I think my setup is extruding too much, which is jamming up the nozzle. It runs reasonably well for 2 loops prior to getting to Benchy, then it starts to jam shortly after, and all extruding stops. The stepper keeps spinning, and cutting into the wire, making a mess. When I remove the wire, it’s frozen in place, and has a slightly larger shape, as if it pushed the melted material back up the nozzle and re-froze. The test-IE-part gcode example runs the longest, I think because the extruder stepper never reverses, but eventually jams as well. The nozzle ends up running thru the material already extruded on the heat bed, eventually jamming and stripping the wire.
The recommended firmware DEFAULT_AXIS_STEPS_PER_UNIT {200,200,4535.44,200} were perfect for X, Y, and Z axis. I tested each with a ruler visually, and the movement was just as expected. I tried 200 for the extruder stepper with no luck. I tried 100, 50, 25 and still think there’s too much feeding. Rambo 1.4 3dprinting firmware.
I’m using prusa slicer on linux. I copied all of the settings from the import-extruder page when setting up the Benchy gcode. 100 feedrate on repetier.
I’m using several year old inland PLA, so I ordered some new hatchbox PLA to test as well. Temp at 230c per the max recommendation on the box, 60c on the heated bed. 1.75mm wire, measured with calipers to confirm.
Is there a way to slow the extruder stepper down more? I saw some references to jumpering extruder driver, or 1/32 microstepping mode, but no details how. I’m missing something here.
I tried the 215 and 202 for temp. It extruded, but crashed earlier on both prints. Same issue.
How wide should an extruded line be? my first layer is 1.65mm wide.
How can I tell if it’s set for 32 microsteps? There are some vague references for the ramps board, but I’m using the full size rambo with new firmware from @Jeffeb3
I guess if it’s 32 times too big, then setting the steps per unit to 200/32 = 6.25 might work. I’ll try that now.
I tried the extruder steps per unit of 6.25, 1.0, and 0.25. The lowest rotations it will go is 1/4 turn per 1 unit click in repetier. No change in performance.
I just ran the hatchbox pla at 200 steps per unit, and 200C per box. It seems to work better than the inland pla, but it still ran into the same problems.
It still looks like over extruding. The nozzle ends up pushing thru the first layer. Eventually it fails to continue feeding.
Or it’s failing to move the extruder up vertically to compensate, but up on the Z axis moves the extruder up.
The layers are supposed to be 0.2752mm, and that might be true for the perimeter layers, but everything else is a full millimeter or more. By the time it appears to be the second layer, the extruder is running into the first solid layer, noticeably shaking the machine.
I removed the heat bed and ran it on the table top surface, which should be mostly level and very solid. Same issues there as well.
I asume you use either Rambo 1.3 or 1.4, so 1/16th microstepping. Thus steps per mm should be 100.
Perhaps try calibrating your extruder. Thom has a nice video on this:
I have no idea why this hasn’t been updating when I reflash the firmware. It always shows successful. I always reboot after flashing. It should retain the settings.
I’ll use the M92E100 command to update, and M500 to save. Wow it is super slow at this setting. Thanks for the help!
Marlin does not know it has been flashed. So if you flash it with the same firmware, it will trust the value in eeprom over the default firmware. You can set it with M92. Then save it to the eeprom with M500. That will make it stick while it is powered off.
If you want to force it to be the firmware settings instead of eeprom, M502 will clobber the current setting with the firmware ones. Then save to to eeprom with M500
