Calbrating 3D Printer Fillament

Calbrating 3D Printer Fillament

I have spent a considerable amount of time ensuring that the model is not only easy to print but also fun. Before starting the print, there are a couple of things to keep in mind.

Firstly, the printer needs to be calibrated. I cannot provide much assistance as I do not know which printer you have, but there are numerous resources available. When calibrating the printer, the most important things to consider are squareness and belt paths. We will deal with the e, x, and y steps next.

Secondly, the filament calibration is crucial to successful and enjoyable printing, and it is often overlooked. The best part is that it only takes around 15 minutes.

Meet the AKRO calibration model. This model is included in the package you downloaded and is used to calibrate the flow rate, dimensional accuracy (using M92 steps/mm), pressure advance, input shaping, strength, and temperature.

It is essential to note that everything AKRO prints is done with a 0.8 or 1.2mm nozzle. This is done for speed and strength. A larger nozzle means a higher flow rate, which translates to faster production of parts. It is pretty simple. Additionally, the goal is to produce a part that is almost entirely isotropic or crystalline or pure. When using small nozzles and thin layer thicknesses, higher chances of air pockets and failure points are introduced, resulting in reduced strength. A well-tuned 0.8 nozzle print will always be stronger than an equally tuned 0.4 nozzle print due to fewer nucleation points (air pockets) that create high-stress zones. However, everything can be printed with a 0.4mm or smaller nozzle. If you are curious, I have included a slicer profile in the file download for you to look at my standard settings.

30mm x 30mm Post Shrink Calibration Cube: Plastic shrinks, particularly when it solidifies from liquid form. This cube must be measured after the print has cooled.

100mm x 100mm Preshrink Calibration Cube: Plastic finds it challenging to shrink when stuck to a bed. This cube is ideal for measuring (the base of the cube or closest to the bed) when the print is finished but still attached to the bed. It provides a good understanding of the printer's calibration, more specifically, the printer's X and Y steps/mm, which is controlled by the M92 command for RepRap firmware. We'll delve more into that later.

Extrusion Coefficient Calibration: For default (0) width settings in Prusa or Super Slicer, 0.8mm nozzles should produce two lines and a total width of 1.55mm. For 0.4mm, four lines and still 1.55mm. More on how to adjust these later.

Pressure Advance Ref: As a rule of thumb, if the corner is budging (too much filament at the corners), then there is not enough pressure advance. If it is tearing itself apart (too little filament at the corners), then there is too much pressure advance. There are excellent tools available to further calibrate pressure advance, but this is a good indication of when a change is needed.

Input Shaper Ref: Input shaper is used to mitigate ringing or phantom lines in prints. It is strongly advised to follow the Klipper or Duet guides on input shaper tuning. However, for 99% of printers, a generic setting such as M593 P"zvdd" F40 can handle every possible ring produced by a printer. If this does not make sense, online resources on how to tune input shaper can be helpful. Similar to pressure advance, input shaper can be a great reference for whether something needs to be further tuned. It is common to find that ringing when pressure advance is on is caused by loose belts, belts riding up and down idler pulleys, or something completely unrelated to input shaper, especially if squeaking or ticking is heard during printing.

Strength Samples: Why print something if a 3-month-old baby can break it? Use these little bits to break in half, twist, and mangle to discover if any layer separation or premature breaks occur. Necking and stretching of the plastic should be seen rather than snapping. As a rule of thumb, if it breaks too easily, fan down then heat up. If it is strong but looks poor, fan up then heat down.

How to use this Calibration Method

An excel sheet is included in the download with the following table:

Given

Actual

Ratio

M92 Start

M92 Change

x

30

y

30

e

1.55


Print the sample and fill out the table with the actual values produced and measured with calipers. In the M92 Start section, one needs to look into the config file to find out what (M92) steps per mm the printer uses for x and y. It is recommended to keep these values true to the motor data sheet in the config file and make all filament changes in the slicer pre and post gcode for particular filaments. The calipers always give accurate measurements, regardless of the M92 value.

All slicers, as far as the writer knows, have pre and post custom gcode options in the filament sections. This is what it looks like for Prusa and Super Slicer:

As shown above, in the start gcode, M92 X81.08 Y80.54, and M92 X80 Y80 are listed, while in the end gcode, only M92 X80 Y80 is listed. M92 X80 Y80 represents the steps/mm according to the motor manufacturer of the stepper motors used. M92 X81.08 Y80.54 is what the excel sheet produces when measuring the sample 30mm cube. In this way, one can measure the calibration cube, input the M92 Change as needed and print away without doing any XY scaling to the model or messing up other profiles.

For the above example, the filled-out excel fields would look like this:

Given

Actual

Ratio

M92 Start

M92 Change

x

30

29.9

1.003344482

80

80.26755853

y

30

29.9

1.003344482

80

80.26755853

e

1.55

1.5

1.033333333

0.95

0.9816666667


To tune the extrusion factor, take several measurements of thin line segments (strength samples before breaking them) and record the average in the “Actual” column of the “e” row. Then use simple calculations to obtain a ratio and multiply it by the Extrusion Multiplier used during the sample print. Adjusting the extrusion factor is done similarly, but you only need to change the “Extrusion Multiplier” in the filament profile.

It is surprising that popular slicers do not treat x and y steps/mm the same way they treat extrusion steps/mm (e-steps), considering that all are equally important. However, you can also use M92 EXXX in the same manner used for x and y by keeping the Extrusion Multiplier always set to 1. Since it is available in the slicer, why not use it?

It is essential to note that if your extrusion multiplier was not accurate (you measured something greater or less than 1.55), it would affect the x or y dimension of the 30mm cube precisely by that much. For instance, if you measured 1.6, then the actual measurement of the cube, based on how much the machine was told to feed, would be 30.05 (30mm + ½ extrusion error + ½ extrusion error).

Print another test object quickly to verify the measurements, and that's it! You can now print all AKRO products with the assurance that they will fit where they are supposed to.

Back to blog

Leave a comment