Warts and Seams

[jimc]

that is unreasonable. you could be severely over extruding. what is your ext width set to? did you mic the filament and get that value set? how about your multiplier setting? the jog dial on the control panel...is that set at 100%. when you extrude 2 lines side by side by side the edges should just touch. the nozzle should not be pushing plastic around. do you have any pictures you could post?


the e3d and v4 arent your solution. they would be a nice upgrade at a later date but there is something else going on. you should be able to get very nice prints with the v3b. its a proven design that the m2 and many other printer companies use so 1000's of people are running a v3b or a hot end of similar design.

[pyronaught]

that is unreasonable. you could be severely over extruding. what is your ext width set to? did you mic the filament and get that value set? how about your multiplier setting? the jog dial on the control panel...is that set at 100%. when you extrude 2 lines side by side by side the edges should just touch. the nozzle should not be pushing plastic around. do you have any pictures you could post?


the e3d and v4 arent your solution. they would be a nice upgrade at a later date but there is something else going on. you should be able to get very nice prints with the v3b. its a proven design that the m2 and many other printer companies use so 1000's of people are running a v3b or a hot end of similar design.

Ext width is manual .35
Ext multipler is .9
jog dial is 100%
filament miced at 1.75 pretty consistently and is set to that on the Other tab

a wall width of 1.17 in the CAD is printing at 1.38 though, using three parallel lines to produce it.

The print below is the best attempt so far using my test model which checks both curved and flat walls as well as cutouts. The flat walls actually don't have any stop points in them except for one or two at the very edge, so it doesn't really check for warts on non-curved surfaces very well. Coast is 7mm, retraction is 3, M201 acceleration is set to 900, print speed is 4800, outline underspeed is 50%, layer height is .2 and wipe/restart distances are turned off because they were not having any effect. Wipe would usually cause a set of two bumps separated by the wipe distance instead of one, so it actually made the problem worse. One interesting thing in the window cutouts is that on the right edge you can see the blob problem effecting the edge, but on the left edge it does not happen. A reversal is happening at each edge, so I don't know why there would be a blob going one direction but not the other.

bump_test.jpg (330.42 KiB) Viewed 10919 times

[insta]

Go download slic3r, and the 20x20mm calibration cube from Thingiverse.

Create a new profile with 1 perimeter, zero top, zero bottom, zero infill, 0.2mm layer height. Leave everything else default. Mic your filament perfectly (measure 8 times in 4 different spots, rotate calipers 90 degrees), throw out the highest and lowest numbers, average the rest. Set the extruder multiplier to 1.00 and the filament diameter to 2 decimal points. Slice the cube, which should generate a single walled part.

In the GCode there will be a comment towards the top that says something like "; Perimeter wall thickness = 0.35mm". Measure the cube wall (again 8 times, but in 8 places, average all 8) and compare it to that wall thickness. The ratio between them is how far your esteps are off (for this filament...), which is the last thing we haven't considered. If you measure 0.44mm of 0.35mm thickness, you are extruding 125.7% of the amount of plastic you should be, so your esteps should go down by (100 / 125.7), or 79.5%. Take that number (with the values YOU calculate) and multiply it by the current esteps value in firmware, then save it and run the test again.

Once it's dialed in, you can then use the extrusion multiplier for the same effect. Run this same test again, but re-slice the gcode with different extrusion multipliers when switching to different plastics.

As an example, when I ran this test, I only did it with extrusion multiplier to begin with. I noticed all my PLAs were like 0.8789, 0.8801, 0.8705. So I lowered the esteps to 88% of their current value, and then my multipliers were a much more reasonable 1.01, 0.99, 1.02 or whatever.

[Dale Reed]

Ext width is manual .35
Ext multipler is .9
jog dial is 100%
filament miced at 1.75 pretty consistently and is set to that on the Other tab

Using the default 0.35 mm nozzle, it's actually difficult to produce an extrusion width of 0.35 mm. Auto width usually puts the width at more like 0.42 mm. I would expect an extrusion width of 0.40 to 0.45 mm to be more reasonable, if you want to set it manually. Using 0.40 mm as the width, you should be able to produce a wall thickness of exactly 2.0 mm with five lines of filament.

In addition to insta's suggestions, please change the extrusion width to something more in the range given or set it to auto. Then try his calibration procedure.

Dale

[pyronaught]

So when you say modify esteps in the firmware, are you talking about the steps per mm assigned to the extruder? Is the way you are changing that by sending an M92 command from the communication tab in S3D or is the Arduino IDE required to make header file changes, recompile and upload?

I just noticed that the firmware settings dump that occurs when connecting to the machine also tells you if you have a 24v machine or not.

[ednisley]

The ratio between them is how far your esteps are off

I highly recommend computing the number of steps per millimeter of filament directly, by measuring the amount of filament moving through the drive gear, not indirectly from the width of the extruded thread. That way, you have an actual number that represents a measurable quantity, not an arbitrary value based on other factors.

Remove the hot end, put a mark 150 mm from the top of the extruder, manually extrude 100 mm of filament, measure the distance "x" between the mark and the extruder, and compute the amount of filament that actually went through: 150 - x. For example, if there's 60 mm remaining, then the extruder actually moved 90 mm of filament.

Then adjust the steps per millimeter value: new step/mm = (100/90) * old step/mm.

Iterate two or three times to get a step/mm value that you can verify in the future when something odd happens and the extruder motor starts turning twice as fast as it should.

That actually happened to me (and others): the old Thing-o-Matic had Allegro stepper controllers that failed by shifting permanently from 1/16 microstepping to 1/8 microstepping, no matter what the DIP switches said. Knowing that the extruder suddenly started running exactly twice as fast as it used to was a big debugging help!

The Extrusion Multiplier is the only arbitrary value in the whole collection of actual measurements. Normally, it will be around 1.0, but that's not a hard and fast rule. For example, the default Makergear firmware value fudges the step/mm value to force the visible value.

More details on the measurements:

http://softsolder.com/2013/05/07/m2-vs- ... culations/

Now, if you wind up with a truly offbeat Extrusion Multiplier, that may point to a problem elsewhere in the setup... but you should be able to find it, because all the other values come from actual measurements.

[pyronaught]

I agree that the steps per mm setting for the extruder should be just that-- the number of steps required to physically move one mm of filament through the extruder and not fudged up or down to compensate for anything else. Let the extrusion multiplier do all the fudging, even if that means it isn't always hovering around 1.0.

[jimc]

pyro, do you still have your settings that you used for the pic above? if so can you export it, compress it as a zip then post it up here so we can see where you had everything set for that model? alos not sure if you are aware but rsilvers posted his profiles a long time ago. alot of people say they have had good luck with them or atleast a real good starting point.


http://www.thingiverse.com/thing:269637/#files

[pyronaught]

My measured value for extruding 100mm was consistently 98mm, so the estep multiplier would be 1.02.

Filament diameter averaged around 1.74mm

Set screw on the extruder pinion gear was tight.

factory file attached.

[msmollin]

Any chance it's maybe the filament? Have you tried different filament? I recently switched over to some silver PLA from Makergear and it's printing beautifully using mostly stock PLA profile in S3D. I just upped the extruder temperature to 225C and dropped the bed temp to 50C. Rounded faces, cones, and overhangs looks great, and where I can find the seam they're very minimal. Any chance you can link to your models? I'm curious to try out and see what's happening.