Bad surface finish on the glass side with S3D & ABS ?

[Mind]

Hi All,

Just got my M2 after 4 weeks of waiting (YAY!). My prints need to go underwater, so they have to be ABS, not PLA. I am using S3D. Klapton tape and ABS goop. Sticks so hard, I need a light tap with a hammer to break the bond!

Problem: I am not getting 100% infill. In fact, I am getting a wierd pattern on the layer that goes onto the glass. See pic

Settings:
100% infill
0.1mm layer height
Nozzle DIA: 0.35mm
Extrusion Multiplier: 0.9
Extrusion Width: 0.4mm (measured)
Filament: 1.75mm (measured)

Bottom that is on the glass. Also, the first few layers are messed up... rough edges etc.


Other side of the part. Note the outside finish is pretty smooth, but not perfect! Sorry it is slightly out of focus (macro with a cell phone is horrid!)


What did I change incorrectly ? Should I bump up the extrusion multiplier back to 1.0 ? Won't that cause ridging/etc and make the outside surface finish worse ?

[jimc]

the odd pattern on the glass is because you dont have the bed gap set correctly. the first layer is being printed too far from the glass. the other is hard to see but appears to be over extruded outer layers and or blobs. the pic is a little fuzzy so i cant see exactly. check filament doa, lower your multiplier. if using s3d add some coasting. it can also be from damp filament as well.

[ednisley]

Also, the first few layers are messed up... rough edges etc.

They're too sparse, because the first layer had too much vertical space. The nozzle will move vertically by the correct distance between each layer, so, after a few layers, it'll start looking pretty good... but the first few layers will be rocky as a fixed extrusion volume must fill too much vertical space.

The bottom of the overhang looks underfilled, although that may be an artifact of the support structure. How did you measure the 0.40 mm extrusion width?

Using 100% infill demands absolutely correct calibration of everything. You'll get better results with 90% infill: the infill won't be totally stuffed and the perimeter quality should improve.

[Mind]

the odd pattern on the glass is because you dont have the bed gap set correctly.

Exactly what should the gap bed be set to ? The only reference I can find to this is the M2 setup video where they show using a bent business card. That is pretty sloppy and open to interpretation (not to mention business cards come in different thicknesses!)

How did you measure the 0.40 mm extrusion width?

The 0.4mm width of the filament was measured by extruding some in the air, letting it cool, then using a digital Vernier caliper.

[ednisley]

Exactly what should the gap bed be set to ?

I set it to the same 0.20 mm that I use for the thread thickness, so that the first layer comes out exactly the same as all the others, but opinions differ. In any event, you're absolutely right about the business card thing; a round spark plug gap gauge or set of feeler gauges will work much better:

http://www.amazon.com/s/ref=nb_sb_noss_ ... plug%20gap

You must set the first layer height with everything stabilized at operating temperature, a process that involves fiddling with dangerously hot hardware. I do a coarse room-temperature setting, print a set of four thinwall boxes (one in each platform corner), measure the height of the single-layer skirt around the outside, and adjust the platform height accordingly.

The skirt also provides a quick check of platform alignment: if it's 0.20 ± 0.05 mm all around the platform, I know the first layer will be very close to spot on.

extruding some in the air, letting it cool, then using a digital Vernier caliper.

Unfortunately, that diameter doesn't equal the width of the actual thread as it's laid down in the object: you must measure the plastic thread in situ after it's extruded. A recent discussion in another thread goes through the steps:

viewtopic.php?f=11&t=1893#p8482

The single-thread walls in the thinwall box tell you what you need to adjust the Extrusion Multipler.

As a bonus, the box height lets you cross-check the first layer height: a 5.00 mm tall box will have 25 0.20 mm layers. If it's not exactly 5.00 mm tall, the difference tells you how far off the platform was for the first layer.

All this sounds really tedious, but changing a distance by 0.05 mm matters...

[jimc]

yup ed is right on. i use a set of feeler gauges and have found .005-.006" seems about right. if you dont have a set you can grab em from your local auto parts store for a couple few bucks.

[jsc]

Just a note on Ed's claim that 0.20mm should give the same gap as all the other layers:

In fact, if you look at your generated gcode, you should see that the first layer is extruded at a height of 0.20mm, not at 0. So it is in fact theoretically correct to set up for a zero gap clearance (when hot). In practice, I've found that jimc's recommended gap of 0.006" (cold) works fine for me, and I don't know what that translates to at printing temperature because I don't like to fiddle with things when they are at burn temperature.

[ednisley]

the first layer is extruded at a height of 0.20mm, not at 0

Good catch.

I've set my startup G-Code to put Z=0 at the platform. All the layers are 0.20 mm thick, which means the G-Code shows the actual height above the platform and the first layer comes out spot on.