My new filament yields 0.2mm short in every dimension of this 20x20x10 calibration cube. What could be the cause?
Here are the perimeters of the problem. My Ares 3D printer was shipped with delta rod shorter than is set in firmware. With the originally shipped filament, the calibration cube came out ~1mm short in every dimension. Then I fixed the firmware to match the rod. The cube came out within 0.1mm.
I knew that the new filament was 50µm thinner than ø1.75mm when I saw the first calibration cube. I thought the 0.2mm shortage was due to under-extrusion because my slicer was still set to ø1.75mm. So I adjusted slicer to match ø1.70mm. Under-extrusion is largely fixed. (Sign in - Google Accounts) But the new cube still came out 0.2mm short in every dimension, no more, no less.
My next thought is thermal contraction. But shrinkage should be proportional to the dimension in such amorphous materials.
Is there another firmware (Marlin) setting that needs adjustment?
Calibrate your e-steps and adjust extrusion rate in slicer
Thanks, @Hoang_Nguyen . Is e-steps something in firmware? I’m pretty new to these terms.
I suspect that extrusion rate would not affect build because adjusting for filament diameter is effectively an adjustment of extrusion rate, and that did not change build size, at all. (Down sizing filament in slicer by 2.8% is a 4.7% increase in flow rate; as is, the build size is just 1% in horizontal dimensions and 2% in vertical.)
esteps is in firmware. If your new it’s probably easier just to play with the extrusion “multiplier” in your slicer. Try increasing it and print out the test print again and see if it improves.
Try measuring a move (manually) with thaz calipers. Best over long distance like almost full platform length. That much out of underextrusion and you would see holes in a top infill. If the height is of by 0.2 on a delta its certainly not e-steps or extrusion multi.
@VolksTrieb That’s what baffles me, too. On the other hand, the previous filament gives at most 0.1 discrepancy. (In some builds, the caliper cannot detect even 0.02 discrepancy.) So properties of filament do play a role.
If it’s the filament not filling out like it should, then you need to adjust the flow rate. Other clues is if the top layers don’t pack as well as they used to. You can often adjust the flow in the LCD panel on a printer if it has one. The computer host might have a flow rate adjustment slider or dial. There is even a flow rate adjustment g code you can enter if you have a console window to the printer.
Print a larger cube. If error is the same it’s backlash potentially. If the error is larger could be thermal contraction or esteps. To rule out esteps do what @VolksTrieb suggested.
Edit: nevermind Esteps, just saw it was a delta. Delta’s have their own nuances.
It makes little sense to randomly twiddle with flow rate until you’re certain of your movement accuracy, which you won’t get the way you’re measuring.
An easier way to do what @VolksTrieb suggested is to just print two of the same object spaced apart, say two 10mm cubes (or cube-holes) 40mm apart and then measure the distance same-edge to same-edge.
If you have an extrusion problem, it should manifest on both objects and cancel out. If you measure 40mm same-edge to -same-edge, then you can move on to twiddling with flow rate or whatever.
On a Rostock Max (a similar delta printer) this is adjusted via the “rod length” in the firmware. Ive had to tweak it too when making up custom magnetic rods.
Is that one of those plastic bodied calipers? Avoid at all costs. Too much frame flex to trust any measurements you get from them. Save up for at least halfway decent measurement tools. Sorry, my inner metrologist is showing.
@Richard_Kranium That’s a good suggestion. I tried two 10x10x20 towers spaced 20mm apart but with no base. Because I didn’t not put them on a base, resultant distance is even harder to measure. From what I can tell on the blue tapes, the distance is quite accurate. (30mm - really rusty 3D design here.) The height, though, again sees a 0.1 - 0.2mm shortage despite double the height. I think this means the mechanics and firmware are OK, but I can add a bit more flow to it. (Another mistake in the quick test is to use 12% infill - same in the new calibration cube test.)
When I get around to test again, I’ll use a larger distance, with a base, and 100% infill.
@Alan_Devine I have already adjusted for rod length with the previous filament.
@whackyhack as @Eclsnowman has suggested, try a larger print (not necessary a cube, just anything that has a decent volume for a stable print but with a dimension 2 or 3 times this one). Knowing whether you have a 1% error, or a 0.2mm error is going to tell you a lot about where the problem lies.
@Mark_Fuller 's remark about measurement is on the mark: I was contending with tens of microns when the caliper’s precision was one hundred microns. Additionally, uniform discrepancy in width and height can be the result of systematic errors in measurement.
So I bought a caliper with 0.01mm precision - which I didn’t think existed because my concept of caliper remained in the vernier age. (It is also stainless steel body.) Surprise! The GP3D filament’s diameter is also on the mark. Not only that, but the cubes are not short in any dimension when measured with the new caliper. (Sign in - Google Accounts) So back to drawing board to fix underextrusion.
