Massive under extrusion

[andyfarthing]

Beta 3 - seems to have a problem with under-extrusion on prints, but seemingly at random. Tried about 6 or 7 prints, and only 2 succeeded.

I spent a day playing around with settings both in firmware and on Cura, but nothing seemed to have any effect. At random, a print would turn out okay even though I didn't make a change to any settings (just playing around with bed level).

I can push filament through manually fine, so pretty sure there isn't a blockage. Hitting the 'extrude' button in OctoPrint also works fine - it just seems to be during an actual print.

Reverted back to the original Wanhao firmware and problem went away instantly.

[Silverquark]

Hm weird. Maybe a coincidence? Havent heard that from anyone else and never experienced it myself. The actual print Code also didnt change at all from beta 1 to beta 3. When does The print fail? On first layer? If so then it's probably just a leveling issue

[andyfarthing]

It's possible - I'll run a few more prints on stock to see if it pops up again. If I can't reproduce it on stock then I'll flash back to this and see what happens.

They all failed on the first layer, so you could tell straight away. It looks like it would with a clogged nozzle, but I don't see how I've managed that if it is! I've ordered a new nozzle to try, just in case.

I did see a mention on the Facebook group that somebody else had an under-extrusion problem after flashing, which is why I thought it worth mentioning. They didn't comment with a follow-up though, so don't know how that progressed.

I'll run some more tests tonight and let you know how I get on.

[Silverquark]

Its probably just the nozzle being too close to the print bed.
Where did you see that?

[miikael]

I have somewhat similar problem.
I printed few test items and saw that there were some strange problems in prints. Sometimes underextruding and sometimes overextruding (left blobs of filament in random places).
So i tried to to e-step calibration. Sent over octoprint the command to extrude 100mm of filament, but it actually extruded only about 40 mm. So calculated new estep with the new result.
After applying the new estep number, i got almost same result with the command to extrude 100mm of filament. (little bit more filament was extruded, but not much)

The extruder gear is not skipping also, looked to that too and also manually pushing the filament, it comes out fine.

[andyfarthing]

Facebook comment here: comment

As an update - after a few successful prints the same problem has now popped up in the stock firmware, so I'm not really sure what's going on. I've triple checked the bed level, replaced the nozzle, PTFE tube, tried different temps, different slicers etc. and no change.

Does anything carry over in the EEPROM when flashing between the stock firmware and this one? I'm just wondering if I've got some gremlins somewhere that are lingering... I've tried the factory reset but that hasn't made an improvement.

I'll try doing the e-step calibration as above in a bit, and see what result I get.

[Silverquark]

Hm weird the only thing that may get transfered between flashes is preheat preset values. But they are stored outside of marlins normal save/load range. So that shouldn't interfere with anything.
Can someone with this problem try sending a "M221 S100" and see if that changes anything?

[andyfarthing]

I flashed back to this firmware (thought I may as well) - the first few prints turned out fine, but then the problem appeared again. The only change I made between prints was to run the PID autotune - could that have anything to do with it?

Ran the estep calibration and had the exact same result as miikael - only about 40mm was extruded. Changed the steps to 118 (which is what it calculates as), and a tiny bit more than 40mm was extruded.

Interestingly, if I run the "G1 E100 F50" command (from the wiki) to extrude, the motor runs backwards and ejects the filament. Any idea why this might be?

Running "M221 S100" doesn't have an effect, unfortunately.

[Silverquark]

That's really weird. Autotune shouldn't have an Effect on it at all. All it does is run the M303 autotune command with the given parameters.
M303 C10 S200 U1 would be the command for 200°C.
About the wrong Extrusion. If you are in absolute move mode and you give The command G1 E100, it will not extruded 100 but rather try to go to position 100, just like x and Y axis would do. Try doing a G92 E0 before. That will gell the extuder he is ar position 0. Or switch to relative Mode with G91

[andyfarthing]

It might just be coincidence, but that's the only change I made.

If I run "M221 S350", prints look fine (maybe a little over extruding, but certainly closer to what I expect). If you look at the image below, the top part (and bottom outer part) is run with "M221 S100", and the inner bottom with "M221 S350". The two blobs are where I paused it - ignore those!

[andyfarthing]

Running the G92 E0 command before extruding sorted out the backwards motor, thanks. It still only extrudes 40mm though!

[Silverquark]

Im currently working rebasing the firmware to a fresh marlin 1.1.3 build. No idea if this can fix the issue. But it will definitely clean the code up a lot and may fix some things I forgot to add from the original marlin source. Once I got it ready I will provide a download. Should be tomorrow or the day after. Let's hope that fixes it

[chschmit]

When will Beta 4 be available?
I'm running Beta 3 and still having underextrusion issues like mentioned above despite resetting all of the EEPROM settings that were used prior to Beta 3.
I've even gone back to the original Wanhoa firmware and the EEPROM settings work just fine. I've since gone back to the Beta 3 firmware and LCD update the underextrusion is again happening.

[Deneteus]

Any updates on this under extrusion issue? I noticed it was under extruding and found your Github while looking for a solution.

[johnboiles]

FYI Cura has had an issue since ~2.6 where the filament diameter defaults to 2.85 even if the machine definitions specify 1.75 Ultimaker/Cura#2270

After a Cura update I had issues that looked exactly like @andyfarthing's but it turned out to be a Cura problem. When testing between i3PlusPlus and Wanhao firmware, make sure to use the same gcode.