What to tune to get smooth sides on prints?

[Buurman]

I wonder, even following most popular tuning guides, most are looking into layers, and I think my printers are pretty well tuned.

Yet, my prints are not super smooth on the sides sometimes.. 

What is the most important setting here, I expect ... belts, but this 110hz tuning thing, is not my kind of thing, its so open to how hard you pull the belt, what sound you get .. I feel..

So, if belts are probably the most important, how do you "tune" your belts, or maybe other things to get smooth sides.. 

Today I will try to do lots of tuning, and maybe fill up this post to help someone else out, if it get me anywhere useful.

[smirk]

I've never been convinced by the whole "pluck your belts, they should should like this musical note" thing. Printers might be things of beauty but they're not musical instruments which have been designed to allow the bits that vibrate to do their  thing (vibrate). Whereas with a printer there's always something that damps the resonance, if you think how many belts run within mm of the length of an extrusion. There's also the unknown of belt length, the  length of the string (belt) controls how much it vibrates. So you can't really universally say a belt should vibrate at x Hz (unless you know the length and a bunch of other things). Perhaps deflection would be a better measure (how far can you pull the belt) but to do that properly requires a spring guage and naturally some notion of how much measurable force is required to move the belt X mm (or inches, microfurlongs or hides which are clearly better measurements)......but as you will have noted no such defined value exists. Clearly there is a opportunity for you to create your own mesaure.  I would suggest something like the belts should have x microBuurmans of deflection per microfurlong. I'm sure that could be an SI unit.

There's also the resonance compensation bit within Klipper to explore (I've never tuned it, always taken the defaults) but I imagine there's extra that could be squeezed out of that as belt elasticity/wiggling/microBuurmans will have a bearing on that.

[mvdveer]

Agree that tuning belts is very operator dependant. Following tuning guides, this is what you are supposed to do. However, I have NEVER hit that 110Hz EVER. And every time I try, I get a different result. Input shaping may yield better results, and thanks to nero3dp this has now become so much easier. 

My Pi pico's have arrived and I will be building a couple of these. (I need a backup...) 

Watch out @smirk - I coming for the other half of Scotland 

[smirk]

@mvdveer are those the wireless picos (well wifi enabled)?

[mvdveer]

@smirk three of the 5 are wifi enabled yes.

Here are the specs;

 

Raspberry Pi Pico W - (Wireless) supports WiFi 802.11n 2.4GHz out of the box along with BLE in the future. Best of all this board is manufactured by Raspberry Pi, which means excellent resources and a very stable port of MicroPython.

The wireless antenna is an onboard, similar to that used on the Raspberry Pi Zero boards. The pins are standard pitch, so it's easy to solder pins on and get going with solderless breadboards. For industrial users, the Pico W can be soldered in place using the castellated mounting holes 

Board Specifications:

RP2040 microcontroller chip designed by Raspberry Pi

On-board single-band 2.4GHz wireless interfaces (802.11n)

Dual-core ARM Cortex M0 processor, flexible clock running up to 133 MHz

264kB of SRAM, and 2MB of on-board Flash memory

Castellated module allows soldering direct to carrier boards

USB 1.1 Host and Device support

Low-power sleep and dormant modes

Drag & drop programming using mass storage over USB

26 multi-function GPIO pins

2×SPI, 2×I2C, 2×UART, 3×12-bit ADC, 16×controllable PWM channels

Accurate clock and timer on-chip

Temperature sensor

Accelerated floating point libraries on-chip

8×Programmable IO (PIO) state machines for custom peripheral support

 

[Simon2.4]

1 hour ago, smirk said:

So you can't really universally say a belt should vibrate at x Hz (unless you know the length

I second that!!

[Buurman]

2 hours ago, smirk said:

 I would suggest something like the belts should have x microBuurmans of deflection per microfurlong.

Hahahahaha, made laugh there.... lets invent a new standard.. sure!!

[Buurman]

52 minutes ago, mvdveer said:

Agree that tuning belts is very operator dependant. Following tuning guides, this is what you are supposed to do. However, I have NEVER hit that 110Hz EVER. And every time I try, I get a different result. Input shaping may yield better results, and thanks to nero3dp this has now become so much easier. 

My Pi pico's have arrived and I will be building a couple of these. (I need a backup...) 

Watch out @smirk - I coming for the other half of Scotland 

ORdered and will be using this!!!! thanks!! I didnt feel like putting extra wires into the printer, so this is a perfect solution!!!

[claudermilk]

2 hours ago, Buurman said:

Hahahahaha, made laugh there.... lets invent a new standard.. sure!!

Obligatory XKCD link. 

https://xkcd.com/927/

I get pretty good sides on my prints within the limits of FDM printing (i.e., there's *always* going to be layer lines). I just make sure the printer is all squared up, bolts tight, bed is leveled (Klicky + auto level). I have run input shaper, repeated every time I change the printer or the table it's on. I tensioned the belts using the tension meter tool found here: https://www.printables.com/model/46639-tension-meter-for-the-gt2-belts-of-i3-mk3s-or-prus I did print in PETG as suggested.

[Buurman]

57 minutes ago, claudermilk said:

Obligatory XKCD link. 

https://xkcd.com/927/

I get pretty good sides on my prints within the limits of FDM printing (i.e., there's *always* going to be layer lines). I just make sure the printer is all squared up, bolts tight, bed is leveled (Klicky + auto level). I have run input shaper, repeated every time I change the printer or the table it's on. I tensioned the belts using the tension meter tool found here: https://www.printables.com/model/46639-tension-meter-for-the-gt2-belts-of-i3-mk3s-or-prus I did print in PETG as suggested.

on what length of belt do you test it? I mean on a 40cm of belt the tension is different than 10cm of belt (on the same belt).

But I like the idea for sure.. the part "you need use Prusament PTEG" ... genius! hahahaha

[Simon2.4]

1 hour ago, Buurman said:

I mean on a 40cm of belt the tension is different than 10cm of belt

Actually, belt tension is not affected by the gantry's position. just place the gantry in a way where you can use the tool with ease.

[Buurman]

2 hours ago, Simon2.4 said:

Actually, belt tension is not affected by the gantry's position. just place the gantry in a way where you can use the tool with ease.

You know I will test that  but I will print it, so strange the made it a gcode download... but guess a little edit goes a long way...

[Buurman]

the gcode version is annoying, took out the beginning part, replaced it with PRINT_START

But gotten an extrusion error, instead of fixing, looked for an STL version...

Found versions in STL - Tension Meters in 6mm and 9mm.

https://www.printables.com/en/model/115460-belt-tension-gauge-source-file-included/files

[Buurman]

On one printer i think I got the smoothness pretty tuned down...

The other one, needs a lot more love... 

 

These are NOT direct light, cheating pictures, this is just how it is, looks better in real life.. 

[Buurman]

@claudermilk these belt tensioners are absolutely amazing, they should be in the tools set of a Voron built.

You DO need to print them in PETG, I tried PLA+ but its too stiff, so it will give a lower value, but at least you can make them all the same.

Put it on a belt, and you also see direct result with just a small twist of your screwdriver, so they really work.

Sorry for the out of focus pictures, stupid iphone kept on switching between macro and normal.. 

 

[smirk]

Looking at the picture that looks roughly 1.278926 uBuurmans per MicroFurlong, possibly 1.278928 as you say the picture is a bit blurry. Definitiely a good and visible way of getting all the belts to a consistent tension.

[Buurman]

Slowly going through all the tuning parts.. my lord.. its days of work...

https://github.com/AndrewEllis93/Print-Tuning-Guide

But getting clearly better and better results...so its worth it, I can advise it!!

First tuning the bottom, now tuning the top layer... Extrusion Multiplier time..

My setting used to be close to the last two blocks, so, I expect improvement. 

These blocks are tuned to make the bad effect worse, read well when printing them. (12 top layers)

[Simon2.4]

52 minutes ago, Buurman said:

Slowly going through all the tuning

I've made so many changes to my printer since last I did this... once I'm done with my current modding, I need to do it again.

[smirk]

5 hours ago, Buurman said:

so its worth it

Appreciate the effort going into this!