Klipper firmware is considered to be a speed
at quality boost. Features like Input Shaping and Pressure Advance should level-up every
3D printer. But how much can you REALLY profit from using Klipper? I’ll give you the answer
comparing four low-end bed slingers before and after Klipper installation! Faster printing
is one the big things in FDM at the moment. Manufacturers like Fokoos and Anker advertise
their newest bed slingers being five to six times faster, which is… in my opinion…
quite misleading: What really matters are the gains in actual print time… Klipper
might be one of the most advanced Firmwares for FDM printing at the moment. Using input
shaping along with higher accelerations as well as pressure advance should improve the
print times significantly without sacrificing print quality. Is it really worth the efforts?
I checked it out with these four cheap Bed Slingers: The printers for this video were
provided by Geekbuying, BIQU, Fokoos and Kingroon. The BIQU B1, the Artillery Hornet and the
Fokoos Odin-5 F3 can be considered as quite typical Ender 3 clones. The Kingroon KP3S
is a cheap and interesting smaller scale printer I wanted to include into this comparison.
Unpackaging and building these printers was quick and easy…except for the BIQU B1: There
is quite a bit more to assemble and to adjust correctly, which could cause potential problems
for absolute beginners. Concerning the spares and accessory you get with those printers:
Not worth mentioning the differences. Everything is provided in order to get started with these
machines. But, instead of using the reskinned or old CURA versions provided with the printers,
I went for an easier and better way. All of them are i3-style printers – just like the
Prusa MK3, which comes with thoroughly optimised profiles in Prusa Slicer. As a baseline for
the stock Marlin performance, I simply used the 0.2mm quality pre-set of the MK3, which
works perfectly on almost every bedslinger. The only settings that need to be adjusted,
are concerning some basics for the machine itself, extrusion and retracts as well as
the automatic cooling settings. Setting a viable “Max volumetric speed” value is
very helpful, as it automatically limits the printing speeds according to the performance
of your extrusion system. In order to check if there is any unexpected behaviour on these
printers, I did a quick flow rate test on PLA+ with Stefan’s helpful tool. The KP3S,
BIQU B1 and the Artillery Hornet perform at around E3D V6 Level. The Fokoos Odin, with
its Volcano clone and custom extruder, performs is around 35% superior to the other printers.
So, no surprises concerning extrusion. For the automated cooling settings, I printed
two test models iteratively on each printer in order to determine the minimum layer time
and the minimum print speed for each printer: The circular cooling concept of the BIQU B1
works very well, followed by the 4010 setup of the Fokoos Odin. The air duct attached
to the 5015 fan of the KP3S would profit from some optimization concerning the air duct
design. The Artillery Hornet though is underperforming: It might have two 4010 fans, but the airflow
is dispersed way too much here. I did a bunch of different prints on the stock printers
and documented the print-times. All in all, it’s a good mix for the comparison before
and after installing Klipper, as there is different emphasis on the prints – like
acceleration, speed or cooling performance. All prints came out well except for the Kingroon
KP3S. I was not able to get rid of the stringing and there were some visible extrusion flaws.
End of the story: The hotend was not properly assembled, leaving a gap between nozzle and
heat break. Therefore, the PTFE inliner of the heatbreak got damaged. I didn’t receive
a spare part in time, but I went on with the tests on the KP3S. But now, let’s finally
check out the Klipper boost on our low-end bed slingers! The Klipper documentation is
very well done and also very detailed. There is a ton of background information, which
is great on the one hand, but might scare off some users because of the amount of text
to read and understand. Some of the best Klipper tutorials you can check out additionally,
are made by Teaching Tech and 3Dprinterbeginner. Basically, the procedure is the same for every
3D printer or mainboard. If you follow Klipper’s official documentation step by step, you’ll
definitely succeed. I’ll show you why it is worth it… even for cheap low end bed
slingers! Input Shaping will help to multiply accelerations and speeds - at the same or
even better print quality. Pressure advance is a measure in order optimize the filament
flow, enabling quality printing at those higher speeds. Last but not least: Configuring firmware
settings in Klipper is very convenient in comparison to Marlin. You simply edit a text
file and restart the firmware. No recompiling and re-flashing… Though, in order to take
full advantage of Input Shaping and Pressure Advance you need to do measurements and calibrations,
as the settings need to match to the used hardware setup. For input shaping, I use the
automated measuring, as it more precise, but it requires additional hardware such as an
accelerometer and additional wiring. Also, very important to know: Klipper does not support
many displays natively, that connect to the printer’s mainboard. The “Klipper Screen”
addon supporting displays directly connected to a raspberry pi is an alternative you can
check out. The results and calculations of measuring resonances and Input Shaping gives
in some extend an evaluation of your printer setup. This concerns the construction itself
as well as the build quality, the weights of the moved masses in X and Y as well as
the belt tensions. As a result, you get recommendations for the maximum accelerations for each direction
as well as for the settings of the frequencies for the shapers. Acceleration should be validated
afterwards doing a test print procedure, which increases the acceleration in dependency of
the layer height – By judging the resulting ringing and smoothing, this helps you to find
out the max acceleration in order to not loose print quality. Feasible accelerations for
those cartesian printers range from around 3000mm/s² up to around 4000mm/s². Using
nowadays marketing terms like from Fokoos and Anker, this would be “six to eight times
faster” in comparison to the 500mm/s² of most Ender 3 clones. This will never ever
ever ever transfer into the real gains in print-times – we’ll talk about that later.
For all printers I also tweaked the extruder and z-movement settings. Thanks to pressure
advance, which is retracting all the time anyways, retract lengths can be lowered to
somewhere from 1 to 3mm – depending on bowden or direct setup. Also, retraction speeds and
accelerations can generally be set way higher than the stock values. A fast and save value
for PLA is somewhere between 70 and 100mm/s for retract speed and around 3000 to 5000mm/s²
for retract acceleration. Thanks to pressure advance and input shaping we can also go with
decisively higher speeds in our print settings. This time, I use the 0.20mm speed system preset
for the MK3 as a baseline. I did some test prints and, in the end, I increased the speeds
up to 100% depending on the feature. We are still limited by cooling and the max. volumetric
speed of the extrusion system, because Klipper won’t improve the hardware magically. I
repeated the set of prints I did before on stock firmware, but now with all the Klipper
features enabled and with the improved settings in Firmware and slicing. The quality is at
least as good as on marlin and the slower prusa 0.20mm quality pre-set. But how much
faster is Klipper concerning print times really? Klipper boosts these printers, being cheap
bedslingers, up to two times faster, which definitely is a tiny bit mind-blowing, but
far away from the marketing of Fokoos and Anker. Smaller prints like a Benchy for example,
which are cooling limited due to short layer-times, do not profit that much from Klipper, because
again: it doesn’t magically boost the cooling performance. Though, it drops the print-time
significantly all the way ranging from at least saving 30% up to 50% of the overall
print times. This can make a huge difference: This might help for your daily routines, your
progress in projects or your 3D printing business might become more profitable as you could
double the throughput of your machines. So: Is it worth it? From my point of view: Definitely,
because I am impatient and I love to go fast! But, I also have quite some experiences with
Klipper and the necessary calibration procedures. Furthermore: I mostly don’t really use or
need displays on my 3D printers as I always have a smartphone or a PC nearby. In the end
Klipper Installation and calibration is quite simple, but it can take time and a lot of
things to learn for beginners. On the other hand, reading, trying, understanding and succeeding
with new topics is fun in the end, right? However, there is good news for the future:
The commercialisation of Klipper for the mainstream has started and we’ll see Klipper out of
the box on more and more printers in the future. The only thing that concerns me a bit: Even
more than Marlin, Klipper needs well-tuned settings for Firmware and Slicing. After testing
some bed slingers the last months, I have to say: That’s not really the parade discipline
of cheap printer manufacturers… Bu,t we’ll see! RIGHT HERE ON 247 PRINTING! Oh, and don’t
forget to leave a like and comment and subs…blablabla – you know what to do!
