andrewellis93/print-tuning-guide

print-tuning-guide's Issues

Pressure Advance Calibration tool

The tool is very good, there is one small issue. At the beginning, before setting pressure advance, You set Z to 0.25 (or first layer height). At this point a Voron 2.4 is above the Z homing switch, it's better forst to travel to the starting point and then lower Z to print level. See third line in the code below.

M106 S0 ; Set fan speed
SET_VELOCITY_LIMIT ACCEL=4500 ; Set printing acceleration
G0 Z0.25 F9000 ; Move to z height
SET_PRESSURE_ADVANCE ADVANCE=0 ; Set pressure advance

Input shaping

Cura does have per object Extrusion Multiplier (Flow) settings.

Hi, I am following your guide and got confused by:

Cura and Prusa Slicer do not have per object Extrusion Multiplier settings so each test cube will need to be printed separately.

You can configure a per model setting, it seems to me that this is what is needed for the EM calibration.

Single Page Guide

Is there an easy/quick way to combine all the guide sections into a single, continuous page? That makes it very easy to search keywords (via browser) as well as just scroll down for the next step, rather than clicking links continuously or jumping browser tabs.

Bulging on seams

Hi,
what can cause this bulging on seams?
I use Ellis standart 45deg 15mm/s profile

thanks for hints

Issues With SuperSlicer Profiles

So, I just went throught this entire guide and everything went nicely, until I used the superslicer profiles. Whenever I slice with them, I get an error that Python 3.1 cannot be found in the C:/ directory, even though I have python installed. Though this error occurs, a gcode file still exports, albeit when trying to print it, the bed doesn't heat during the print, and the print usually stops during the start of the print. Any help in fixing this is greatly appreciated, thanks!

TEST_SPEED macro accuracy issue with hall-effect endstop

If the carriage is too close to the hall effect endstop for the first homing, the microsteps can potentially be miscounted during the G28 sequence.

Might want to mention this potential issue in the work instructions or modify the macro to account for hall-effect endstop users.

Can you please add some guide for the labeled extrusion multiplier cubes?

Hello,

I see in the test prints folder that you have some labeled cubes and I like the idea to have them visible even after you mix them ( cause if you print multiple normal cubes you can start to mix them) but I would like if possible to have a section in the guide on how to use them. For now I don't have an idea what extrusion multipliers to set on each one to be able to test it properly.
Thank you @AndrewEllis93

PrusaCode will not working

latest version of PS

Twisted/skewed frame

Based on
https://discord.com/channels/431557959978450984/582347488938164240/993958021283725433
showing an unusual bed mesh

and on the answers
https://discord.com/channels/431557959978450984/582347488938164240/994063694436782201
"Mesh looks like a twisted frame to me. Could the feet arrangement have changes. Try adjusting feet or shimming?"

https://discord.com/channels/431557959978450984/582347488938164240/994121022536175676
"Actually I remember the mesh was flatter when the printer was on a desk, then I moved it upstairs and placed on the floor"

which suggests a skewed/twisted/uneven frame, would it be possible to extend the page about gantry squaredness to a more general squaredness of the outer frame (even vertically), with tips to check and correct?
Of course it's less relevant for Vorons, since they have 4 point bed screws and a gantry independent from the outer frame, but your guide is very good and used also by people with different printers!

In the case of that post, the author with the issue has a glass bed so that can be used as reference, but given the photo posted, the gantry is using 8 mm rods so it will sag, making even a perfect frame show some bowing.

It's tricky, but maybe with your experience you can extend the guide by providing at least tips.
If you provide an example mesh with the issue, some people may even realise they are affected by it and take actions!

Commit description

Hi, thanks for your work first of all.
I followed the guide and I got your PIF profile and from time to time I come back to check what changes/improvements are there.

I find however very time consuming to open each single commit to find out, all of them are called "update" with, usually, nothing else.

I think that giving more meaningful titles and a line or two of description would help the (several) users of your guide and profiles to follow better!
I understand it would take you a bit more time for each commit so you may not want to follow the suggestion, but I thought it was worth to mention it to you so you are aware of it.

Inconsistent Order of the first Print Tuning steps.

This is just a minor thing that I noticed while following your guide.

On the main Readme it says this: "Essentials. Read/follow these in order shown." and then it shows the following things:

• Build Surface Adhesion
• First Layer Squish
• Pressure Advance ("Linear Advance" in Marlin)

And in the Build Surface Adhesion guide it says right at the beginning to first ensure that the "First Layer Squish" is correct.

So wouldn't it make more sense to list the First Layer Squish Guide before the Build Surface Adhesion Guide in the main Readme?

FR: "Back to Table of Contents" on page bottom

It's a bit more natural to click the "Back to Table of Contents" link when you've finished reading the article, at the bottom. That way you wouldn't need to scroll all the way up to go back. It's not a big thing, but it would be a nice quality of life improvement

hyperlinked Amazon product page for BTT Smart Filament Sensor is incorrect

The hyperlink included in line pause_resume_filament.md#L64 forwards the user to the incorrect Amazon product page.

TEST_SPEED macro on Delta printers

Hi
I found this via the Vector 3D channel on YT and really love your work..
So much so that all 5 of my klipper printers have been worked on today.
But....
The TEST_SPEED macro do no work on the Delta printer, I have been in the macro and tried to set it up, but its beyond my understanding unfortunatly.
What should be considered when testing on a Delta printer???

Bulging caused by Arachne

Hey Ellis,
I recently tried to troubleshoot a print of mine since it always had two layers that bulged.
I tried some of the methods you suggest but nothing seemed to help.
A closer inspection of the part in the slicer finally lead me to the solution.
The bulging was caused by the Arachne slicer settings in SS. Therefore I think it might be worth adding a note about inspecting the outer perimeters in the slicer using SS/PS Width view since it highlights the issue quite clearly.

With Arachne, layer width view:

Classic width, layer width view:

Retraction Test 45 degree bug

The picture says more than I can write. I have setup the profile according to the guide and this was one of the last steps.
Info: SS Version 2.4.58.2

Test pattern prints in the wrong place

I have a voron 0
It has Bed Size X: 120 mm
It has Bed Size Y: 120 mm
But your test pattern prints outside these dimensions

Ever tried using "max line overlap"?

Appreciate your contribution here - you've created the best tuning write up I've come across. I wanted to ask if you've ever touched the "max line overlap" feature or not. Here's my experience with it:

One parameter I've used with good effect is the "max line overlap" filament setting in superslicer. Merill occasionally mentions it in discord threads, but I don't see too many people talk about it that much. SS/PS/Slic3r obviously wants to space lines to fill all gaps between lines, but a filament like PETG would never really flow smooth enough to fill out like this. As a result solid layers would often overextrude and bulge, and the traditional way to offset this was to reduce the flow multiplier. Reducing the flow multiplier would decrease the dimension of an object, and altering shrinkage to correct that would just move the dimensional inaccuracy to an inside dimension. By tuning with the max line overlap I've been able to get near dimensionally accurate 10.00 x 10.00 x 10.00 cubes with equally accurate interior dimensions with the same setting. It's like a game of adjusting max line overlap to reduce any multi-layer solid infill bulging. Then adjust shrinkage to get the interior vs exterior dimensions balance out (like 9.98 on an inside dimension and 10.02 on an outside dimension). Then adjusting extrusion multiplier to get to 10.00 on both. If extrusion multiplier changes, then max line overlap can be adjusted to compensate. One place where I always had issues before using this approach was thin strips of solid infill. Even adjusting the infill/permieters encroachment to 0, the solid infill would still always bulge.

The newer Superslicer "Solid fill overlap" feature probably provides similar capabilities of the "max line overlap" feature, but probably better to use the max line overlap setting on the filament as it's usually the filament's flow parameters that need the adjustment. Since using this method I usually have my extrusion multiplier closer to 0.98 - 0.99 and the max line overlap can get as low as 20% depending on the flow of the filament. 20% might sound really low, but it really looks identical to a .93 extrusion multiplier and 100% max line overlap.

PETG overhangs tips

Hi ellis! Thank you for your time and your fantastic guide!
I not have a Voron, I have a modded davinci 1.0 pro and I use pets from first day. It is know for bad bridging (not so bad if tuned well) and overhangs. When there is a small piece with overhang it curl up and this ruin my prints, (bulges on wall). Petg wants cool down naturally between layers. I tried some values for temperature (higher temperature to prevent stress after extrusion and lower to prevent curling), fan speed ecc... but nothing works well. if I can I print multiple object so it cool down itself, but isn't always possible...
have you any tips about? thanks

Broken PA & EM link?

First, thank you for this guide!

I believe the method section has a broken link titled "throw you off further." to what is now called The Relationship Between Pressure Advance & EM.

Pressure advance and part mechanical strength

Hi there, not sure how else to comment/contribute, so apologies for opening an issue :-) .

I am in a process of doing extensive testing and research using Voron 2.4, what i've found with Pressure Advance is that it directly affects mechanical strength of the part (Logical!) and your amazing article suggest to always stay on the lower side of the PA value... this can and will mask the issue, the print will look nicer as there will be no gaps close to corners BUT, this will cause serious "perimeters fusing together" issues making the parts much weaker.

What i did as a test, i keep printing "[a]_xy_joint_cable_bridge_igus.stl" from Voron 2.4 repository as this part has everything, 90 degree walls, 45 degree walls, holes, bridges... and what i have found is if the EM is correct and PA too low, i can easily break the part with my fingers, when the PA is higher (in my case cca 0.06) for Fillamentum ASA i can not break the part no matter how hard i try. same setup, PA different.

So yes, lower values do look nicer and more correct, but the part is mechanically weaker.

Thanks,
Alek

Kudos

This isn't an actual issue so you can close it after reading.

I just wanted to say well done. Your guides in general have always been excellent, but I finally got around to going through your gantry squaring guide in particular, and it is exceptional. I was just thinking to myself the other day how rare it is in the opensource community to find great instructions accompanying projects (through no fault of their own obviously; individuals contributing to such projects are already going above and beyond if they provide any sort of guide at all, not to mention writing good instructions often takes a very different skillset/mindset than what is required for actually working on the project itself). Your own guide covered absolutely everything that was needed for the process in both text and image-based formats, and did so clearly and succinctly, without including anything that seemed superfluous. I think the final step really exemplifies what I mean where you even had the consideration and forethought to include a reminder to change back the idle timeout. It's a small thing, but it's the kind of detail that most guides gloss over.

Thank you for the fantastic work you do here :)

Gantry Squaring Suggestion

During the process of rebuilding my V2, I decided to re-square my gantry. I found the Z locks quite useful:
https://github.com/VoronDesign/VoronUsers/tree/master/printer_mods/tallman5/z-locks/

They hold the gantry at a set and even height relative to the frame. They also prevent the gantry from falling if the Z belt is loosened a bit too much.

I also did a squaring with the GE5C mod installed, and this, with the Z locks installed pretty well locks the gantry in place. Thus, just loosen the attach points, and retighten and it is perfect with in an N95/95 tolerance. I wrote about it here:
https://github.com/IRTrail/Voron-Stuff/blob/main/Voron_2_Gantry_Squaring.md

At any rate, the Z locks help, a lot. Just a suggestion.

Input shaping: before PA

Wonderful guide.
Would you please edit it to make it clear that input shaping should be performed before PA, since even small amounts of ringing will affect the appearance of corners and lead to wrong estimates?
Thanks

The best way to calibrate pressure advance!

I've just stumbled across this from a 3 month old reddit post. I've always had issues getting my pressure advance right until I found your tool. 5 minutes and my problems are gone. Really easy to read the results. The website is easy to use. What worries me is how long will it be hosted. Would be epic if this able to be downloaded and ran offline.

Determining Maximum Volumetric Flow Rate

Stefan, CNC Kitchen on YouTube, has an excellent video that uses a tool he built for working out flow rates. One of his viewers translated that tool to a web page.

I needed to adjust the [extruder] max_extrude_cross_section klipper setting for the tests.

Print Tuning Guide

https://github.com/AndrewEllis93/Print-Tuning-Guide/blob/main/articles/extrusion_multiplier.md

Final G28 after BED_MESH_CALIBRATE?

Andrew,

Thank you for this excellent guide - I am utilizing it to get my printer (V2.4 300mm^3) ready for PIF service.

Quick question - I am adding BED_MESH_CALIBRATE to my workflow as well as heat soaking. Do I need to add a final G28 before I start printing?

e.g: heat soak --> BED_MESH_CLEAR --> G28 --> QUAD_GANTRY_LEVEL --> G28 --> BED_MESH_CALIBRATE --> G28 --> print

Best regards,

Mark Steele - V2.2024

TEST_SPEED macro that prints

I find what limits my acceleration the most is using pressure advance with a boden tube. Pressure advance seems to ignore the acceleration limits of the extruder. It would be cool if you had a similar speed test macro that printed to help find the printing limits, especially while using pressure advance.

Question: Is temperature calibration included in this guide?

Hi :)

I couldn't find temp tuning and I wondered what its relation would be to pressure advance and input shaper.

Thanks for putting this great guide together!!

Theo

Instructions for 24V fans using tachometer pin don't work well

Simply plugging the tachometer pin into a free endstop port doesn't work well. The pullup voltage on the endstop pins is too low compared to the 24V power to the fan, so the fan spins constantly, sinking current through the tachometer.

My suggestion to fix this is to add the following:

• BAT85 Schottky diode facing away from the MCU (just like with the inductive probe)
• 10k resistor pullup to 24V on the fan side of the BAT85. I used one of the available always on FAN ports to get the 24V.

This solved the issue for me. It wasn't particularly pretty having all of those components along the tachometer wire, but it worked. The fan behaved normally, and the macro to detect fan stoppage worked as well.

TEST_SPEED macro considerations for CoreXZ machines (ie, Switchwire)

Hi. Finished tuning my V0.1 and moved on to my SW.

The TEST_SPEED macro reported wildly different positions for the X-axis upon completion of the test. Scratched my head for awhile, then realized that the Z motor contributes to X motion. If the Z beginning position is not returned to the position at the initial GET_POSITION state, then X position will be way off (more than 6000 microsteps in my case).

Not sure how to best handle this. Simplest seems to do the GET_POSITION after a full home (G28) at beginning and end, but I don't know if that introduces other problems. I may just try to undo the G0 ... Z{bound+10}...

In any case, it is probably a good idea to mention in the docs that for CoreXZ machines, the Z position values should be compared as well.

Thanks for the guide! My V0.1 is purring like a tiger!

Frame temperature compensation

Hi @AndrewEllis93,

first, thank you very much for your great tuning bible ;-)
At the moment i´m struggeling with the first layer consistence if my Trident 250 is heating up over time.
I´ve got the idea to adjust the gcode offset in dependeny of the frame temperature. I think it is the same way you do it.
Can you please share your START_PRINT macro so that we can see how you realised the frame temperature compensation?

Best regards
Markus

fix Link to macro

Print-Tuning-Guide/articles/determining_max_speeds_accels.md

Line 58 in 11d9497

The macro is available [:page_facing_up:here.](Macros/TEST_SPEED.cfg)

Consider mentioning nonlinear extrude for RRF on "PA/EM Oddities"

PA/EM Oddities mentions

Assuming EM is well tuned, this is often caused by a slight 📄 flow dropoff at higher print speeds — since most people print their perimeters faster than their top layer.

and then offers a couple options to address this. It is omitting RepRapFirmware's Nonlinear Extrusion setting, which could also be used to compensate this in firmware. Since the RRF documentation is sparse, and no good way to tune this exists, I would expect this to appear low in the list of mitigations.

Example voron2 klipper configs missing

Hi Andrew,

I have recently been building my Voron 2 and your "backup klipper configs" were a huge help for me. I just wanted to look up your thermistor setup and noticed they were gone. Any chance you could upload them again, maybe add them to the print tuning guide next to "my setup"? Those were really helpful to me!

Thank you for all your articles and writeups. Your mentioning of belt slack on the individual z-stepper units actually made me fix the probing issues I have had :)

best
Alex

small typo (swapped letters)

https://ellis3dp.com/Print-Tuning-Guide/articles/first_layer_squish.html

A_ZDJUST must be Z_ADJUST?

Determining Maximum Speeds and Accelerations

Hi,

Thanks for your great guide!

I wonder if you could give a "safe" acceleration value to use when testing speed. I assume if the acceleration value is chosen too low, the maximum speed will never be reached depending on printer size?

Pressure Advance Calibration Tool

M190 & M109 are issued before print start, which doesn't work for us TAP users. Basically it heats up goes to the QGL and bed mesh, etc, but the nozzle needs to be <=150, so it drops the temp and then, since the M109 command was already complete, it doesn't heat back up and the print fails.

FR: hope we can include in the pressure advance lines method a bed for delta printers

hope there is an option to change the bed shape to circle for our delta printers

https://realdeuce.github.io/Voron/PA/pressure_advance.html

EM Automation with PrusaSlicer

Hey, thanks for the great work! Sorry for creating an issue here, but I wanted to sail across my input without colliding with your writing and styling.

I came up with a solution for EM automation in prusa and I'd like to share it. You may even want to integrate it into the guide.
The idea is to use sequential printing, which lets you use a "Between objects G-Code" in the printer settings custom G-Code options. Since the slabs are only 3mm high that shouldn't cause a big challenge for any hotend.

Needed are three macros, the first one only to ensure the variable is properly set and to allow for step and start percentage settings. Could be used in the start code or issues manually. The other ones increase or decrease the EM with each object succession.

[gcode_macro EMTUNE]
variable_em_step: 2
variable_em_value: 100
gcode:
    {% set EM_STEP = params.STEP|default(2) | int %}
    {% set EM_VALUE = params.VALUE|default(100) | int %}
    SET_GCODE_VARIABLE MACRO=EMTUNE VARIABLE=em_step VALUE={EM_STEP}
    SET_GCODE_VARIABLE MACRO=EMTUNE VARIABLE=em_value VALUE={EM_VALUE}
    M221 S{EM_VALUE}

[gcode_macro EMPL]
gcode:
    {% set EM_STEP = printer["gcode_macro EMTUNE"].em_step | default(2) %}
    {% set EM_VALUE = printer["gcode_macro EMTUNE"].em_value | default(100) %}
    {% set EM_VALUE = EM_VALUE + EM_STEP %}
    SET_GCODE_VARIABLE MACRO=EMTUNE VARIABLE=em_value VALUE={EM_VALUE}
    M221 S{EM_VALUE}

[gcode_macro EMMI]
gcode:
    {% set EM_STEP = printer["gcode_macro EMTUNE"].em_step | default(2) %}
    {% set EM_VALUE = printer["gcode_macro EMTUNE"].em_value | default(100) %}
    {% set EM_VALUE = EM_VALUE - EM_STEP %}
    SET_GCODE_VARIABLE MACRO=EMTUNE VARIABLE=em_value VALUE={EM_VALUE}
    M221 S{EM_VALUE}

Cause of Pockmarks

I was getting pockmarks with some new asa and I'd never had the issue before but I saw someone say to reduce retraction. Even though the actual retractions were not near the pockmarks, it still fixed the issue. I think this would be useful to add to your guide.

No fan speed in new PA tuning tool

Sliced two files. Both at 50% fan speed entered into the online form.
Both files sliced with 0% fan speed.
Two M106 S0 commands found.

Line 94 and 514 in the attached gcode file.
Screenshot of settings used included.

TheGoo_PA_Tune_003_004.txt
TheGoo_PA_Tune.txt

Pressure advance tool

In the Pressure Advance / Linear Advance Calibration Tool the End G-code doesn't work as intended. When changing to something else than the preset it doesn't take. The generated gcode still contains the preset gcode, which is "PRINT_END ; End macro. Change name to match yours".

Dynamic acceleration / square corner velocity (jerk)

Hi Andres,

I was missing this part in your tuning guide, so thought to mention it.

With SuperSlicer there is the possibility to set accelerations per extrusion type (e.g. external perimeter, infill, travel).
The square corner velocity (SQV)/jerk can not be set this way in the slicer. It can be done with a post-processing script, or with slicer g-code substitution or in the printer firmware. I used the latter and made an override of the M204 macro in Klipper. It deduces both 'acceleration to deceleration' and SQV form the requested acceleration by the slicer (M204).

What do you think about using high accelerations and SQV only for travel?

Stepper stutters while doing the speed test

When I use the macro for Klipper to perform the speed test my x axes runs all the way to the right "crashing" in the right "endstop" and then starts stuttering while waiting for the hotbed to run towards the front. After the speed test the steps are off by more than 10k bc of the stutter. I dont know how to fix this :/ I looked through the macro Code but could not figure out why it is doing this.
Using an Sovol SV06.

Hi , i read your Print guide, great work, but 1 misstake

hi, i read your print guide, it says that cura can't do PA. It has been able to do this with a plugin for 2 years.
Here the Link to Cura Plugin for Pressure Advance :
https://github.com/ollyfg/cura_pressure_advance_setting

Consider additional to layer shifting section

Under: Troubleshooting > Layer Shifting > Mechanical

Consider adding detail about checking drive pulley setscrew. It took me ages to track down a stupid setscrew that was just loose enough to cause layer shifts after some direction changes.

Klipper resume after power loss