direct drive extrusion and consistent flow

I'm having a problem with my direct drive extruder.

Printer specs
- I'm using an mk7 and direct drive, stainless steel from qu-bd
- nema-17 stepper, 5mm shaft ( ordered from makerfarm, pretty sure this is the source [www.circuitspecialists.com] )
- Merlin firmware, current-ish from git repo ( ish meaning about 1 week ago I did a git pull )
- Repetier host mac, 0.56 build 53
- slic3r bundled with repetier, 0.9.9
- Matterhackers 1.75 light blue abs - [www.matterhackers.com]
- 235C extrusion temp
- 110C bed temp
- extruder configuration [www.thingiverse.com]
- Note, there is a non-adjustable spring pushing the bearing against he filament
- e3d direct drive, 1.75mm and 0.4mm nozzle

Following the instructions at [reprap.org] and I got an interesting issue. What I see when printing 20x20x10 cubes is somewhat like a tiny bumps in a banding pattern. As I tuned the e steps the bands got tighter, then suddenly changed direction. The extrusion diameter is approx right ( slicer says 0.67, I'm getting 0.68 - close enough for me or should I try to get it exact? ).

When I compare my printed parts against parts from other printers I can see little bumps. While watching the prints, I was able to correlate the bumps to extruder pulses - basically As those pulses occur I see the bump, then it thins out and repeats. Its likely back pressure. I had thought the idea behind e steps and the extrusion multiplier was to basically tune the extrusion based on the filament ( 1.75 abs ) and the nozzle ( 0.4mm )

I have included links to images showing successive prints after tuning the extruder steps.

[dl.dropboxusercontent.com]
[dl.dropboxusercontent.com]
[dl.dropboxusercontent.com]
[dl.dropboxusercontent.com]

What I can't tell is if something I'm doing is wrong, something with the hot end, extruder or both/all?

The parts are coming out almost the right dimensions. The 20x20x10 is coming out 20.03x20.03x9.97 so its a little bit off. I'm not able to measure the difference in width between the bumps but my guess is there is a little extra ( .2mm or something, not a lot )

I have checked the bearing, it rolls very smoothly and without force. I do not hear anything funky or high pitched from the extruder motor. I have checked the gear, I don't see any ground or chewed filament - little fleck or two here and there but nothing major.

Any ideas?

Edited 1 time(s). Last edit at 03/09/2014 05:46PM by bilsch.

Hello bilsch,

Well I see a lot of views on your post, with no one chiming in, so I will give you my take...

Interesting print pictures (quite an assortment of perimeter print defects)... It looks like Infill pattern types on your perimeter side walls, with the diagonal lines, crossed diagonal lines/rectilinear lines, honeycomb, and archimedean-cord patterns etc. on the perimeter wall sides of your prints. Wow, your perimeters are all over the map.

Barring any unknown mechanical issues, and seemingly that you have checked your axes for smooth movement and alignment, and assume you are not trying to print at some ridiculous print speed >40mm/s that may generate harmonic vibrations/ringing/etc. in your machine that could represent itself as some weird repeatable pattern in your printed perimeters. I believe your issue is directly related to the same old problematic slicer, Slic3r (any version beyond >Slic3r v7.02b is TOTAL GARBAGE !!!). Not sure about the MAC Slic3r versions, but the PC versions are a mess !

Check your gcode output from Slic3r with your favorite visual gcode viewer, or try Web Based gcode Viewer , mostly likely your issues are visible in a gcode viewer, as produced by Slic3r, a total garbage slicer. PLEASE read my little rant on Slic3r... Out of Control bug list, time to get a clue....

My initial take to your issue is after viewing your print pictures, somehow Slic3r is translating your set infill pattern to the perimeters (90 degree off axis).

Curious as to your findings as to what you found viewing your Slic3r gcode output.

Edited 1 time(s). Last edit at 03/10/2014 01:20PM by ShawnT98027.

Easy way to see where the problem is coming from - try another slicer such as Cura.

Also - you should be calibrating your extruder by measuring the length of filament pulled into the extruder. Measuring the width of the output is not the most accurate way to get precise flow control.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Thanks all for the responses.

So, one thing I did by accident. I had thought I should measure the amount of filament extruded through with the nozzle off which is how I had done the tuning. When I re-assembled the hot end I forgot to tighten things up all the way and ruine the heat block/heat break. Thankfully, the e3d is modular so I'm only replacing those pieces vs the whole damned thing ( about half the cost, plus filastruder has the parts vs getting them shipped across the pond! )

I was wondering about the slic3r behavior and had been working on kisslicer configuration but have been trying to keep the changes contained to the same components/products. No sense in just thrashing ( which I do anyway I guess )

Anyway, the replacement parts should be in late in the week. First thing I'm going to try is to continue with the repetier/slic3r combination. Maybe there was a partial clog in the hot end somewhere? Mis-aligned or something internally maybe?

While I have the hot end mostly apart I'm going to re-do the e-step and then work on the fine tuning with the replacement parts. Will work my way out from there.

I loaded the gcode from the last print to the gcode viewer from ShawnT98027's post. I need to look a little more closely. I could see the lines but I was having trouble seeing if they were straight or bumpy.

The print speed I've been using is 30mm/sec. It seems that if I go really low ( 5mm/sec ) it exacerbates the issue. This seems to suggest harmonics may indeed be at play here, but could also be hotend related.

Will post back later in the week with results. I'm guessing it'll be friday or so.

What you've described and what appears in the pictures could be caused by the stepper motor configuration for the extruder stepper. It looks as though the stepper is not microstepping properly.

Some questions:
1) What stepper motor controller are you using?
2) What is the number of microsteps that you have configured on your stepper controller?
3) What is the current that you have set on your stepper controller?
4) What voltage are you using to drive the stepper motor?

Sorry for such a delay in responses!

Ok, so to answer madscifi
#1: Pololu A4988 , ramps 1.4 board
#2 1/16 steps on all
#3 I have not messed with the current, trimpots are whatever they were when I got them. I'll have to check with a multimeter
#4 12 volt

I checked the voltage at input and it was dipping down a little bit below 12. I've adjusted the pot on my power supply and its now delivering 12 volt without issue. The ampere capability of the power supply is 30A at 12 volt ( [www.lulzbot.com] )

Quote
madscifi
What you've described and what appears in the pictures could be caused by the stepper motor configuration for the extruder stepper. It looks as though the stepper is not microstepping properly.

Some questions:
1) What stepper motor controller are you using?
2) What is the number of microsteps that you have configured on your stepper controller?
3) What is the current that you have set on your stepper controller?
4) What voltage are you using to drive the stepper motor?

Ok, so now for the long list 'o stuff I've been trying over the past few weeks!

I thought maybe the issue was with vibration caused by a slightly loose fan shroud. I have printed a new one which snugs up against the bottom of the extruder mount ( [www.thingiverse.com] ). Other than any vibration caused by the fan or movement during printing it should be pretty stable.

At the local 3d printing meetup, there were a few theories:
1) the extruder spring felt a little loose. I printed a little shim to tighten things up, things improved a bit
2) Lots of playing with e steps and speed/flow rate via repetier

We went from 100% on speed and flow to various configurations ( increase speed/decrease flow rate, decrease speed increase flow rate etc )
In the end we manually tweaked the flow rate to a point where the angular pattern went away - now it is almost vertical.

Follow-up tuning at home I managed to get the e step value right about where we were doing it on the fly ( photo attached from a debug ). I can't seem to get much further - any 3 step tuning even at fairly small incriments ( 1/2 % step tunes each time ) and I start getting the pattern following the angular type thing.

The guy I purchased the printed parts found a very interesting link on a solidoodle forum and it looks like there is another experiencing very similar effects to what I am. Sadly, I don't think I can get 1/32 on a ramps 1.4 board - anyone know if otherwise?

[www.soliforum.com]

Well, that was it - funky pattern is gone!

Quote
bilsch
The guy I purchased the printed parts found a very interesting link on a solidoodle forum and it looks like there is another experiencing very similar effects to what I am. Sadly, I don't think I can get 1/32 on a ramps 1.4 board - anyone know if otherwise?

[www.soliforum.com]

So you fixed the problem by switching the driver board for one that support 1/32 of a step? Or something else?

madscifi,

No I have not changed the stepper driver chips. I don't even know that I could - ramps 1.4, only have jumpers for 1/4, 1/8, 1/16 ( [reprap.org] ). I think I'd have to get a smoothie board to go 1/32 or 1/64

So, changing the e steps until the bumps were forming a vertical pattern. Then I started adjusting the pot while printing a 20x20x100 hollow cube ( made it high so I could test more changes quickly ). I now have almost no detectable pattern - little bit of tweaking left there, but you have to look really close to see it and at least I can't feel the difference.

I did not change the slicing application. The effects I was seeing indicated likely a mechanical problem of some kind which it was. The one thing I have noticed ( and expected ) is that the filament drive should always be turning ever so slightly while extruding. What I noticed and heard were short pulses and they lined up with the bumps. Now that I have made the changes I see exactly what I had expected - slow and steady turning of the drive gear

Tuning 3d printers seems something of a fine art Lots of options and easy to screw up

Quote
madscifi
So you fixed the problem by switching the driver board for one that support 1/32 of a step? Or something else?

That makes sense. Thanks.

Quote
bilsch
Sadly, I don't think I can get 1/32 on a ramps 1.4 board - anyone know if otherwise?

Hi,

I did change my pololu A4988 on ramps 1.4 to DRV8825 at 1/32 microstep. You can use the same jumper on the ramps to decide of your microstep ratio. Take care to put it in the correct orientation, the trimpot is located on the opposite edge of the drive.

It doubled my resolution as I can see it and reduced my moire problem. Also the stepper motor run quieter. This drive can handle higher current than the A4988 without heatsink (1 amps vs 1.5 amps). That compensate for the lower torque of the higher micro-steps. In addition this chip manage better the decay than A4988 depending on your stepper (noise and moire effect).

In my opinion, you need to put your hands on and adjust the trimpot. It is definitely fine art to ajust the vref voltage, low enough to get smooth motion and low noise and high enough to prevent skipping steps.


Good luck with your printer

---

solitutopia.com

Edited 1 time(s). Last edit at 04/09/2014 03:12PM by SolidUtopia.