Hey folks, Alan Mandic, MandicReally here. I've got some tuning to do on a 3D printer, namely
this big blue monster right here. Not this one! I'm talking about the Elegoo Neptune 3
Max, but what I'm going to show you is the same process
I apply to every 3D printer I use, no matter the firmware
or configuration of the machine. Let's get into some tuning. I am not at all going to claim
that this is the end all, be all tuning guide for a 3D printer. This is the way that I approach it
& the way I'm going to approach it on this machine. I need to do it anyway. So I figured I'd bring you folks along
so maybe you could learn something. Or maybe you have some input you could leave in the comments down below to let me know
how you would do things differently. One of the first things I do
on a new printer out of the box is tune the extruder step count, aka E-Steps. On Klipper,
we would generally do Rotation Distance. I'm running Marlin on this machine,
so I'm going to do E-Steps. So I'm first going to heat up the hot end
to extrusion temperature on this machine. Then I'm going to mark the filament away
from the extruder assembly. I'm using a pair of calipers
that I will set against the top of the extruder housing and mark
120 millimeters of length. Now I'm only going to be extruding 100 millimeters of filament
for this testing process. But if I marked at 100 millimeters
and I was over extruding, I would actually go past that mark
and not be able to measure the distance that I over extruded by. Now that I'm heated
and I have the filament marked, I just need to command the machine
to extrude 100 millimeters of filament. Lucky for me, the menu system
on the Neptune 3 Max allows me to enter a custom value
and I can easily do this with my machine. Preheated the filament loaded and marked. I extrude 100 millimeters of filament. From here on, I would recommend
writing down the numbers you're working with
so you can keep track of them. I'm going to go into the menu system
of this machine and see what my E-Step number is. In this case, it's 415. Then I need to measure how much filament
was actually extruded. So I'm going to go from the top of this
housing up to the line that I marked previously 120
and see how much that distance is. In an ideal world, there would be exactly
20 millimeters there. It's usually not exact though. In my particular case
I'm measuring at 20.45 mm. Which means if I subtract
that from 120 millimeters, I extruded 99.55 millimeters of filament, when I commanded this machine
to extrude 100mm. I'm going to take these numbers, I'm going to plug them into a formula
to give me my new step number. The Teaching Tech calibration website
from Michael has a calculator on it. So if you're not the greatest with math, it can really help you out
getting through this process. For your information, this is the formula
and it does require the order of operations to be followed. So that means that I first need to take
the commanded extruded amount, 100 millimeters and divide it
by the actual extruded amount, the measurement that I took. In my case, 99.55. I now need to multiply that number that
I just got by my current E-Step number, which is 415,
and that will give me my new E-Step value. On the menu system of the Neptune 3
Max I can only go in one step increments, I cannot go in decimals. If I was using a laptop or a computer
to remote into the firmware, I could go a little finer on my adjustments, but
honestly, I'm not that worried about it. I recommend you do this three
maybe four times in a row because you're going to be narrowing in
on a good number. Your variation between the way you measure
filament from one time to another has some room for error. So you're going to get finer
and finer results the more you experience you have with this
and the more tests you do back to back. In fact, I had previously tuned this machine's
E-Steps, but in the course of filming the process for you folks, I came up with a different value
and ended up changing my settings. If you're 3D printers, menu
system does not allow you to see Steps or adjust them, you will need to use a computer
and a USB cable with some software to terminal into the machine
and issue G-code commands. The Teaching Tech calibration website
has a pretty good guide on how to do this. Before I move on to the fine filament
tuning, I'm going to install a BondTech 0.8 millimeter nozzle onto the Neptune
3 Max so I can do some really big prints for my review of this machine
Let's get this thing installed! With that 0.8 millimeter nozzle installed
we're ready to start tuning filament. We are going to be tuning the filament
to this specific printer. So whatever values you get from
the test, I’m about to show you, apply to THIS filament, on THIS machine. They don't necessarily apply to that
same filament on a different machine. It might give you a baseline
starting point, but I absolutely recommend testing individual materials
on individual machines. And just because this PLA has
one set of values does not mean that this PLA
from an entirely different manufacturer, and even a different colors can play into
this, is going to want the same thing. When we're tuning in Filaments, of course,
one of the most important things we can worry about
is temperature of extrusion. Most spools have a number
on the side of it, telling you what temperature range
the filament will fall into. But I know say the PolyTerra PLA that I'm printing right now
says 190 to 230 Celsius. That is a huge range! Those ranges on the spools are a ballpark
to get you started, but they are not going to give you the proper extrusion temperature
for your machine, with your filament. You're probably familiar with it, you’ve
probably seen it before, but a temperature tuning tower
is a great way to test out materials and see what is going to work for you. There are a couple of ways
to go about this. You can download an STL and manually
modify the G-code as you go up in height to adjust the temperature you’re
going to be running for the extrusion. Or you can download G-code off of random
sharing sites, maybe for your machine
it's already sliced ready to go. I don't generally recommend
downloading and running G-code, but for temp towers
it's usually fairly tested. Or you can use a calibration test
generator such as in-built into SuperSlicer
or SoftFever fork of Bambu Studio. That's what I'm running here
now, and it's going from 230 Celsius at the bottom to 190 degrees
Celsius at the top. We'll take a look at what results
we get out of this thing when the temp tower finish. We get to take a closer look at it. And that's basically
what we need to do is look at the thing. I just will start looking at all of these and seeing
which one looks the best to me, visually. The bottom one at 230C
I can see some clear part curling on the ramp of it, which tells me that it was a little
too hot and not enough part cooling. A common problem on larger and faster
printing. The other end of the spectrum, the 190C
has the least stringing on the print. It has less part curling, the top surface
looks pretty darn good to me. All around that looks really good. But that little tower inside
a little window is not just for stringing,
it's also for strength and layer adhesion. So I'll take a pair of pliers
and I'm going to push on that little tower and try and break it off. I do this for all of the temperature towers, checking
which one has the best layer adhesion. Now, this is absolutely
just a judgment call. How it felt to me, to me to 220 actually felt like the most effort
to break the tower off. This is a Matte PLA so it doesn't have
the greatest layer adhesion, so it wasn't that difficult
to break any of these out of here. Now, I could use the Teaching Tech
Calibration Website to go generate a new tower, say between 190C and 215C in smaller increments
to really dial in on a good temperature. For this, I really feel like
what I'm seeing here is aligning with what I know of this material
on other machines that I run. So I'm probably going to just do some test
printing between 210 to 215 to try and balance out
strength and stringing and curling on this material with a ballpark idea
of what temperature I want to be running. It's time to move on to the other tests. I'll use this temperature moving forward. I'm going to start by tuning
Linear Advance. I find that Linear Advance between nozzle
sizes can vary pretty dramatically. So I want to start there
and not have my previous 0.4mm nozzle setting affect me moving forward. If you're not familiar with Linear Advance or Pressure Advance, what it does
is actually reduce the amount of extrusion force as it comes up to a corner
or the end of an extrusion line. Think about having a bottle of ketchup
and you're squeezing it onto a plate. As you're
drawing a straight line with that, when you get to the end of that line,
if you do not reduce the amount of squeeze you're having
and you just pick it up and stop, you'll probably end up with a blob
at the end of your line. That's
what Linear Advance is trying to combat. It is tapering off the extrusion force
toward the end of the line that it knows is going to be there so that you don't get
that blob or as you're turning a corner, it has to reduce speed just enough
that that blob could then occur there. To test here, I'm going to use the new
Ellis Pattern Generator instead of lines drawn on your bed
or a tower printed off of it. It's a bit of an object
with a handful of layers where individual arrow shapes are printed to change the values at different point
and see what results you like best. This is honestly pretty easy to set up. I was able to go through here and dial
in how I wanted, add in my bed dimensions, designate what firmware I'm
using on this machine, and I also adjusted the values of the intervals
between the Linear Advance tests. So this pattern does a handful of things. It has a 90 degree corner on it,
so we can look in that corner for a bulge or for the extrusions to narrow
too much as they are on the end of this test pattern. And then at each side
we have an end of the line so we can look for the change in extrusion there as well. There's also a number grid on the side
to designate what the different Linear Advance values were so we can pick which
ones the best and know the value for it. In this particular instance,
it looks to me like 0.054 is where I need to be
to get the best result. The corner could be crisper,
but because I'm using a 0.8 millimeter nozzle, the diameter of that corner
is probably always going to have at least a little bit of a radius to it. So I'm fine with this. This has good consistent extrusion
all through the arrow, the corners, about as crisp
as it is on any of these lines. So this is the one that I'm going to go
with. Savings
value is really pretty straightforward. On some machines you can do it through the menu system
to put it into the firmware and save it. The Neptune
3 Max does not have that option and I really don't like doing it that way
anyway. As I said, this value can differ
from one filament to another, so the better way to do it
is to put it into the Start G-code of a filament profile. Specifically for this filament
in your slicer, I'm using some Polymaker
PolyTerra PLA for this testing, so I'll create a profile in my slicer
with that name. And then in the Start G-code
for the filament, I will put in an M900 space Kand then whatever value
my pressure advance was. In this case, mine is 0.054. That's for Marlin firmware. For Klipper, it's the string of words here
that you enter your K value into. With Linear Advance out of the way
it's pretty downhill from here. Next thing I do is Extrusion Multiplier
a.k.a Flow as it's called in Cura. This is where we are fine tuning the
extrusion rate of individual materials. Getting the mechanical system dialed in
with E-Steps is one thing, but then individual filaments
will flow differently. Some folks skip E-Steps altogether
and purely use Extrusion Multiplier
to offset for mechanical differences. I like to tune both. The way I'm going to do this is prints
and rectangular chips of filament. It's very straightforward to do it
this way. I use the SoftFever fork of Bambu Studio
as my primary slicer as of late, and it has a built in calibration
test to do this. But you can really do this on any slicer by individually
printing a handful of these chips. If you want to do this manually, I'll show
you how to in PrusaSlicer real quick. You don't even need a model to download. Just add a shape
which is a rectangular box. In this case, I'm just going to squish
the Z height of it down because I don't need to print
a big calibration cube to do this. Make sure my top and bottom
layers are set to a value that's going to leave some sparse
infill in the middle of this print. So it's not a solid 100% Infilled object. And then I'll go into my filaments profile
and a set my Extrusion Multiplier. I'll start at one which is 100%(.1),
and then I'll work my way down. All the SoftFever calibration test is
doing is taking the work out of this. It's just creating individual objects for you and setting the values
on each of them separately. Unfortunately, PrusaSlicer cannot do that. I don't believe Cura can either.
Then we go to the printer. We print off
our individual chips of Filament, whether we do it as a full test pattern
like I did, or you can do it one by one if you have to. Once I have these handful of chips
printed out, I can start to look at them closely and look for inconsistencies
or issues in the extrusion. Such as this -9 chip right here. It has gaps between the lines on the top
infill that's telling me this was under extruding
while printing. On the other end of the spectrum,
I have the 0 chip here which is a 100%(.1) multiplier and that has ridges in the top
infill where it looks like the nozzle was digging through filament
a little as it was printing. Which tells me this was over extruded. I'm looking at all these chips looking for the point where this is
extruding not too much and not too little. We're looking for the Goldilocks zone. I personally find that Polymakers
filaments, in general, flow really well. Meaning they need to dial back
the Extrusion Multiplier a bit. In this testing right here
I'm seeing -5, -6, and -7 looking best. -7 I see some little under extrusion
at the ends of the lines, but I usually try to favor
on the lower side of the extrusion. But that's me wanting good looking prints. If I'm more worried about strength,
I might favor a little more on over extrusion. That way
our layers are really squishing together, so that -5 or even maybe -4
might be a better value. This is one of those things
you're going to have to learn to eyeball and figure out what works for you, what feels best, or what works
best in your form of printing. If you're printing models
that you really want to look good, you might want to under extrude
just a little bit so you get nice
cisp corners and less blobbing. If you're more worried
about the functionality of a print versus the looks of it, you might want to over
extrude a little bit. There's absolutely room for interpretation
as to what order you should do these tests in, as each of these variables
can affect one another. But the next one, you should wait
until after you've done your Temperature, Linear Advance and Flow Rate testing,
and that is Retraction! I say that because Retractions are kind of a Band-Aid for things
that the other tests are solving. If your temperature is really dialed
in, you've got a good Linear Advance and your flow rate
is where you want it to be. You may find
you need little to no retraction. With a 0.8 millimeter nozzle
having a pretty large orifice that’s unlikely in this situation,
but it can happen. The test I'm going to use for Retraction
is off of the Teaching Tech Calibration page. You can go in there
and configure your own retraction values, set your temperatures, add Start G-code,
everything you need. This generator worked beautifully, and it's what I would recommend
using for Retraction tuning. This will generate a tower that has different retractions
for different heights. In my case, I'm doing .2, .4, .6, .8, 1 and 1.2 millimeter
because I am running direct drive. As with all the other tests,
it's time to print this thing. The point of the retraction tower
is that the different heights, it will adjust
the amount of retraction used. So I generally like to start with a low
amount of retraction and work my way up so you can see the improvement as things
travel upward and the more retraction. And you can clearly see that
as we went up in .2 intervals from 0.2 to 1.2,
there is a clear improvement here. I expected a little better from this,
but with this machine it looks like
I need 1 or 1.2 millimeters of Retraction. I decided to run this test three times to test out temperatures
combined with retractions. So I did 205, 210, and 215 as my temperatures, and I really didn't
find much of an improvement. So I think I'm going to stick with that
right around 215 to keep my layer adhesion up,
but still get some of the better properties of the slightly cooler
temperature. With those tests out of the way,
it's time for the ULTIMATE TEST! Print testing. Slicing up a design, throwing a print down
and seeing what results we get out of it so we can confirm that all of what we've
done is going in the right direction. Since I'm running a 0.8 millimeter nozzle,
I'm going to do a 200% Benchy and 0.4 millimeter layer height. Let's get this thing printed
so we can look at the results. I came back into the studio
after this finished printing, saw the dramatic lighting
and just had to film it this way. So is this a perfect Benchy? By no means. But for 200% Benchy at a 0.4
millimeter layer height that printed in just over 2 hours,
I think it turned out pretty good. Looking closely at this thing,
the biggest thing that stands out to me is the Retraction and Linear Advance
are still not quite dialed in. And this is exactly why I printed this,
because those tests are great for getting you in the ballpark, but they don't really translate
to the real world all the time. Honestly, looking at this,
I think I could stand to go be a little more aggressive on my Linear
Advance, bump it up a little more and dial back
my Retractions, because I'm getting just the tiniest bit of a blob
with a tiny bit of space after it. I could be wrong though,
and that's something I need to play with as I further tune
this machine in. That's kind of the way it goes
with 3D printer tuning. You're always going to be evolving
and changing what you're doing as you move forward. Our results weren't perfect.
I'm not perfect. 3D printing in general
is kind of an imperfect science. We're spewing out molten plastic,
forming solid objects out of it. There's a lot of variables at play, I guess where I'm going to wrap it up for
this one folks, I've got to get print testing on the Neptune 3 Max so I can get
the review out for this machine. I got some big stuff to lay down! So I'm going to get to that. If you want to check out another video
where I did some printer tuning in it, check out my previous Workshop Vlog
where I worked on my Voon 2.4, FLSun V400, and Project Re-Animaker,
got those printing better! Get Subscribed to ensure your 3D prints
don't fail. It's not a guarantee
but it can't hurt. See you folks!
