I've got the "opposite" problem, all my parts have one "slice layer" more in z axis. If i print a 50 mm with a 0.14mm my part will be 50.14mm (same for 0.19mm). Is i a bug in the last Z-suite software?
I've got the "opposite" problem, all my parts have one "slice layer" more in z axis. If i print a 50 mm with a 0.14mm my part will be 50.14mm (same for 0.19mm). Is i a bug in the last Z-suite software?
You should count the layers then you would find that you have 24 layers less!
Instead of 357 layers (wich is 50mm / 0.14 mm) you would count 333 layers (wich is 50mm / 0.15). This is because Z-Suite slices with less resolution than specified.
Z-Suite minimal resolution is 0.1 (instead of 0.09) and the other resolutions are 0.15, 0.2, 0.3 and 0.4 while the profiles are named 0.09, 0.14, 0.19 etc.
If you are talking about what you measure it's just a dimensional issue of the print and has nothing to do with layer count / slices.
Now let's see what happens next. When I mentioned this some weeks ago, my posting was deleted and the thread was closed ;)
I just wonder how much sense it makes to pretend this misleading resolution naming was a secret as the Z-Suite software let’s you easily count the number of actual layers.
I’d be also interested to know if it’s a coincidence that the fix of the Z issue happened at the same time as the hardware profiles for m200 were removed and the HW profile was hardwired to v02/trapez screw.
You should count the layers then you would find that you have 24 layers less!
Instead of 357 layers (wich is 50mm / 0.14 mm) you would count 333 layers (wich is 50mm / 0.15). This is because Z-Suite slices with less resolution than specified.
Z-Suite minimal resolution is 0.1 (instead of 0.09) and the other resolutions are 0.15, 0.2, 0.3 and 0.4 while the profiles are named 0.09, 0.14, 0.19 etc.
If you are talking about what you measure it's just a dimensional issue of the print and has nothing to do with layer count / slices.
Now let's see what happens next. When I mentioned this some weeks ago, my posting was deleted and the thread was closed ;)
I just wonder how much sense it makes to pretend this misleading resolution naming was a secret as the Z-Suite software let’s you easily count the number of actual layers.
I’d be also interested to know if it’s a coincidence that the fix of the Z issue happened at the same time as the hardware profiles for m200 were removed and the HW profile was hardwired to v02/trapez screw.
Probably explains why I was getting unexpected changes in height while tweaking geometry......
I did some checking to see if this is true or not; I think you're correct - everything seems to be shifted in the z-suite profiles by 0.01mm for the m200. Ive included my testing below for your own conclusion.
I think it's helpful when tweaking models for exact fits (you'll have unexpected changes in number of layers if not multiple of true slice thickness due to aliasing).
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Below is a microscope shot of a print set to 0.14 mm layer with a gauge above it. I got .15mm layer height instead of 0.14mm...
To double check, I counted every layer on the piece (72 layers) using the microscope and calipered it..... .15mm again
To check Z-Suite, I made this phase block or whatever you want to call it. It's a segmented block with each segment repeating. The special part about the block is each segment is slightly tilted, so it accounts for for vertical positional phasing and should highlight aliasing. Each segment repeats every 10mm, with .9mm solid and .1mm space. Therefore if you slice at a true 0.1mm, you should get 9 solid layers and 1 empty layer. Scaling the block by 90% lets you check for 0.09mm layer height. Scaling the block by 140% lets you check for .14mm layering, 150% checks .15mm etc... in all cases should always get 9 solid layers, 1 empty later. Because of the tilt for each segment, the transitions points between each layer will also be shown. If the slice height does not match the the appropriate scaling, the block will wigg out.
Set z-suites at 0.14mm, block set to check for .14mm. Layer transitions aren't correct, layer count not correct..
Change block to check for .15mm and set z-suite for .14mm layers. Layer transitions aligned, layer count is correct.
For z-suite set for 0.09mm layering......... the 0.09mm check block doesn't match, 0.10mm layer block matches.
Checked with z-suite versions V0.04, V1.0.0, and V1.5.0.2. Results same across all (M200 both hardware versions in V1.0.0, no difference)
Checked Inventure in v1.5.0.2 .... z-suite set for 0.14mm, does correctly match the 0.14mm check block.
Below is a microscope shot of a print set to 0.14 mm layer with a gauge above it. I got .15mm layer height instead of 0.14mm...
To double check, I counted every layer on the piece (72 layers) using the microscope and calipered it..... .15mm again
Much effort! :-)
Wouldn't it have been much easier to slice a simple object (e.g. cube / cylinder with 10mm height) and count the layers in Z-Suite with the layers view function (where you set a pause)?
That way you would see that with 0.09 layers you get exactly 100 layers (the layer percentage shown in Z-Suite is always the value of the layer count that way).
Since the M200 does not reslice what it gets from zcode, the printer always exactly prints what Z-Suite slices.
Means: The M200 never printed 90 microns resolution with the Z-Suite versions that have been published so far.
Why are you even bothering to count layers ?
Read post 5 back on page 1 from Revitorio, just tweak up the Z height adjustment by 1% or so until you get the printed height correct.
Much effort! :-)
Wouldn't it have been much easier to slice a simple object (e.g. cube / cylinder with 10mm height) and count the layers in Z-Suite with the layers view function (where you set a pause)?
That way you would see that with 0.09 layers you get exactly 100 layers (the layer percentage shown in Z-Suite is always the value of the layer count that way).
Since the M200 does not reslice what it gets from zcode, the printer always exactly prints what Z-Suite slices.
Means: The M200 never printed 90 microns resolution with the Z-Suite versions that have been published so far.
I should have used the microscope first, that only takes a few minutes, but I didn't have the camera battery charged :P
The slicing stepping thing you suggested was the first thing I did....but checking a model for differences in dimensions near the resolution of the slicer, you have to be careful because of the nyquist limit and because of potential tweaks like layering offsets or something...thought maybe something like that was going on....so I made the phase block (or whatever you want to call it), which also has potential hiccups depending on the slicer algorithms, but seemed to work fine in this case
Why are you even bothering to count layers ?
Read post 5 back on page 1 from Revitorio, just tweak up the Z height adjustment by 1% or so until you get the printed height correct.
Because when you have more than one surface your trying to adjust, you cant use a global Z offset. Also if you know the true layer thickness, you can make the change as a multiple of the layer thickness (if not exact multiple, then youll get unexpected changes due to aliasing when trying to adjust near the slicer resolution).
Actually I think nobody really cares if the resolution is 90µm or 100µm, but this makes it even weirder that Zortrax tries to keep this a secret.
Indeed if Zortrax is so interested in users not finding out about this, they should have fixed any Z height issues immediately so nobody would have felt the need to investigate..
Anyway, tweaking Z height to work around a bug shouldn't be necessary. But at least users shouldn't try to find workaround based on wrong layer height assumptions.
BTW: there is no layering offset apart from the very first contour/infill layer but Z-Suite shows it as one layer anyway.
You are working on the assumption that each layer, diameter of material, sits exactly on top of the previous layer,
like a vertical stack of roller bearings.
It don't work like that.
The new layer melts / sinks a little into the previous layer in order for it to stick.
So layer counting is actually irrelevant.
Actually I think nobody really cares if the resolution is 90µm or 100µm, but this makes it even weirder that Zortrax tries to keep this a secret.
Indeed if Zortrax is so interested in users not finding out about this, they should have fixed any Z height issues immediately so nobody would have felt the need to investigate..
Anyway, tweaking Z height to work around a bug shouldn't be necessary. But at least users shouldn't try to find workaround based on wrong layer height assumptions.
BTW: there is no layering offset apart from the very first contour/infill layer but Z-Suite shows it as one layer anyway.
Other than it's nice to know when trying to adjust for size of parts. I don't really care about it either because the quality of the final output from the machine is awesome, which is what I really care about, and the difference between 0.09mm and 0.1mm is likely imperceptible to the eye, they probably technical reasons for it (probably to prevent layering aliasing).
You are working on the assumption that each layer, diameter of material, sits exactly on top of the previous layer,
like a vertical stack of roller bearings.
It don't work like that.
The new layer melts / sinks a little into the previous layer in order for it to stick.
So layer counting is actually irrelevant.
The layer to layer spacing drives the volume of material ejected from the print head. Also I used a gauge.
So layer counting is actually irrelevant.
Actually you did not get the point (yet) and make the same mistake as the guy I answered originally: You mix up math and physics.
If you talk about layers it's talking about resolution (the math), not physics (shrinking, layer bonding etc.).
If an object of 10 mm is sliced the slicer creates layers according to object height / resolution.
If the resolution is 100 micron it creates 100 layers, if the resolution is 90 micron it creates 111 layers.
And because it's that simple you just can count the resulting layers in Z-suite.
That said: The M200 is never printing with a 90 micron resolution, max. achieved is 100 micron.
My original answer therefore was: Don't talk about layers if it is a dimensional issue (that's where your arguments pop in).
The layer count is always object height / resolution. If you want to adjust your printed objects dimensions you can use the Z-Suite offsets in advanced printing options and / or add or remove height.
Other than it's nice to know when trying to adjust for size of parts. I don't really care about it either because the quality of the final output from the machine is awesome, which is what I really care about, and the difference between 0.09mm and 0.1mm is likely imperceptible to the eye, they probably technical reasons for it (probably to prevent layering aliasing).
I agree about the difference being somewhat neglectable. I don't agree there could be any technical reasons to call a 100µm layer a 90µm layer. The printhead moves up 100µm, there is no reason whatsoever to call that 90µm and every other slicer on this planet calls a 100µm resolution just that: 100µm. And most obviously, this is not about aliasing. Even though the height of a single layer is off by only 10µm, the number of layers differs significantly even for objects that are just 5mm in height.
Side note: for the Inventure, the slicer actually uses 140µm layer heights while for the m200, it uses 150µm with the same "140µm" setting. So which reason would exist for that?
I agree about the difference being somewhat neglectable. I don't agree there could be any technical reasons to call a 100µm layer a 90µm layer. The printhead moves up 100µm, there is no reason whatsoever to call that 90µm and every other slicer on this planet calls a 100µm resolution just that: 100µm. And most obviously, this is not about aliasing. Even though the height of a single layer is off by only 10µm, the number of layers differs significantly even for objects that are just 5mm in height.
Side note: for the Inventure, the slicer actually uses 140µm layer heights while for the m200, it uses 150µm with the same "140µm" setting. So which reason would exist for that?
By technical reason, I mean the reasons why the m200 seems to running the layer height it's using - wasn't referring to the marketing/stated specifications side of things.
For the height issue you're talking about, I think I'm not observing that. Mine are usually off by small amounts for larger, for thin parts they're usually pretty close to (within a few of the layer setting). I sometimes have to adjust by very small amounts when trying to get slip/press fits..
For the Inventure, the layer resolutions appears to be correct. I'm suspecting the m200 is shifted by .01 due to mechanical aliasing, which would appear as banding on a 3d printed part (it's possible a bug or something else like resonance their trying to avoid or something, or maybe i'm totally wrong, dunno). The stepper motor has a specified step size, combined with the z-screw, gives a specific minimum stepsize for the z axis. If the commanded step isn't a multiple, then you can get weird effects.
I'm suspecting the m200 is shifted by .01 due to mechanical aliasing, which would appear as banding on a 3d printed part (it's possible a bug or something else, or maybe i'm totally wrong, dunno). The stepper motor has a specified step size, combined with the z-screw, gives a specific minimum stepsize for the z axis. If the commanded step isn't a multiple, then you can get weird effects.
I'm not sure if I understand that way of reasoning. The layer resolution is 100µm for the 90µm setting on the m200 and this 10µm difference exists for all resolutions.
There is no doubt about it as it can be and was proven by different approaches.
Of course there is probably some reason for not using the specified values - whatever that may be.
However, there was and is absolutely no reason to change the name instead of the specification.
The only problem I have with this whole matter is that Zortrax wasn’t honest regarding this from the start.
Maybe they feel it’s too late now to admit this little marketing trick. But I’d rather prefer to have the facts on the table now instead of continuing this charade forever.
I'm not sure if I understand that way of reasoning. The layer resolution is 100µm for the 90µm setting on the m200 and this 10µm difference exists for all resolutions.
There is no doubt about it as it can be and was proven by different approaches.
Of course there is probably some reason for not using the specified values - whatever that may be.
However, there was and is absolutely no reason to change the name instead of the specification.
The only problem I have with this whole matter is that Zortrax wasn’t honest regarding this from the start.
Maybe they feel it’s too late now to admit this little marketing trick. But I’d rather prefer to have the facts on the table now instead of continuing this charade forever.
Something like that can be hard to fix after the machines are out, but if you're designing a new machine it's a simple change to a different screw pitch or motor or whatever. And it might not be that, maybe something else like resonances they're trying to avoid or a bug, or trade off for 10% faster printing....or maybe I have it wrong... dunno
I'm not too worried about it from my perspective. I posted because I thought it might help people trying to size parts to tight tolerances, 0.01mm wont make a visible difference I think, but a full layer thickness difference can mess with fits.
A more simple answer is when sorted alphabetically the Zortrax is sorted to the bottom. When sorted by layer resolution and alphabetically the Zortrax sits near the top. Simple but understandable, I am not saying that is what happened but it is a possibility.