Is it really faster out of the box?  

RE:

Why is the infill on the MK4 at 20% and on the Mini at 15% ?

It will not make much difference but this affects the print time.

Here is a video with a practical test.

As the printers and profiles in prusa slicer are now.

Mini, MK3(x), MK4 and the XL.

RE:

Mini+ was at 20% and MK4 at 15% ... i think this was the default for Prusa Mini+ and MK4.
Thanks for the video.

RE: Is it really faster out of the box?

With the same infill 15% i get the following:

Mini+ =  8h 28m (no difference)
MK4 = 8h 12m

I was expecting better speed improvements...

RE: Is it really faster out of the box?

For a 175x175x50 box, the difference is pretty good.

Mini 14h43m

MK4 10h23m

Difference of 4h20m.

But I am sure there will be settings with turbo speeds to come. 😉 

RE: Is it really faster out of the box?

Hello,

I would like to stress the concept that Cartesian printers like the Mini, MK3S or MK4 could not reach the high speed of CoreXY (like the XL, X1 Carbon, Voron) due to mechanical constrains.  So, don't expect any future firmware update that can enable a MK4 to reach a speed of 400 mm/s 🙄  and 15000 in acceleration: the printer mechanic would explode 😓

Looking on the current MK3S+ PrusaSlicer profile settings, in QUALITY mode, we find external perimeters at 25 mm/s and infill at 80 mm/s. The SPEED profile raises the values to 200 mm/s for infill and 35 mm/s for external perimeter.
I would expect for MK4 HI-SPEED values to be 280 mm/s for infill and 90 mm/s for external perimeter that together with all the low speed values presents in other settings could reach something around two time the MK3S speed, a good average of 120-150, 200 at best mm/s. 

Regards

RE: Is it really faster out of the box?

They haven’t released the input shaper part of latest release I imagine soon 

RE:

Posted by: @antimix

I would expect for MK4 HI-SPEED values to be 280 mm/s for infill and 90 mm/s for external perimeter that together with all the low speed values presents in other settings could reach something around two time the MK3S speed, a good average of 120-150, 200 at best mm/s. 

well... a reliable 280mm/s for infill in an ipotetical highspeed mode will be great.

RE: Is it really faster out of the box?

You guys are too optimistic, I think.

Consider that the MK4 is basically the same as the MK3 series regarding mechanical structure, sliding bed, smooth rods ,X,Y and Z steppers. In all fairness, speeds should be on the same ball park. If, eventually, Prusa manages to enable correctly input shaper on the MK4 ( adding accelerometers and sorting out the software implementation )  the new printer would be able to reach faster speeds keeping the same quality, but don’t expect night and day. A gain of 25-30% seems more reasonable.

Don’t forget also that printing speed is limited by the capability of the hotend to supply molten filament. In this regard the Nextruder of the MK4, according to several reviews, has a similar performance to the V6 of the Mk3/S/+, so no real gain here.

In top of that, you must consider the maximal  volumetric speed. Look at the this setting on the profiles for  PLA, PETG, ASA,PC, TPU, etc..  For Prusament PETG ,PC, or PC-CF  the values range from 8 to 10 mm3/s. That means that with a 0,4 nozzle and 0.2 layer height you can’t surpass 90-110 mm/s, regardless of the speeds you set in the printing configuration. Of course, you can override the max.vol. speed setting, cranking it up at your wish, but those values are there for some reason.

RE: Is it really faster out of the box?

Two things to consider.  One, what do you mean by "faster".  Two, can it become "faster".

In theory, right now, the MK4 isn't any "faster" than the MK3 or Mini.  You can't use the build in slicer presets to judge whether it is "faster" or not.  I have presets for my MK3 that make them print objects in half the time of the slicer presets.  I think when we ask about "faster", were looking for shorter print times at a high quality.  Right now, without input shaper, the MK4 isn't any "faster".

There are a few limits on shortening print times, but for small detailed objects, the limit really comes down to acceleration.  Without input shaping, you can still get short print times on a Prusa, but you'll have tons of artifiacts/ringing/ghosting/etc.  Once they implement the input shaper functionality, you should be able to get much better performance from a MK4 than from the mini or MK3.

The second limit is volumetric flow, or how fast you can melt filament.  Right now, the hot end that comes with the MK4 isn't a high flow hot end.  So, you can't melt plastic significantly faster than you can with the other printers (that's my understanding, anyway, I haven't tested myself).  So, if you are printing larger objects that don't have tiny details, and you're using bigger layer heights, the printer is going to run out of melting ability before you hit a ton of speed.  You can get over this now by using an adapter and a CHT nozzle for higher flow, but I haven't tried this yet.

The extruder on the MK4 seems to be an improvement, from my few tests I'm seeing much better quality on flat vertical walls.

Once the input shaper is enabled, and they release a high flow hot end (or you use an adapter and use something like a CHT nozzle), the MK4 SHOULD BE FASTER at a high quality.  You should be able to run at faster speeds and acceleration, and melt plastic fast enough to keep up.  Will it be as fast as a Bambu, or other core x/y machine?  Probably not, but having owned a Bambu and a Voron, I will take reliability and consistancy over the fastest possible speeds.  The faster machines are great, but also nowhere near as reliable as my Prusas (in my experience).

So, getting a MK4 today, if you are just using the stock profiles, will not be a night and day difference.  If you're using the stock profiles, depending on what you're printing, you likely can adjust them NOW to print objects in a shorter time period.  However, once they've figured out the bugs and have input shaper enabled, the MK4 should reliably print things in a shorter amount of time.

RE:

By "faster" I meant that changes had been applied to the standard profiles to allow, with the same print reliability, greater speeds (for this reason I wrote "out of the box"). Then obviously by changing the parameters you can change the performance, but I had read, superficially, that changes had been made to allow at parity of quality, better printing speed, so I asked. I did not know, however, that some of these innovations (input shaper) are still not there.

RE: Is it really faster out of the box?

I'm sure once they have input shaper developed and implemented, they'll have presets which definitely shorten print time by a good amount, but that's the main thing that's going to let you keep the quality high at higher speeds.  Their other improvements are all great, but don't necessarily help you get faster printing, just better quality overall.

RE: Is it really faster out of the box?

Finally MK4 is starting its race with the new IS enabled firmware ... 😉 

RE: Is it really faster out of the box?

I print the same prints with mk3s+ and mk4, need a lot os similar stuff printed.

I'd say the mk4 is about 20% faster for the things I print right now.

RE: Is it really faster out of the box?

Posted by: @aake-e

I print the same prints with mk3s+ and mk4, need a lot os similar stuff printed.

I'd say the mk4 is about 20% faster for the things I print right now.

The 20% (without IS) comes from the fact that Prusa has stored different settings for the MK3 and MK4: f.e. 20% vs. 15% infill and Gyroid vs. Rectilinear.

RE: Is it really faster out of the box?

Posted by: @marcelismus

 

Posted by: @aake-e

I print the same prints with mk3s+ and mk4, need a lot os similar stuff printed.

I'd say the mk4 is about 20% faster for the things I print right now.

The 20% (without IS) comes from the fact that Prusa has stored different settings for the MK3 and MK4: f.e. 20% vs. 15% infill and Gyroid vs. Rectilinear.

Is it? I don't have the MK4, but on PrusaSlicer 2.6.0 beta 4, I loaded up a model I've printed. I got all times for 0.4 mm nozzle, 0.2 mm QUALITY preset. MK4 defaults to grid (not rectilinear), while MK3S+ defaults to gyroid (same as my MINI+). Both the MK3S+ and MK4 presets used 15% infill by default, so all times are 15% infill. Times are in hour:minute format.

MK3S, gyroid: 20:29
MK3S, grid: 16:55
MK4, gyroid: 17:47
MK4, grid: 13:59

So even without input shaping, the MK4 is faster than the MK3S, for the same infill type and same infill %, with the MK4 13% faster with gyroid vs gyroid, and 17% faster with grid vs grid. So the MK4 is faster, even with the same infill type and %, without input shaping, than the MK3.

Then for fun, I threw in the MK4 input shaper profile, and got:

MK4 IS, gyroid: 11:56
MK4 IS, grid: 08:30

However, this is just for fun, because there's no QUALITY profile with input shaper, there's one and only one input shaper profile (so far). So these times are probably best compared to the SPEED profile the non-input-shaper profiles have. Eventually I'm sure there will be input shaper profiles that have the same options for quality or speed as the existing ones.

RE:

Both the MK3S+ and MK4 presets used 15% infill by default, so all times are 15% infill.

Interesting. The 0.2 mm SPEED profile for the MK3S is 20% grid default. This means the same model was now more than a 24 hour print! Probably the MK3 always used to be 20%, and they updating it to be 15%, but missed the 0.2 mm SPEED? (My MINI profiles all default to 15%, so if these used to be 20%, they must have decided 15% was good for infill defaults)

Dropping it to 15% grid like the rest, the MK3S SPEED is 14:39. The MK4 0.2 SPEED, 15% grid is 12:20.

So when comparing apples to apples, the MK4 without input shaping appears to always be faster than the MK3.

RE: Is it really faster out of the box?

On this video, you can see how the MK3S is really very very similar to MK4.

This youtuber used the MK4 profiles and GCODE with small changes on settings on the MK3S, and surprise surprise... they got performances very very similar to the MK4. 😍 

You can see how they printed on an MK3S the same first perfect PLA layer sheet as the MK4 does (but it is on an MK3S with PINDA !!)

But even better, they used the MK4 settings on MK3S and they were able to speed up the print and transorm the MK3S in a fast MK4 (but with no input shaping).

In the video, PrusaSlicer  planned a print for an MK3S in 7h:46m estimated. Then, they applied the MK4 0.6 FAST profile on the MK3S (that had a 0.4 CHT nozzle) and they just changed only the layer height to 0.32; in this way they were able to speed up the print on the MK3S from around 7 hours to  3h:26 min ! 😲 

Basically your old MK3S is much, much, better than you had ever though. 😊 

Sorry for the Italian audio without captions, but you can see the results. And they used a small server rack as cheap enclosure (around 100$).

Regards

RE: Is it really faster out of the box?

It seems like there are so many misconceptions, and lack of understanding, in how printers work, and what makes them "fast".  What makes the MK4 physically move faster than the MK3?  Pretty much nothing.  The newer stepper motors may allow it to travel faster (that's an area I don't know about), but the two printers are practically the same when it comes to movement.

We're not interested in whether the MK4 can move faster.  We're interested in whether the MK4 can print faster, with high quality.  The one main thing that allows for that is input shaping.  That compensates for the fast motion, to eliminate ringing in the parts.

I think a lot of people get hung up on the presets for the machines.  The presets do not define what the machine can do, they are there to give you a reliable configuration that will (presumably) work well every time.

You can go into the settings for you MK3, increase the limits for acceleration and speed, and print things in half the time as you'd get with the presets.  I've been doing this for some time, with great success.  You'll also see many more artifacts in your prints.  How much this matters depends on your end goal.

The main improvement in the MK4 that will allow for faster printing has nothing to do with the movement hardware of the printer.  It's the Input Shaping.  That's the game changer in terms of printing faster on a prusa.

If you print with an MK3, using the presets, and want it to print faster, that's easy.  Increase the acceleration limits to 3,000.  Increase the speeds to 150 (or 200, or whatever).  At some point, you're limited by the flow of the hot end (which is the same for the MK4).  You can get faster speeds, but you are going to see ringing/ghosting, and other artifacts.  If these don't bother you, than you're golden.  Depending on what you're printing, you may not even notice them.

Here's what the MK4 is actually giving you in improvements:  Firmware (input shaping, pressure advance, other misc small changes).  Extruder system (more consistent and reliable pushing of filament).  Nozzle sensor (easier first layer calibrations).  Upgraded steppers (reduction in vertical artifacts, which some people may not even notice).  Nozzle system (easier nozzle changes).  There are a lot of good improvements there, and it's definitely a step up in many ways.  However, what you aren't getting is a redesigned motion system.  The movement of the machine is basically identical to the MK3.  So, let's understand that the "faster" part of the MK4 has almost nothing to do with the actual machine, it's the software (input shaping).

RE: Is it really faster out of the box?

Lots of talk about waiting on mk4 to use input shaping. Just throw klipper on your mk3 with an old pi3 and buy a cheap adxl345 and you'll gain all the same benefits without waiting or "upgrading" to essentially the same printer. You'll still be limited to approximately 11mm/s flow rate of the e3d v6 unless you upgrade to something else (the revo isn't an upgrade) but there's nothing stopping you from throwing a mosquitto magnum or dragon hf on there once you realize the prusa firmware is only holding you back. Don't be afraid to tinker. You can always flash the old marlin back on there if you need to.