Precise printing thickness (print height)

Thoughts about Raise3D, 3D printing and making in general.
User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Precise printing thickness (print height)

Postby Vagulus » Sun May 23, 2021 2:05 am

How does ideaMaker deal with the issue of printing a specified model thickness (printing height)?

I use 0.8mm nozzles with a designated layer thickness of 0.64mm. Obviously (or so I thought) this means that the depth of material deposited on the print bed must be in increments of 0.64mm. Well ... that's what I thought ... then I devised this testpiece
210523 3D Printing Thickness Testpiece.png
210523 3D Printing Thickness Testpiece.png (4.53 KiB) Viewed 1614 times

with steps set (nominally) at 3, 5, 7, 11, 13 and 17mm. Printing it produced these results
210523 3D Printing Thickness Chart.png

Previously, I had always thought that the heights were pretty accurate although I must confess that I had not checked. These figures show that the heights deposited are only very approximate and I guess it does not matter when I am printing ornaments and suchlike. However, it does matter when I am trying to print mechanical prototype models. I really need much greater accuracy that I have achieved here. I have attached my STL file so that you can repeat my experiment and I'd love to know whether your results are comparable.

What I find even more interesting is that the heights are not in integer increments of layer height. How, for example, does ideaMaker tell my printer (a Pro2) to print 3mm of material by depositing 5.429 layers at 0.64mm per layer? I can't make any sense of this and I'd appreciate all help.
Attachments
210523 3D Printing Thickness.stl
(2.62 KiB) Downloaded 26 times
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sun May 23, 2021 11:40 am

Question, how did you determine the layers count value in your chart?
Did you read the gcode or use gcode preview to determine layers counts?

You aren't wrong for thinking this, this is how a slicer is supposed to work, there is no percentage of a layer.
I use 0.8mm nozzles with a designated layer thickness of 0.64mm. Obviously (or so I thought) this means that the depth of material deposited on the print bed must be in increments of 0.64mm. Well ... that's what I thought ...


Example, I sliced a 20mm test cube using your nozzle size and layer height and in gcode preview it tells me layers and final height.
Ideamaker is going full integer- as do most slicers.
Screen Shot 2021-05-23 at 7.19.34 AM.png

So, 10mm object STL height VS 16 layers times 0.64mm does equal 10.24mm
That's gcode VS actual real world print, but the point is, the behavior is correct, final object height as per what the slicer and resulting gcode is round up or height+.

The next question is about your profile since I see one area where the user can change things is first layer height.
In your 0.8mm nozzle profile, did you then also edit first layer height?
Screen Shot 2021-05-23 at 7.28.31 AM.png


Now if you want to get into real world gcode VS actual print, that has a number of variables.
Mechanical Z gap/ bed leveling at the start of the print- ultimately the exact nozzle to bed gap when starting the print.
Slicer setting for first layer height beyond the mechanical Z gap
Flow rate settings affecting total volume of plastic per layer and is cumulative error. In other words, lesser filled layers results in final height being less than expected per gcode. Over extrusion results in nozzle plowing through layers as layers build up more and more cumulative error.
Plastic type and percentage of shrinkage from hot liquid state to cool solid room state.
Filament diameter variations in the spool of filament resulting in variable volume into the print.
Any extruder slippage from gripping the filament error.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sun May 23, 2021 11:53 am

Again, just looking at your chart:
I think you may not be accounting for mechanical Z bed gap and/or first layer height values.

It's working just as expected. Your real world results are just about what I would expect.
Yes, suddenly precision shocker- 3D printing is not as accurate as claimed in advertising.
Z by far is the most misunderstood aspect of all of this. Between the integer layer of layer height and all the mechanical variables of first layer nozzle to bed gap being user adjustable and extremely variable, and finally extrusion volume errors typical, well yes, it can be off by a country mile.

A user can tune and get much closer and more accurate printing, but that requires understanding the variables and then adjusting them. Both mechanical and in the slicer. Also, massive experimentation, print, adjust, reprint. Also, changing variables like volume and first layer height incorrectly to attempt to tune final Z height can result in print problems like elephant foot, poor bed adhesion, affect XY dimensional values and so on. It's all a tradeoff and typically, the tradeoff that is most acceptable is slight variation in final Z height.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sun May 23, 2021 12:11 pm

Also, there are plenty of discussions about how variable the Z axis homing can be for a given printer.
That's basically Z gap error variation before you ever print anything out of the nozzle.

On the N series with a mechanical switch and a threaded thumbscrew with spring bumping it, there was significant variability to the point it caused people some issues. A workaround on that printer was to put a nut on the threaded bump stop to prevent wiggle. That is a highly effective mod, just requires you loosen the locking nut to adjust but again, highly effective simple mod. https://forum.raise3d.com/viewtopic.php?t=1466#p12701

Pro2 series uses an optical sensor that since it requires no pressure or even contact, in theory is higher precision and more repeatable accuracy. That said, I'm not aware of any precision enhancing modification discussions specific to this printer, but may have just missed them.

E2- I'm not sure, don't own that printer.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sun May 23, 2021 1:02 pm

We need to make more columns in your chart and again, your final value column of layers is wrong and the wrong way of doing it.
Nominal STL height, Actual print measured value, Gcode layers @ layer height, gcode expected height, error
So we take a nominal height of 17mm, divided by 0.64mm layer height = 26.5625 layers, but remember, that should be an integer so we round UP= 27 layers.
To prove the round up behavior and gcode layer height, I scaled the 20mm test cube object to 17mm high, sliced with my custom .8mm nozzle 0.64mm layer height profile and got 27 layers and 17.280mm gcode print file.
Screen Shot 2021-05-23 at 8.20.48 AM.png

Your chart shows your real result at 17.424mm and let's assume (I know, I hate that too) gcode was like mine 17.280mm a difference of 0.144

Let's do a line that appears to have the most error (highest non integer portion in the layers column), 3mm nominal, actual 3.475mm.
3mm/0.64=4.6875 layers, so we again round up to 5 layers @ 0.64mm=3.2mm gcode expected value. 3.475-3.2=0.275

Ok, let's shoot middle now as I'm trying to best understand cumulative error here. In other words, you made an object, a one piece object with expected final layer heights. It printed at the same time layer by layer with simply less layers for lower heights and more layers on the upper heights. The idea is, we might be able to identify a baseline error that represents Z gap error intruded mechanically at the start of the print, but then also, if we define the error, then be able to see a cumulative error of per layer that either shows underfill or overfill of the incremented 0.64mm layer nozzle height.
We already see a trend:
First height is 3mm, we are over but what matters is actual VS expected, and by expected, I mean what the gcode produced by the slicer expected- not the original object 3mm.
3mm, error is 0.275, in 5 layers.
17mm, error 0.144, in 27 layers

So, 11mm object height, 11.097 actual, 11/0.64=17.1875layers, so in this case, round down to 17 layers. 17*0.64mm=10.88mm gcode height minus result of 11.097mm actual=0.217mm error.

We could go more, but the pattern is there. Remember, this is expected gcode height per the slicer due to integer layer rule VS actual print result.
5 layers 0.275mm error
12 layers change 0.058mm
17 layers 0.217mm error
10 layers change 0.073mm
27 layers 0.144mm error
The pattern is a decreasing error PER number of layers. This implies that 2 things expected happened.
#1 there was an initial hardware gap on first layer that added a baseline error to the entire print.
#2 layers are slightly under extruded or compressed resulting in cumulative less than expected layer height the taller we get.

Again, there are 2 trends just per the above numbers. The raw error is continuously decreasing between the first height of 5 layers, and the final height of 27 layers, 0.275mm down to 0.144mm. There is a second increasing cumulative error that in the first 12 layer change from 5-17 layers, we changed a total of 0.058mm cumulative error or 0.00483 per layer. In the next jump, 10 layers from 17-27, we saw a change of 0.073mm or 0.00730 per layer.

Also, I have to be honest here- charting 3 digits???? No, I'm sorry, I'm calling it, you did not, nor would I, nor would most anyone at home get 3 digit accuracy in measuring mm. At best 2, realistically 1 given the surface and nature of 3D printing.

My gut says you have a roughly 0.1mm or more baseline Z gap error before ever printing.
We have a slight cumulative under extrusion that given a tall enough part you end up being short, and small parts, you see this error much greater.

Again, you have 3 basic errors or variables:
First layer nozzle to bed mechanical gap- something you can adjust and correct.

Integer round up or down nature of layer height due to how a slicer creates gcode- something you have limited control to change other than choosing layer heights that results in less rounding jumps.

Minor over or under extrusion per layer cumulative layer height error that adds up with more layers-another variable you can attempt to tune but caution, this also affects XY dimensional wall thickness results- basically flow rate affects this most.
Last edited by Jetguy on Sun May 23, 2021 1:32 pm, edited 2 times in total.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sun May 23, 2021 1:04 pm

Also- hats off to you for a good technical discussion starter. You clearly are getting into some of this.
That's good thinking and beginning to peal back the layers of the onion of how 3D printing actually works.

User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Re: Precise printing thickness (print height)

Postby Vagulus » Tue May 25, 2021 7:54 am

Thanks again Jetguy for all that work. Once again, there is a lot to digest. I'm going overseas for a few days but I'll get on to this when I get back.
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Tue May 25, 2021 12:40 pm

Again, you have 3 basic errors or variables in final Z height of a print:
First layer nozzle to bed mechanical gap- something you can adjust and correct. The problem is, you are taught and by the manual this is a number- say 0.1 or 0.2mm or some other value. It's a combination of how flat your bed is, how straight the motion gantry doesn't sag in the middle, and your ability to level and adjust both the bed and the Z gap endstop adjustment. However, the focus taught was bed adhesion and yes critical, but NOT dimensional accuracy starting point. In other words, you might be creating a thick or thin first layer for best adhesion, and then NOT accounting for that new added error in Z height. This is the foundation of the print so again, the tradeoff was and has been better usability to even get a print to stick VS possible small Z height final object error.

Integer round up or down nature of layer height (whole number layers) due to how a slicer creates gcode- something you have limited control to change other than choosing layer heights that results in less rounding jumps. Again, yes, you were taught correctly, it does print in full integer layers. You apparently assumed because the actual print final value was NOT an integer that it was printing not in integer and that was your false assumption.

Minor over or under extrusion per layer cumulative layer height error that adds up with more layers
-another variable you can attempt to tune but caution, this also affects XY dimensional wall thickness results- basically flow rate affects this most.

User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Re: Precise printing thickness (print height)

Postby Vagulus » Wed May 26, 2021 3:17 am

A few spare minutes and I had a look at the first error type.
My print nozzle clearance is reset from ).2mm (recommended) to 0.28mm to fix an over-adhesion problem. I adjusted my First Layer Height from 0.64mm to 0.72mm to allow for this then reprinted the test piece. This gave these results
210526 Printing Thickness Chart.png

Layers Ind.[/b] refers to the layer count taken from ideaMaker Preview.
[b]Layers Calc.
refers to the printed (actual) thickness divided by layer thickness of 0.64mm.
Comparing this to the original values (before the change to First Layer Height ) shows little change in accuracy
210523 3D Printing Thickness Chart.png

I expected more change in accuracy.

Notes:
1. You can see the rounding effect between ideaMaker's number of layers and the calculated number.
2. I agree about the number of decimal places in my measurements. I am aware that three decimal places is a bit of overkill, but I am also aware that computers work in a lot more decimal places and having three should reduce calculation errors a bit. That's my theory anyway ;) !
--------------------------------------------
God so loved the World that
He did not send a Committee

User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Re: Precise printing thickness (print height)

Postby Vagulus » Fri Jun 04, 2021 2:59 am

Let's look at this another way.

My printed (corrected model) gives these results:
210604 Layer Errors.png

Take the top line: the actual height of the object is 3.505mm composed of 5 layers. Layer thickness is set at 0.64mm. 5x0.64=3.2.

What would cause the discrepancy of 0.305mm?
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sat Jun 05, 2021 3:22 am

Vagulus wrote:What would cause the discrepancy of 0.305mm?



0mm height is the bed touching the nozzle.
We cannot print like that and risk damage of an XY nozzle movement (example homing XY with the nozzle already touching the bed) while homed at TRUE Z0 so we "cheat" and make the Z homed position a gap above the bed. So homed at mechanical now becomes Z0.2 or whatever you mechanically set.
Now, does the SOFTWARE making the gcode or the homed position in firmware take amount for this mechanical and logical offset?
Answer is NO!!!
No, the home command simply marries up the I AM HERE coordinate in firmware memory to 0mm when you home.
So the printer thinks it's at Z=0 when homed, even though the nozzle is clearly some gap above Z=0.
So you are off even before EVER printing a line of gcode.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Sat Jun 05, 2021 10:57 am

Again, it's a simple mismatch of assumed coordinates.
The printer thinks it's at Z=0 when homed, even though there is a mechanical Z gap intentionally set and needed by the printer for safe operation. WORSE, that gap is a variable set by the user's skill and variation from one printer to the next and one user to the next.

The slicing software produces gcode without nary clue one what the true gap is. If it doesn't know what that gap is, it cannot adjust the number of layers or rounding such to end up at a closer actual height. Don't think that because first layer gap settings are affecting first layer they are fixing this discrepancy- they aren't in the slightest and reading and understanding the gcode can prove this.

So here's the bigger problem. I just pointed out the technical flaw and exact point where reality (the bed and nozzle gap) is one thing, and the logical coordinate system causes a discrepancy. This is actually true on most printers- even SLA resin printers. For millions of reasons, we set a small gap rather than the printing surface directly in contact with XY. Most firmware is written to home and then automatically zero the axis. Again, with some degree of certainty, most every brand of printer I've touched in 10 years, this was the dirty secret error introduced just by some early on precedents in 3D printing. People mostly just accepted it.

So fine, there is a discrepancy but the slicer knows it right? Well, probably not. Think about it, your prints are printing and sticking to the bed now. If you go changing the coordinate system in the firmware, then this simple bit of logic applies. Z cannot be negative values. If you send a negative value Z gcode to motion control firmware it will ignore the command. Same concept, if the firmware says I'm at 0.2mm and I tell it to move to 0.2mm it doesn't move because the current and target values are identical. And then we get into gcode omission. You notice if reading print gcode, thousands if not millions of lines have no Z value. That because we print in layers and if the line doesn't require a z move, then we send less characters and less information for the firmware and motion planner to deal with. But as an example, first layer doesn't have Z moves it starts at logical Z0. Second layer then is the first Z move. However, if we again are at a non-zero value, and the second layer move is even say in your case, Z0.65 and you have fixed the Z0.3 error by using the G92 command to say G92 Z0.3 at the end of the starting gcode https://marlinfw.org/docs/gcode/G092.html well again, second layer will only move 0.35 not 0.65 thus elephant footing -jamming a full layer liquid volume into a half layer height means it has to squish somewhere, and that somewhere is out sideways.

There may be a combination that works. Giant hint, there may be other slicers that handle this differently and possibly better accuracy.

User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Re: Precise printing thickness (print height)

Postby Vagulus » Sun Jun 13, 2021 11:46 pm

Jetguy wrote:Now, does the SOFTWARE making the gcode or the homed position in firmware take amount for this mechanical and logical offset?
Answer is NO!!!
No, the home command simply marries up the I AM HERE coordinate in firmware memory to 0mm when you home.


Orrite!

I said before that the Homed clearance of the nozzles over the print bed is set at 0.28mm rather than the default 0.20mm to cure an over-adhesion problem. I also said that I had adjusted First Layer Height on ideaMaker from 0.64mm to 0.72mm. This, I thought, would compensate for the increased gap leaving the top of the first layer where it should be. Am I incorrect in this?

Perhaps the real question here is, "What changes when First Layer Height is altered?
Is there any change? Does the change alter the depth of the first layer of deposited filament as I expected?
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Mon Jun 14, 2021 12:20 am

I want to help but part of this is you learning yourself.
That means reading gcode.
That's not too hard, we are looking for a single line and really a value a Z axis value AFTER the starting gcode.
I'll do one, and then you do your experimentation.
;Sliced by ideaMaker 4.1.1.5050, 2021-06-13 20:11:41 UTC-0400
;Dimension: 305.000 305.000 300.000 0.400 0.400
;Plate Shape: 0
;Origin Center: 0
;Extruder Offset #1: 0.000 0.000
;Filament Diameter #1: 1.750
;Filament Diameter #2: 1.750
;Filament Compensation #1: 90.00
;Filament Compensation #2: 90.00
;Filament Density #1: 1240.00
;Filament Density #2: 1240.00
;Bounding Box: 122.000 182.700 136.598 168.398 0.000 48.050
;Model Gap: 0.000
M221 T0 S90.00
M140 S60.00
M104 T0 S205.00
M109 T0 S205.00
T0
M190 S60.00
G21
G90
M82
M107
G28 X0 Y0
G28 Z0
G1 Z15.0 F6000.0
G92 E0
G1 F140 E29
G1 X20 Y0 F140 E30
G92 E0
G1 F6000.0
M117 Printing...
M1001
;TOTAL_NUM: 1
;PRINTING_TIME: 0
;REMAINING_TIME: 5969
;LAYER:0
;Z:0.300
;HEIGHT:0.300
M106 S0
;PRINTING: 3DBenchy.stl
;PRINTING_ID: 0
G1 F1200 E-1.0000
G0 F6000 X127.551 Y146.709
G0 F1500 Z0.300

So there is the magic line, the first printing move layer height is 0.3mm ABOVE the homed position already gaped some distance above the bed.
For reference, this was the profile with a 0.3mm first layer setting. So, it matches the layer.
Pro2 firstlayer03.jpg


So then just for giggles, I made it 0.1mm first layer in the profile and exported a slice and the gcode follows and first layer is 0.1

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Mon Jun 14, 2021 1:01 am

Vagulus wrote:I said before that the Homed clearance of the nozzles over the print bed is set at 0.28mm rather than the default 0.20mm to cure an over-adhesion problem. Wrong, you should adjust first layer flowrate to adjust an over extrusion problem if you know the gap is right. I also said that I had adjusted First Layer Height on ideaMaker from 0.64mm to 0.72mm. This, I thought, would compensate for the increased gap leaving the top of the first layer where it should be. Am I incorrect in this? YES

Perhaps the real question here is, "What changes when First Layer Height is altered? Answer: The actual first line of Z axis move in the gcode matches whatever this setting is. First layer height=Z homed gap physical setting+first layer setting
Is there any change? Yes, proven in the above post. Does the change alter the depth of the first layer of deposited filament as I expected?
No, not as you expected. Here lies the issue.

3D printing 101
3D printing is 3D, a height, a width, and a length, and actually a 4th value filament length to get volume.
The nozzle squirts a noodle of hot plastic the diameter of the nozzle (ignoring die swell at the moment).
Ideally you print with a layer height ratio somewhere near 1/2 nozzle diameter. The logic and reasoning behind this is to squish the round hot noodle into an oval for more surface area contact with either the build plate for first layer, or for other higher up layers, bond to the previous layer. In addition, this squish bonds adjacent parallel infill strands together on the sides because the paths overlap (extrusion width).

But here is the part to get:
The slicer creates straight lines of gcode that represent a 3D actual line of extrusion. (Logical= Physical). So when the slicer figures out the pattern for a layer and makes a gcode segment, that segment has a height (layer height), a width (nozzle width or *user defined value) and a length. That forms a logical volume of 3D space (height*width*length), so the slicer does the math with what filament diameter you gave the slicer (1.75mm nominal but should be user entered to reflect actual diameter of the wire filament) and from that if figures out the exact same volume of filament to equal the volume of the space. That is the E distance value also in the same line of gcode.

Last, flowrate edits either the direct linear distance of filament OR tells the firmware to do it.
Point being, flowrate 100%= execute the gcode as written for E value.
Greater values flow MORE distance of filament than the line of gcode implies thus over extruding more plastic causing the plastic to squirt wider than normal (elephant footing) or make up for a gap uncounted for in layer height!!!


Again, let's go over where you went wrong.
You said
the Homed clearance of the nozzles over the print bed is set at 0.28mm rather than the default 0.20mm to cure an over-adhesion problem.

So you wanted dead accurate final print results and adjusted a mechanical setting not accounted for anywhere in the software to fix an adhesion problem and break worse a final height error.
The correct answer would have been to adjust first layer flowrate to LESS than 100% in the slicer and adjust until you got the adhesion you wanted without elephant footing or under extrusion to the point it didn't stick or infill bond properly.
Again, you are already on day one adding a gap not accounted for anywhere, the idea is to reduce this gap to the bare minimum if accuracy is your goal not ease of printing or adjustment. Doing it the right way VS the easy way.

Then, we got into a whole new level of hurt in this additional error
I had adjusted First Layer Height on ideaMaker from 0.64mm to 0.72mm.
First off, layer height should be a direct function of nozzle size with an ideal ratio of 1/2, however, can can vary and go up to 3/4 with a tradeoff of less bond and layer strength or lesser, to 1/4 ratio, with more trouble because you are now smearing the strand wide so width and plowing of the tip of the nozzle becomes a possible tradeoff in surface quality. A 0.8 nozzle is ideally 0.4mm layer height. 0.2 and 0.6 being the min and max recommended respectively, so 0.64 is already above. Then you added another bit and raised it to 0.72. Well, if we had the nozzle darn near touching the bed, raising the nozzle from the homed position for that layer would not even begin to squish the round hot noodle of plastic oval.

What might be happening is:
You've not measured your filament and likely it's larger than 1.75mm more like 1.82 to 1.83ish.
You might be running a different flowrate than 100% Ideamaker filament profiles default for 94% flowrate.
When you increased layer height, the slicer increases the volume of each segment of gcode and thus makes longer E values and the printer follows them. So again, increasing layer height not ONLY increases the nozzle height, but also increases the volume of plastic squirted during that line of gcode.

What you wanted to do:
Go back to a close as possible homing gap without scraping the bed anywhere. The closer you get, the higher the risk of scraping so only you know how flat your bed is, how much you can adjust this gap.
Once that's set, now make first layer match your ideal layer ratio. If 0.6mm is where you want to be, then if you get a 0.2 homed gap, first layer would be 0.6mm-0.2mm homing gap= 0.4mm first layer height setting in the slicer.
Then you print experimentally to get the correct first layer flow right so that it prints at the correct nozzle height and you end up at the correct volume of plastic to not elephant foot, but also not under extrude.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Mon Jun 14, 2021 1:29 am

Vagulus wrote:What would cause the discrepancy of 0.305mm?


Answer: a rather tall adjusted first layer gap of 0.28mm to start with.
the Homed clearance of the nozzles over the print bed is set at 0.28mm rather than the default 0.20mm to cure an over-adhesion problem

And then, adding an extra tall layer well outside recommended nozzle to layer height ratio of 0.72mm
I had adjusted First Layer Height on ideaMaker from 0.64mm to 0.72mm.


So yes, not surprised in the slightest- 0.3mm knowing we started with 0.28 error.
The 0.02 error is expected roughness or error in measurement.

Jetguy
Posts: 3227
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Precise printing thickness (print height)

Postby Jetguy » Mon Jun 14, 2021 5:10 pm

The lessons to be learned:
#1 this slicer (Ideamaker) and many others do not account for the mechanical homing based nozzle to bed gap.
#2 First layer settings aren't complicated, it simply puts that value for the first layer move from HOMED position. As seen in #1 homed position probably isn't true nozzle touching the bed true 0mm. Thus, all Z moves are gcode value+ homed gap+slight mechanical variation or plastic flow.
#3 understanding layer height, nozzle or extrusion width, filament diameter, and how flowrate works. Also, being able to read gcode and find a value.
#4 just basic general expectations- up in arms about a 0.3mm Z error with a 0.8mm nozzle?? Already showing how most people unknowingly probably already have a 0.2mm or more error with a 0.4mm as stock and are clueless about it. Bottom line expectation any error in XYZ less than the nozzle size, let alone half of it, you are doing stellar. You can chase your tail all day and yes, one can dial in tighter- but the print waste and time to get there- not worth a fraction of a mm.

User avatar
Vagulus
Posts: 225
Joined: Mon Mar 25, 2019 5:27 am

Re: Precise printing thickness (print height)

Postby Vagulus » Sat Jun 26, 2021 1:52 am

Jetguy wrote:Again, you have 3 basic errors or variables:
First layer nozzle to bed mechanical gap- something you can adjust and correct.

I mentioned that I had adjusted my gap clearance from 0.2mm to 0.28 mm to fix a model adhesion problem. You said to adjust the first layer flowrate rather than the clearance. I finally got some spare time to work on that.
210626 Three Tests 480x360.png

(A large copy iof this photo is at https://1drv.ms/u/s!ArMxdUYEh9fKv37RTcXO-ApA3gwY?e=dhfDQR.)
All printed with 0.8mm nozzle. All 40mm square.
1. first layer height 0.72mm; first layer flowrate 120%; gap 0.28mm.
2. 0.64 / 120 / 0.28.
3. 0.64 / 120 / 0.2.
That takes the machine back to my standard settings for 0.8mm nozzles. Problem is that I can't see that the adjustments made any difference.

The original adjustment of first layer height to 0.72mm was because a change of BuildTak resulted in the models sticking far too much to the bed - I made a large hole in a new sheet of BuildTak getting the first models off it (which annoyed somewhat). Going back to 0.64mm first layer height has had no effect. Perhaps a bit of use has settled the BuildTak down. :o Interesting, but no matter - everything seems to be working just fine ... however: ...

you will see that the diagonal threads (this is the bottom surface in each case) do not blend to the border threads. (They do just fine on the top surface.) Which of the layer settings should adjust this? If I adjust something to fix that blending problem on the bottom layer my top layer gets fouled up.

Thanks
Attachments
210626 Layer Settings.png
--------------------------------------------
God so loved the World that
He did not send a Committee


Return to “General”

Who is online

Users browsing this forum: No registered users and 2 guests