Bad filament with changing diameter causes an uneven flow of plastic out of the nozzle. Besides buying very high quality filament (do we really want to be dependent on professional Filament producers?), what other ideas and methods could be used to mitigate the differences in diameter?

Could a pellet extruder be a solution?

  • 2
    $\begingroup$ This is way too broad. You're not only asking for all possible ways to deal with bad filament, you're also asking about pellet extruders and seem to want a discussion on whether we need to be dependent on professional filament. $\endgroup$ Mar 30, 2016 at 17:35
  • 1
    $\begingroup$ I'm only interested in ways to get better results from bad filament. I mentioned pellet extruder and dependence only to put the question into perspective. I don't want to start a political discussion about the dependency. Do you have a suggestion of how to reformulate the question to make this clearer? $\endgroup$ Mar 30, 2016 at 20:41
  • 1
    $\begingroup$ Perhaps provide details of what machine or extruder type you're working with and re-title the question to something like How to compensate out-of-round filament?. Putting the emphasis on the out-of-round issue may narrow this question down. $\endgroup$
    – tbm0115
    Mar 30, 2016 at 20:54

1 Answer 1


I am only going to answer the part of the question about how to improve the results from low-quality filament:

There have already been projects going on dealing with the issue of varying filament diameter. A sensor that detects the filament's diameter and sends its value to the printers firmware can compensate for the varying volume of the filament pushed into the hotend by simply changing the extrusion speed. You could think of an extrusion multiplier or a change in the steps/mm value. Of course this has to be calibrated for the distance the filament needs to travel from measurement to hotend.

A good first read can be found in the following project:

And the corresponding Thing page:

  • $\begingroup$ As I understand that sensor (light and measuring shadow) it only detects deviations in one direction(wider) not in the other (thicker) . I hope you get my meaning. And how should that work with transparent filament? Do I need to train the sensor to the filament? $\endgroup$ Mar 31, 2016 at 21:36
  • $\begingroup$ I am not following up on the project too much, but yes, the sensor only detects one direction. However, due to the manufacturing process by extruding, it might be safe to say that the filament can always be taken to have a circular cross section. There might be problems with transparent filament, which either might be taken care of by the calibration process. You could however use a lightsource that your filament is not transparent for, if the sensor can still pick that up. $\endgroup$
    – kamuro
    Apr 1, 2016 at 6:36
  • $\begingroup$ Can you explain why "it might be safe to say that the filament can always be taken to have a circular cross section" ? I remember perfect roundness to be hard to achieve. And if it is not perfect round, then variations in roundness might occur for various reasons, right? $\endgroup$ Apr 2, 2016 at 20:31
  • $\begingroup$ I would assume that, being extruded from a circular nozzle, the filament does not see any reason to produce a non-round shape. Dirt on the nozzle, varying cooling over the cross section as given by airflow, not cirular nozzle and asymmetry in the force given to the molten plastic in the nozzle at the point of extrusion are some reasons, however, which I could imagine impairing the the roundness of the filament. $\endgroup$
    – kamuro
    Apr 4, 2016 at 6:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .