Hi all,
I have been absolutely happy with the Vertex so far and I’m trying to push it a little further every day :).
Today I tried to print the Maker Faire Robot (65pct) url=http://www.thingiverse.com/thing:331035/#files[/url].
This model is printed all in one go. I tried several settings (slow speed, slightly altering the width setting for filament) but I always run into this issue: where I need to see a small gap (see screenshot), the walls touch just a bit which then results in a robot that cannot move its joints anymore.
I have measured this with other prints as well, holes are just a bit too small, while the counter parts are just a bit too large.
We’re talking about just a couple of hundreds on mm here.
If you look at the other side of the print, it doesn’t look that bad at all I would say:
I there a way to improve this or am I just simply hitting the boundaries of the printers capabilities?
This one is printed with a 0.2mm layer thickness, 20mm/sec speed, PLA on 200C.
I noticed something similar when trying to print objects with screw tunnels inside.
The problem is that the while the outer dimensions are exact the inner dimensions of the tunnels are too small and I had to drill them out for the screws to fit.
This problem depends on the actual size of the “inner” objects. For example at 2,9mm(in blender) I got about 2,54mm tunnels on the printed object. And the bigger these screw tunnels I made the less I had to adjust/correct it.
I haven’t googled it yet but I think it must be a software bug(Cura, Firmware, &)
Try ev. to reduce the nosle-temperatur, so the gape between the movable parts is enough wide … I think it’s to hot
-> see my post with same robot at 185°C http://forum.velleman.eu/viewtopic.php?f=64&t=13281
I think that in my case the thermistor is indicating too much!
I measured the real temperatur with my ir-camera and a adhesive tape on the brass heating block around the nosle … but difficult to measure precise!
Try to decrease the temperature by little steps …
[quote=“raby”]It’s indeed a slicer issue. Especially with round holes.
You could try to lower the flow and increase the speed (just a tiny bit).[/quote]
This is discussed in the literature. Holes are approximated with polygons, and always are sliced to be too small. I don’t understand enough about the problem to know why it cannot be easily fixed
Well, that turned out quite nice!
I started a few prints earlier to see what happened when I reduced the flow and found that with 70% it still printed nice solid areas which didn’t seem to touch. With these adjusted settings I printed the full model again. Layer thickness of 0.2mm, 25mm speed, 70% flow (1.75mm filament diameter), 200C.
The quality is really much better, moving parts are moving and the overall quality increased as well.
I see some small issues but it’s getting better and better every day :).
Thanks again for the advise!
Great prints and thanks for sharing your settings. Surprised that flow rate could make such a difference. I want to try the mini car model from le fab shop that has the same issues with moving wheels.
It is a combination of many things. I just tried to lower my temperature from 200C to 190C while keeping the flow to 70%.
That resulted in major under extrusion (I think it’s called this way). So I’m back to 200C again…
Is a print height of 0.2mm a good standard setting?
I’m having huge issues with the nozzle making “scratches” in the surface while printing new layers, the nozzle is “plowing” trough the top of last print layer each time.
I guess I’ve been a bit simple minded by thinking the finer layer the better result?
As the Z axis uses a 8x1.5 mm trapezoidal rod, you can’t use rounded heights. The correct height is 0.2025 or 0.1950 mm. Same for the 0.1mm setting which is an actual 0.105 mm. Only 0.3 mm is correct.
Also, I have measured that my first layers are not 0.1mm each but more like 0.06, 0.07, 0.07, 0.10, 0.10, etc.
I expect backlash in the Z-axis.
Those first layers of that Maker Faire Robot already show you how you are printing.
I have learned that tiny change can make quite some difference.
One more thing, I noticed yesterday that even each color behaves differently too ;).
My Black currently prints with a lower temperature and a slightly higher flow than my Cyan.
I just started to print with White and I see I need to tweak it’s settings too…
Yes h60,
The colors have effect on how the plastics flows but for the majority of the time no tweaks are needed, unless you really want a perfect print.
In regards to the temperature and flow i’ll check with my collegue when i’m back with the office, if he haven’t see this topic already.
This actually (or maybe finally) makes sense to me now (I think).
If I understand this correctly then exact numbers will be corresponding to the stepper motors degree and the height and diameter of the threading?
If this is the case could you please elaborate this a bit more, what numbers are you using to calculate and what about X and Y axis, what are their steps in movement in mm?
Sorry for not knowing the correct terminology and words in English, and if I’m totally out in the blue here please be gentle in your reply
On this site you can compute the various parameters.
If you go to “Steps per millimeter - leadscrew driven systems” the options of the Vertex are :
Motor step angle 1/8.
Driver Microstepping : 1/32
Leadscrew pitch : 1.5 (don’t use the presets)
Gear ratio : 1
It gives the result : 4266.67 (you can find that value in this line (#define DEFAULT_AXIS_STEPS_PER_UNIT {134.74,134.74,4266.66,200} ) of the configuration.h file or the Vertex firmware).
The value 0134.74 for the X/Y axis is computed in the “Belt driven systems” section :
That’s right but it hasn’t the last and most interesting part of the calculator site which is “Optimal layer height for your Z axis”. There you can compute the best (still rounded) layer height matching the one you need.
[quote=“ichbinsnur”]Did you ever do such calculations for X and Y too?
The X and Y axes are way more coarse than z.[/quote]
Excellent point. The Internet can be a wonderful source of information for those who discriminate. It is also a wonderful source for disinformation.
For the stock K8200, at 2560 steps per MM, a one MM error in the Z-Axis is an error of 0.04%.
At 64.25 steps per MM for the X and Y axes, a one MM error is an error of 1.5%.
The numbers speak for themselves, and beg the question as to why nobody ever worries about so-called “optimal” moves for the X and Y axes.
Another telling point: The z-Axis errors do not accumulate. If the first layer is off by 0.04% then every layer will be off by 0.04%, and the overall height will be off by 0.04%. THE ERRORS DO NOT ACUMULATE! “Optimization” of the layer thickness is an urban myth, and a tempest in a teapot.
