Extruder Calibration  

Re: Extruder Calibration

Due to some other problems I started with the calibration of the extruder (steps/mm and extrusion multiplier). Check this link for a very good explanation: mattshub.com/2017/04/19/extruder-calibration

For me was the calibration of the extruder motor (steps/mm) not good. I changed it in the settings of the "printer settings" -> "start Gcode" (because in the fw it is not possible to place it in the EEPROM) bij adding:
M83 ; extruder relative mode
M92 E171.6

Afterwards, printing the cube with 1 perimeter lead to the conclusion that my extruder multiplier is 1, so no correction

For it was an improvement

Frans

Re: Extruder Calibration

Due to some other problems I started with the calibration of the extruder (steps/mm and extrusion multiplier). Check this link for a very good explanation: mattshub.com/2017/04/19/extruder-calibration

That's the method I was outlining above.

M92 E171.6
Afterwards, printing the cube with 1 perimeter lead to the conclusion that my extruder multiplier is 1, so no correction
For it was an improvement

Calculating a new M92 value will give you an extruder multiplier of 1 for that particular type of filament, but you would get exactly the same results with the default M92 value (161.3) and setting your extruder multiplier to 1.06.

If you set your "M92 E171.6" and then extrude 100mm of a different type of filament than you were using when you calculated the M92 value, you won't get 100mm. PLA / PETG extrude different amounts for the same number of steps.

Re: Extruder Calibration

This weekend I will try another type of filament with the same diameter.
If I understand the working of the extruder motor (steps/mm) correct then it should make no difference what kind of filament (with the same diameter) is pushed trough.
By the way, I did try to change the extrusion mutiplier (set it to 1.06) but that give me overextruding and very bad first layers wobbles.

Frans

Re: Extruder Calibration

If I understand the working of the extruder motor (steps/mm) correct then it should make no difference what kind of filament (with the same diameter) is pushed trough.

It's possible my filaments are different thicknesses and accounts for the two different extrusion rates I was seeing, but my new caliper batteries haven't arrived yet.

I assume there is a lot of physics that goes into it, but given that it's pulling the filament through via friction, I would assume PETG and PLA don't have the same coefficient of friction and one would be slipping on the gear more than the other. One would be softer than the other, so it would be letting the gear dig into it a little bit more.

You couldn't extrude a filament that was too slick, or too soft... And PLA and PETG are clearly not the same hardness, and I would imagine PETG is a little slicker too... So I don't find the idea that they'd extrude at ~5% different rates to be such an off-the-wall idea.

By the way, I did try to change the extrusion mutiplier (set it to 1.06) but that give me overextruding and very bad first layers wobbles.

Using what value for the steps/mm?

Re: Extruder Calibration

So as promised I did some measurements:
For all 100 mm set
ABS green Filright result 92 mm
ABS orange Prusa result 92 mm
PLA white Filright result 93-94 mm
PLA silver Prusa result 94 mm
PLA copper Prusa result 94-95 mm

The results are close to each other. There was no slipping of the gear. Diameters are all 1.75
So there are differences. If the extruder delivers constant # rotations, there should be no differences. Resistance of the filament in the hotend can not be the cause.

Re: Extruder Calibration

Resistance of the filament in the hotend can not be the cause.

No, it's the hardness/softness of the filament.

Peter

Re: Extruder Calibration

Hi Peter,

Thanks for replying. I love the discussions and help for understanding the working of my prusa.

So when the # rotations of the extruder motor is set to the same, when using different types of filaments; the pulley won't push through the same length of filament due to the soft/hardness of the filament.
So the case is:
* the softer the filament the "less" filament is pushed through and
* the harder the filament is, the "more" filament is pushed through (not more then theoratically possible by the # rotations of the extruder motor 😆 )

Frans

Re: Extruder Calibration

Sort of, but the filament tension also comes into play. For softer filament, less tension is required.

If too much tension is used, the drive pulley can bite into and damage the filament and cause poor feed through the PTFE tubing.

Peter

Re: Extruder Calibration

i dont know why there's a need to calibrate the extruder . the only reason to do so is if you want to copy slicer setting from someone else , or use some general settings .

i dont care if my extruder extrudes 5mm when i ask it to extrude 10mm .
the only thing i do is adjust my extrusion multiplier for each type of filament i use by checking my 1st layer and top layer finish (live z is constant for all filaments off course).
i store my extrusion multiplier and other settings for each filament and thats all .

i dont copy slicer settings , i do my own tweeking and this way i get constant performance for each filament without ever needing to measure how much filament is extruded when i ask it to extrude x amount .

Re: Extruder Calibration

i dont know why there's a need to calibrate the extruder . the only reason to do so is if you want to copy slicer setting from someone else , or use some general settings .

i dont care if my extruder extrudes 5mm when i ask it to extrude 10mm .
the only thing i do is adjust my extrusion multiplier for each type of filament i use by checking my 1st layer and top layer finish (live z is constant for all filaments off course).
i store my extrusion multiplier and other settings for each filament and thats all .

i dont copy slicer settings , i do my own tweeking and this way i get constant performance for each filament without ever needing to measure how much filament is extruded when i ask it to extrude x amount .

Yup, that's another way of doing it. There are usually quite a few ways that will achieve the same objective (example: changing the filament diameter in the slicer also adjusts the feed).

People will do what they find works for them.

Peter

Re: Extruder Calibration

i dont know why there's a need to calibrate the extruder . the only reason to do so is if you want to copy slicer setting from someone else , or use some general settings .

Yup, that's another way of doing it. There are usually quite a few ways that will achieve the same objective (example: changing the filament diameter in the slicer also adjusts the feed).

So there are obviously a bunch of ways to adjust how much filament is shoved out of the printer, but how many of them affect retractions? I would assume the esteps value does, but I would assume the extrusion multiplier does not.

Does anyone know?

Re: Extruder Calibration

So there are obviously a bunch of ways to adjust how much filament is shoved out of the printer, but how many of them affect retractions? I would assume the esteps value does, but I would assume the extrusion multiplier does not.

Does anyone know?

Yeah, that's probably right, but then again, a retraction of 0.8mm vs one of 0.75 or 0.85 will not make a noticeable difference.

Still it's easy enough for you to test out, it you really want to.

Peter

Re: Extruder Calibration

Still it's easy enough for you to test out, it you really want to.

Pardon my ignorance, but how would one go about testing that? About the retractions, I mean.

I can think of ways, but I can't think of any that don't sound like they'd make my printer explode and/or clog horrifically

Re: Extruder Calibration

1. We know that changing the e-steps affects everything from the point that it is applied to the point where the printer is turned of/resetf or a new value is programmed. This therefore must affect retraqction
2. We can slice G-Code with different extrusion multiplier/flow tweak, different filament diameter, gain etc values and see that the retraction value in the G-Code has not changed.

The default E-Steps is 161.3 steps pre millimetre, but generally this value can vary by about -3% to +7% (according the the notes I have seen on this forum). 7% of the standard 0.8mm retraction distance is a little over 0.05mm. Not really enough to make a difference.

Peter

Re: Extruder Calibration

I spent some time to calibrate my extruder last night, following the standard method of:
(1) Extrude 100mm of filament, calculate an accurate M92 value to make it extrude 100mm of filament
(2) Print the calibration cube, calculate the wall thickness, calculate an extrusion multiplier that fixes it
.....

Is there some low-level difference I'm not understanding between what an M92 command does, and what an extrusion multiplier does?

The same procedure I did... and this was the beginning of a torture! Just wanted to fine-tune my MK2S after upgrading and thought this will help me to get the perfect prints (even with ABS). So I put both numbers (E92 + Extrusion Multiplier) and my prints looked awful (overextrusion). So I started many many test prints and modified all configurations (except E92 + Extrusion Multiplier of course, because I thought this numbers MUST be correct after many many measures) until I just returned the Extrusion Multiplier to 1.0 even my calibration-cube just had 0.41 instead of 0.45 thickness and everything's quite good now (see pictures... just struggling with an overhang/bridging issue with ABS). So I thought either it's not good to do both calibrations OR the instructions for the test (2), the calibration-cube, are not right the way I did it. Because in most instructions they say you have to use the extrusion-width (0.45) for the calibration (what I did). But I also saw a video where they say to use the nozzle diameter (0.4) for calculation what means... everything's fine with my extrusion multiplier. So for me a very confusing thing this calibration-stuff.

So I would be also very interested how other calibrate their system? With both (1) and (2) methods? Or just one? Or how... ? Great discussion!

Re: Extruder Calibration

So as promised I did some measurements:
For all 100 mm set
ABS green Filright result 92 mm
ABS orange Prusa result 92 mm
PLA white Filright result 93-94 mm
PLA silver Prusa result 94 mm
PLA copper Prusa result 94-95 mm

The results are close to each other. There was no slipping of the gear. Diameters are all 1.75
So there are differences. If the extruder delivers constant # rotations, there should be no differences. Resistance of the filament in the hotend can not be the cause.

Probably Skipping/slipping. filament. Put in high power mode, extrude all at same speed. it should be 100% consistent. My non-mk2 printer had this issue due to extruder current being set incorrectly(too low). tighten bolts and see if it's grip as well. try 1-2mm/s extrusion rate for them.

Re: Extruder Calibration

Finally I let extruder steppings to default 161.3. I don't where that value come from perhaps as underextruding is less harm than overextruding and extruder multiplier set to 1 give kind of nice printing to newbies like me. I could set the extruder to 176.3 in my case (+9%) as measured but then I have to decrease the extruding multiplier to get good vase wall width. I prefer to only set the extruder multiplier for different materials. Most of the time it is 1.05.

What I bother is to find a fine and reliable method to set the extruder multiplier. My nozzle is 0.4. If I do a cube in vase mode and a width of 0.45, I can measure if I get 0.42 or 0.48. But I don't find it so accurate. The resolution of measurement is low. It is not like for measuring the 120mm of mattshub.com/2017/04/19/extruder-calibration.

Does someone has a better method to tune the extrusion multiplier? Something to expand the phenomena to somewhat more measurable.

I spent some time to calibrate my extruder last night, following the standard method of:
(1) Extrude 100mm of filament, calculate an accurate M92 value to make it extrude 100mm of filament
(2) Print the calibration cube, calculate the wall thickness, calculate an extrusion multiplier that fixes it
.....

Is there some low-level difference I'm not understanding between what an M92 command does, and what an extrusion multiplier does?

The same procedure I did... and this was the beginning of a torture! Just wanted to fine-tune my MK2S after upgrading and thought this will help me to get the perfect prints (even with ABS). So I put both numbers (E92 + Extrusion Multiplier) and my prints looked awful (overextrusion). So I started many many test prints and modified all configurations (except E92 + Extrusion Multiplier of course, because I thought this numbers MUST be correct after many many measures) until I just returned the Extrusion Multiplier to 1.0 even my calibration-cube just had 0.41 instead of 0.45 thickness and everything's quite good now (see pictures... just struggling with an overhang/bridging issue with ABS). So I thought either it's not good to do both calibrations OR the instructions for the test (2), the calibration-cube, are not right the way I did it. Because in most instructions they say you have to use the extrusion-width (0.45) for the calibration (what I did). But I also saw a video where they say to use the nozzle diameter (0.4) for calculation what means... everything's fine with my extrusion multiplier. So for me a very confusing thing this calibration-stuff.

So I would be also very interested how other calibrate their system? With both (1) and (2) methods? Or just one? Or how... ? Great discussion!

I thought about using (1) and (2) but now I go to only (2).
For the width, I would use the extrusion-width and not the nozzle diameter as explained for slic3r: http://manual.slic3r.org/advanced/flow-math