When I first started 3D printing, there wasn’t much concern for the diameter of filament, by the manufacturer or consumer alike. Average diameter tolerance was 0.10mm more or less than the nominal diameter, which doesn’t sound like a lot but actually plays a fairly large role in how clean 3D prints will actually come out. An inconsistent filament will float back and forth between over and under extruding, which can be confused with Z-banding or other troubleshooting symptoms from the seemingly regular pattern of it. Today, you’d be hard pressed to find a filament that doesn’t have at least a 0.05mm tolerance, but what’s not often talked about is ovality, so let’s talk.

Ovality refers to how oval is the filament; the less oval the more it approaches being a circle and “perfectly” round. Detecting ovality requires a little insight into how filament diameter is measured in the first place:

1-Axis Laser Micrometer

A 1-axis laser will shine a beam on the filament as it is extruded from the extrusion line. As it passes over the transmitter laser, it blocks the light from entering the receiver, which tells the computer how wide the filament is and will ring an alarm that the filament is out-of-spec and oversized or undersized, and will adjust how fast the filament is being pulled to get it back in line. But there’s a problem here, it will only measure left to right, so it may be 1.75mm wide, but it could be infinitely deep and there would be no way for the system to see that with only one laser.

1-Axis micrometers use one laser to measure filament diameter in one direction.
1-Axis micrometers use one laser to measure filament diameter in one direction.

2-Axis Laser Micrometer

Enter the 2-axis laser. This can at least see the width and depth of the filament, but there are still angles that the filament can’t be read from - you could theoretically have a perfectly square 1.75mm filament that to the computer’s knowledge is totally passable. From corner to corner, that square is 2.47mm, which is significantly oversized from the nominal filament diameter and has a cross-sectional area of 3.06mm2, which when compared to the optimal cross-sectional area of 1.75mm filament (which is 2.41mm2), it’s 27% over extruded.

2-Axis micrometers use two lasers to measure filament diameter in two direction.
2-Axis micrometers use two lasers to measure filament diameter in two directions.

3-Axis Laser Micrometer

3-axis laser micrometers are used for our PRO-series line of filament. Filament observed in these systems at their worst can be seen as hexagons. Basically, take the number of axes used to observe filament, double it, and that’s how many sides the theoretically worst filament can have. A hexagon with a width of 1.75mm from one flat edge to the other has a width of 2.01mm from tip to tip. That translates to a cross-sectional area of 2.65mm2, which is 9% over extruded, a 3x increase from 2-axis laser micrometer readings.

1-Axis micrometers use one laser to measure filament diameter in one direction.
3-Axis micrometers use three lasers to measure filament diameter in three directions.

Sampling Rate of Filament Diameter

The other important comparison is how often the diameter is sampled by the micrometer. Micrometers output data at an almost constant rate (thousands of samples per second), which makes it difficult for a person to actually interpret that data. Instead, the data is averaged across some timespan to check the average diameter size, and the fluctuations more or less than 1.75mm in that time. If you are checking less often, that gives the system more time to smooth out its mistakes. Take this chart for example. Let’s say this is a time span of 2 seconds, and this line marks 1.75mm. If the filament is being extruded and has a hiccup where it way over extrudes, then way under extrudes, and steadies, an average in that timespan will put it at 1.75mm. But if we check every 1 second, suddenly you have a section that alarms out for being too large followed by a section that alarms for being too small. The sum of it all is the more often you check the average diameter size, the tighter your tolerances need to be.

Sampling rates can hide the worst filament through an infrequent sampling.
Sampling rates can hide the worst filament through an infrequent sampling.

Having filament that’s reliable is as imperative as having a reliable machine to churn out production parts, and I personally endorse the tighter tolerances the industry is moving to accommodate, but it is important to understand what they actually mean so we aren’t comparing apples to oranges.

I hope this helps give you an idea of what’s going on behind the scenes for your favorite filament manufacturer and encourages you to ponder and dig a little deeper in the future.

Happy printing!