Print setting for parts. Have to use PRUSA's setting for input shaper to work?
 
Notifications
Clear all

Print setting for parts. Have to use PRUSA's setting for input shaper to work?  

  RSS
Category 5
(@category-5)
Eminent Member
Print setting for parts. Have to use PRUSA's setting for input shaper to work?

Prusa always recommends printing parts with 2 walls, and 20% grid infill.  When I did my Bear, they recommended 4 walls, and gyroid at 15%.  The parts are much more robust, and a little heavier.  I was planning to print mine like the bear parts but then I started thinking that the mass and resonance of the parts would vary, and the "one size fits all" input shaping might not work well if my parts are different from the prusa recommendations.

I decided to go with the g-codes that are now available for MK4 upgrades.  What is even weirder Is that the Prusa provided G-Codes use gyroid infill instead of grid, like what is recommended in the print instructions.

Also, I usually use the textured sheet, but it seems like PRUSA hasn't placed all the parts with the visible surface in mind.  Some parts mate on the textured side, which seems like it would be less precise.  Should I be using the satin sheet instead?

Prusa definitely needs to offer an accellerometer upgrade for the MK4 because all of these little variables are going to affect the performance of the input shaper, and I think that's why we are already seeing less definition on edges and fine details with input shaper active.  there is just too much variance in the resonance from printer to printer.

Posted : 15/08/2023 1:36 pm
Diem
 Diem
(@diem)
Illustrious Member

PRUSA hasn't placed all the parts with the visible surface in mind

Correct.  I would expect Prusa to place the parts according to their required strength and the vectors of the forces they are expected to encounter.

Cheerio,

 

 

Posted : 15/08/2023 2:45 pm
jvasileff
(@jvasileff)
Trusted Member
RE:

In one of the youtube interviews with Josef Prusa (sorry, don't remember which one), he mentioned that most of the resonance comes from the belts. That would help explain they think one setting is good enough for all MK4s.

Also note that Prusa supplied parts are printed very differently than what they advise. Their print farm pumps out fairly low quality parts with thick layers, which work perfectly fine(*) for the most part. The Printables page for the MK4 parts even says "Note that print settings are recommended, but you can experiment with your own.". Clearly they don't think the exact settings matter too much, and they should be the ones that know.

(*) The exception to this appears to be the extruder parts. I and others have experienced filament twisting which has been improved by rebuilding the extruder with better quality 0.2mm or better layer height prints. But even this doesn't seem to be a big deal, as the filament twisting does not affect print quality.

Posted : 15/08/2023 3:23 pm
jseyfert3
(@jseyfert3)
Reputable Member
RE: Print setting for parts. Have to use PRUSA's setting for input shaper to work?

And/or in the orientation that allows them to be printed without supports.

Posted by: @diem

PRUSA hasn't placed all the parts with the visible surface in mind

Correct.  I would expect Prusa to place the parts according to their required strength and the vectors of the forces they are expected to encounter.

Cheerio,

 

 

 

Posted : 15/08/2023 5:19 pm
Rainer
(@rainer-2)
Reputable Member
RE: Print setting for parts. Have to use PRUSA's setting for input shaper to work?

Prusa definitely needs to offer an accellerometer upgrade for the MK4 because all of these little variables are going to affect the performance of the input shaper, and I think that's why we are already seeing less definition on edges and fine details with input shaper active.  there is just too much variance in the resonance from printer to printer.

In this thread I have linked the Gcode and instructions to determine the frequencies for input shaper.

Then you will also see that Input Shaper works over a wide range.

The loss of fine detail and rounded corners is a downside with InputShaper. Nothing comes for free.

An accellrometer doesn't change that.

I also recommend the live video from nero3d.

Posted : 15/08/2023 9:39 pm
Category 5
(@category-5)
Eminent Member
Topic starter answered:
RE: Print setting for parts. Have to use PRUSA's setting for input shaper to work?

Eh. There are printers that manage to maintain detail just fine.  It’s definitely a drawback of the way Prusa is doing things.   

Posted : 16/08/2023 12:51 am
Rainer
(@rainer-2)
Reputable Member
RE: Print setting for parts. Have to use PRUSA's setting for input shaper to work?

What should they do differently technically?

I don't know the comparison live. But in videos comparing Prusa with IS with e.g. Bambulab the difference was not big. Mostly even a bit better with prusa.

The only technical downside with the MK4 is the moving bed. With the print size comes more weight. This is attempted to reduce by changing the frequency at Y and decreasing the acceleration.

Prusa uses the same input shapers as klipper. And I'll bet my left egg that Bambulab doesn't use anything else under their proprietary interface either.

If you want to print large, heavy parts quickly with the MK4, you simply bought the wrong printer. For everything else, the MK4 is really good for me.

Input Shaper is not a magic bullet. You don't get anything for free. Everything costs something. In this case, the corners are slightly rounded. The more aggressive the input shaper, the more. But Prusa uses the smooth MZV.

 

Here from the klipper documentary:

Note that at very high accelerations, depending on the resonance frequency and the input shaper you chose (e.g. EI shaper creates more smoothing than MZV), input shaping may cause too much smoothing and rounding of the parts. So, max_accel should be chosen such as to prevent that.

 

Here you can read that too much acceleration rounds the corners too much. Which also means that the corners are always slightly rounded.

Posted : 16/08/2023 6:41 am
Share: