Is non-linear extrusion worth bothering with?

NexxCat

Quick question for anyone who feels like chipping in their thoughts

I'm working on a project which currently can do the following:

Full automatic calibration of an extruder steps/mm value
Check for extruder consistency
Calculate the maximum volumetric flow for the hot end

This is all achieved using a small piece of hardware I've designed (and which will all be open source when it's done), plus a GUI application written in Python and Qt.

One person said that is would be an excellent tool for calibrating non-linear extrusion, and adding it wouldn't be too difficult (once I get my head around the formula), but it seems no firmware other than RRF supports the feature. Further, it doesn't seem very well documented on how to calculate the A, B & L values from a given set of data (feed rate vs over / under extrusion).

From my data and tests, I see a maximum of 2% under-extrusion when really pushing the filament, but by that point the quality of the plastic coming out of the nozzle has degraded greatly. By the time I see 4-5% under-extrusion, the BMG extruder is skipping.

Do people still use it? Is it one of those features that RRF will eventually drop for extra code space?

Thanks for reading!
Simon

Phaedrux

In my opinion, no, it's not worth it. I think if you're fighting that much back pressure you're likely exceeding the melt rate of the hotend in the first place.

NexxCat

@phaedrux Thank you, that's kind of what I figured from the data I got when calculating maximum volumetric flow.

Appreciate the input, I'll drop the idea of adding it

dc42

@nexxcat said in Is non-linear extrusion worth bothering with?:

From my data and tests, I see a maximum of 2% under-extrusion when really pushing the filament, but by that point the quality of the plastic coming out of the nozzle has degraded greatly. By the time I see 4-5% under-extrusion, the BMG extruder is skipping.

IMO, 2% under extrusion at the maximum flow rate you want to use is not worth worrying about. If it was 10%, then nonlinear extrusion might provide some benefit.

It seems likely to me that dual-drive extruders may exhibit less nonlinearity than single drive extruders.

droftarts

@nexxcat the guy who designed the Orbiter extruder did some volumetric flow tests with different filaments at different speeds. See halfway down the page, here: https://www.thingiverse.com/thing:4725897

Ian

NexxCat

@dc42 said in Is non-linear extrusion worth bothering with?:

@nexxcat said in Is non-linear extrusion worth bothering with?:

From my data and tests, I see a maximum of 2% under-extrusion when really pushing the filament, but by that point the quality of the plastic coming out of the nozzle has degraded greatly. By the time I see 4-5% under-extrusion, the BMG extruder is skipping.

IMO, 2% under extrusion at the maximum flow rate you want to use is not worth worrying about. If it was 10%, then nonlinear extrusion might provide some benefit.

It seems likely to me that dual-drive extruders may exhibit less nonlinearity than single drive extruders.

Yeah, that pretty much matches my data, but thank you for taking the time to give your thoughts @dc42, I know you are very busy

@droftarts said in Is non-linear extrusion worth bothering with?:

@nexxcat the guy who designed the Orbiter extruder did some volumetric flow tests with different filaments at different speeds. See halfway down the page, here: https://www.thingiverse.com/thing:4725897

Ian

That is a super useful link. Thank you very much @droftarts. Again, it matches my data, so I know what I've done is correct!