware settings yields questionable returns. (Speaking from experience.) I think this is one of those "must do" steps from the history of 3D printing -- sort of like "magic layerheights" or printing on glass to get a flat surface -- that are mostly NA for a Prusa. We paid the big bucks to get away from that nonsense. Someone will no doubt disagree, but I'm getting good results so will need to see some evidence to the contrary.

i see so theoretically it should be ok to leave extrude multiplier at 0.9 for now for the 3th party PETG but then i should do an average of the filament diameter on the PETG and put it in the slicer then do the cube test and then change the extrude multiplier(from 0.9 for the PETG) to what ever the cube test says is best. 

I would say do what works best for you. Prusa, unfortunately IMO, decided to compensate for some extrusion characteristics in earlier versions of Slic3r by adding the mysterious adjustment to their startup gcode in the supplied profiles:

M221 S{if layer_height<0.075}100{else}95{endif}

This has gone through a few iterations, but basically tweaks the extrusion multiplier (in effect, not exactly the same way) to extrude at 95% for all jobs sliced at greater than 0.075mm. I've gone back and forth on this myself, but have ultimately decided that I'll take responsibility for doing my own calcs, so comment that line out. Having done that, I then do a hollow 2-perimeter 20mm cube print, average measurements of the tops of the walls and use that to calculate my extrusion multiplier. Again, lots of opinions on "the best" way, but this works very well for me. do find that over-extrusion at layer heights above 0.32mm are a problem, so I've replaced the Prusa line with:

M221 S{if layer_height >= 0.32}90{else}100{endif} ; compensate for thick layer heights

I do a lot with big chunky layers and fat nozzles, so this probably wouldn't apply to most people.

Of course, if you start using different motors or other hardware tweaks, doing the 100mm test is definitely recommended. If one of the smart geniuses like @guy-k2 tells you I'm full of it, go with their recommendation!

RE: Calibration calculation alternatives?

I also followed matts method but found it was not good for me. i done my own.

I do also a 2 walls "rectangular parallelepiped" with external perimeter calculated to target a 0.9 total width if i remember (and it is not 0.9/2). i will post my values when i will access my computer.

And now almost all the time the extruder multiplier is 1.

You decrease quickly the mechanical resistance when you underextrude.

RE: Calibration calculation alternatives?

Here my settings as I like rounded wall width for designing my objects which are only functional ones.

Then for a 0.2mm layer height, i set external perimeter to 0.471 to target a 2 wall width of 0.9. It is calculated from here: https://manual.slic3r.org/advanced/flow-math

For nice looking parts, underextruding may be better. But an extruder multiplier of 0.9 seems to me very low for average Prusa printer and filament.