Extruder Steps not being honored

brandonm4

Hello,
I'm trying to calibrate the extruder extrusion amount. No matter what I set the E value to, I extrude about 23mm instead of 20mm when testing.
I have this line in config.g:
M92 X80.52 Y80.78 Z400.00 E329.55 ; I started at 420 and kept doing the 20/ExtrudedAmt * CurrentStep and then saving/resetting/retesting and nothing changes.

I have a Duet 2 Wifi - Firmware 2.02(RTOS).
I do not have a Zprobe, filement sensor or anything like that. This started as a super cheap Greeetech printer that I've slowly been replacing parts on.

The Extruder is a printed on from here: ww.thingiverse.com/thing:1383913
The hotends are E3D genuine .4mm

Please let me know if you need any other info or if I just missed something obvious.

Thanks,
Brandon

Phaedrux

A few things I can think of.

It helps to have as close to correct starting point, which can be calculated with

e_steps_per_mm = (motor_steps_per_rev * driver_microstep) * (big_gear_teeth / small_gear_teeth) / (hob_effective_diameter * pi)

Then for the actual extrusion test are you extruding it through a nozzle and hotend, or with the hotend removed?

Either way is fine, but if you keep the nozzle in place it's important to use a high temp and very slow extrusion speed so that there is no back pressure to cause filament slippage.

It's also a good idea to extrude a longer amount filament. 20mm is pretty short. I would mark 110mm from the point the filament enters the extruder body, and then extrude 100mm of filament at 1mm/s. Then measure the distance of my mark from the body of the extruder. Then use:
new e steps = old_e_steps * (100 / (110 - distance_to_mark))

Phaedrux

Also, what microstepping are you using, and what motor current for the extruder? Improper values for those could also come into play.

For microstepping I would suggest x16 with interpolation and for motor currents its generally recommended to target a range of 70-85% of the rated max current for the motor.

brandonm4

Thanks - upping the temp and slowing the feed rate as well as starting at a much closer number for the initial attempt helped quite a bit. I was able to get it much closer to true.