The Prusa Core One L is so hot right now - does it handle engineering filaments better than the H2C?

3D printing is in a bit of a graduation phase right now. When it was still very new, we were playing around with everything, not really knowing where the limits were yet. And once we had figured out those limits, well, we knew that PLA and PETG were the two materials that were somewhat easy to print, and anything that was more demanding usually wasn't worth the trade-offs. So for

the longest time, we stuck to those materials. But now that we've got those really well figured out, we're starting to make use of that knowledge for the more demanding materials like extra wear resistant filament, high stiffness materials like fiber reinforced polymers, or the classic temperature resistant materials like ABS, ASA, and polycarbonate. I've been sticking to PETG

for probably more parts than I should have. It is mildly temperature resistant but very soft, so I'm wondering: Are we at a point where we can confidently print those engineering materials yet?

So, these are the contestants that I’ll be benchmarking against each other. In the orange corner, we’ve got the Core One and Core One L - both of these have the closable vent grill in the front and exhaust fans in the back, and the L is, obviously, larger, but it also adds a high-power heated bed that serves double duty as an active chamber heater with a pair of fans

at the bottom. Also, really strong magnets in the bed - in fact, they’re so strong that I can see some people struggling to lift the print sheets - but it does make sense. Even with PETG, the reason prints warp on my XL is usually the print sheet coming up from the magnets.

And in the green corner, we’ve got the P2S and H2S. Similar relation here, the H2S is larger and has active chamber heating, but using a dedicated fan heater. Both the H2S and P2S can choose to draw in fresh air or to just circulate the chamber air through their filters.

On the Prusa machines, the filter unit is an optional addon, but it is built so that any air sucked out of the printer needs to pass the filter first.

Filtration is one thing - and it’s important for PC and ABS, as they have pretty nasty fumes - but air management is only one part of why a specific machine can or can not print a part. The core aspect that we’re fighting here is thermal expansion in one form or another,

part. The core aspect that we’re fighting here is thermal expansion in one form or another, but because we’re printing in layers, it adds up, and kids, that’s where warping comes from. So,

really simplified, let’s say that our polymer, the filament, has two really discrete states - solid and liquid. In reality, it’s not as simple, but hear me out. In either state, the plastic will

and liquid. In reality, it’s not as simple, but hear me out. In either state, the plastic will want to expand when it gets warm and contract when it cools down. So, the printer lays down a bead of hot plastic, onto either the print bed or some area of the printed part, and that fresh bead will immediately start cooling down and will want to start contracting. But because we just welded one

side onto something that is already solid, and our bead itself is still “liquid” and therefore soft, it can’t really do anything, it doesn’t have the rigidity to exert force, so as it cools down, its volume will shrink, but all of that shrinkage has to happen in the Z-axis.

But things change once that bead gets to the point where it turns “solid” and hard and can finally produce some force. Again, this is simplified, unlike metals, plastics don’t have a single temperature where they go from solid to liquid, it’s a continuous process, and even seemingly hard metrics like for example the glass transition temperature,

that only tells you when that process happens the quickest per degree of temperature change.

But let’s stick with the simple model because it does work really well. When

that bead becomes solid, it will of course want to continue shrinking as it cools, and now it is rigid enough to start tugging on things. Anything that happens above that temperature, we can sort of ignore, but with every degree that it cools off below that temperature now adds up layer after layer after layer, and when you’re printing onto a layer that has already shrunk

a certain amount, and you’re now adding a fresh layer that is going to shrink again, you get that characteristic banana-style warping. The biggest weakness of PLA is also its biggest strength, and that’s that its transition from soft and “liquid” to hard and solid

at a very low temperature, so most of its cooling happens while it is still soft.

That makes it super easy to print without any chamber heating or temperature control, but it also means that printed parts are super easy to melt in the sun or in a hot car.

On the other extreme, something like ABS or polycarbonate has that transition point quite high, so, the final part stays usable in higher temperatures, but it also means, that if you print it with the same parameters as PLA and let the layers cool down quite far during the print, much more of the cooling and shrinking is now happening in

a temperature range where the plastic is already solid - so you get much higher forces adding up.

The two common ways to deal with this are either by using a filament that, for the same amount of temperature difference, just doesn’t shrink as much. And this is a huge reason why shredded carbon fiber helps with this - the carbon fiber itself actually has a slightly negative coefficient of thermal expansion, so as it cools, it actually expands,

and because it is also incredibly rigid, that means it can negate quite a bit of the expansion that the base polymer wants to do. The other option is, of course, heated chambers.

So you keep the filament the same, but you change how far you allow it to cool down before the next layer is added, and then once the part is finished, you take it out of the printer, and it cools and shrinks as a unit, but it doesn’t have those internal forces adding up that would cause the part to warp. The ideal situation would be if we could instantly cool the fresh

layer to exactly the temperature where it gets solid, and then perfectly keep it there for the rest of the print. But in reality, because it’s a gradual hardening process, even freshly printed filament can produce some force, and if you keep the already printed layers at too high of a temperature, that can be enough to cause what’s known as curling,

where for example overhangs are too soft and get pulled inwards by the fresh layer.

On the other hand, if you cool the layer too much, now the next layer isn’t going to bond to it that well anymore, and your part is going to start cracking at the layer lines.

Figuring out universally usable printing profiles for high-temp filaments or even for tricky situations with regular filaments is a challenge that is as old as 3D printing itself, and more recently, we’ve seen some better software control that helps with some of the compromises you’d otherwise have to make. For example,

PrusaSlicer has much more fine-grained control specific for how it prints overhangs now, so they can better optimize the profiles to work well everywhere else, and then if for example curling becomes an issue, they can tune in just that one aspect separately.

This is the part that I printed. It’s part of a filter holder box, goes in like this, and I

did four different test prints on each machine, using the same print settings, and the closest filament preset that I could find for the filament type. I used Colorfabb XT, which is a hardcore PETG, then Azurefilm ABS, just a standard ABS, Prusament PC Blend, which is a polycarbonate blend is that is a little more easily printable than the pure stuff,

and finally Polymaker PC Plus, which is a lot closer to plain polycarbonate.

And while these are printing, let’s check out this video’s sponsor, PCBWay! One of the harder lessons that I had to learn and that I’m still struggling with is that, as a maker, you don’t have to do every part of the process yourself for it to still be “your project” or “your thing that you built”.

I know how interesting it seems to have your own CNC router, a pick and place, laser cutter, 3D printer, obviously, but for getting stuff done, you’re just going to get a better final result and also save a bunch of time if you outsource the really specialized parts. And PCBWay does just that! The biggest revelation for me has been ordering populated SMD boards - they’re so much

that! The biggest revelation for me has been ordering populated SMD boards - they’re so much nicer than soldering your own through-hole parts, and trying to get good at placing and reflowing SMD is a whole hobby by itself. PCBWay also offers CNC routing, milling and turning, does 3D printing from standard and specialized materials, and even does sheet metal work, which is a pretty

rare offering. If you value your time at all, all of these end up being cheaper than trying

rare offering. If you value your time at all, all of these end up being cheaper than trying to figure it out yourself, so check out PCBWay as your project partner at the links below - and if you need some inspiration, check out their project sharing section at pcbway.com/project

So, these are done, and you may have noticed the Core One L having some issues there, but we’ll get to those in a second. These parts are all labeled on the bottom, and I want to try and rank them from best to worst, basically blind to which machine they were printed on.

Let's start with the ColorFabb XT, and this one unfortunately has some print issues, so I know which printer it is. These other three, though, look pretty much identical. I think

there's some difference in... yeah. So there's a little bit of a ridge there. This one has nothing, and this one has some ridges visible. So, I would say this one is like the best, and these two are pretty much equal. We should also look at how much these warped, and there it is pretty obvious this

warped more than that, so I think this is going to be my ranking for the ColorFabb XT. The worst

one here is the P2S. It looks like the hotend was struggling. The next one up is the Core One, still warped a little bit, but overall pretty good. Then we've got the H2S and the Core One L at the top.

These two pretty much identical, though the Core One L did warp a little less, actually.

Then, moving up to ABS. And this one and this one, these two have some obvious overfilling at the top here where, you know, you can see that it was warping so much that now the nozzle was dragging through the filament. So these two are definitely—well, one of these is the P2S, the other one is the Core One—because those don't have an actively heated chamber. This one, though,

looks a little worse. It has much more visible warping, and we can see that on this side here as well. This one just... yeah, these two are sprung up quite far. These next

two—so H2S and Core One L—which one is which, I don't know, but honestly, they look pretty much identical here. Maybe this one has a little more warp? Yeah, by the slightest amount.

identical here. Maybe this one has a little more warp? Yeah, by the slightest amount.

Let's check out the ranking. The worst one is the Core One. You can definitely see where the part was sticking to the bed and where it was lifting. There's even two lines on this other side here, and yeah, this has quite the amount of curl. The next one up is the P2S: more even warping, but still pretty crooked. Then we've got the Core One L,

and it is beat by the H2S, by the smallest margin.

Then the Prusament PC Blend, and again, this one is very much overfilled at the top, so this was warping quite a bit. And this one as well, but not as extreme. This one still has visible overfilling, and this one is underfilled at spots. So yeah, none of these are great results. In fact,

these pretty much all warped, except for this last one at the end here. This one is still a little warped, but mostly flat. So, the worst one here is the Core One L. Yeah,

you can see how much of that is the P2S. That is still a part that is not very usable. You

can see how much it actually ended up warping up here, so that's a good 3-4mm. One up from that is the Core One, again with the distinct separation lines where it's stuck and where it lifted. And then at the top, the H2S. So this is the only one that is like reasonably flat,

lifted. And then at the top, the H2S. So this is the only one that is like reasonably flat, but still, at the ends, you can see that it kind of is coming up here and still warping.

And finally, the Polymaker PC Plus. This one, though, is the one from the Core One L that I manually adjusted the settings for, so that's not a fair comparison. The fair comparison would be these—that all of them failed. So, these were three tries with the same settings. None of

them stuck particularly well, were pretty flat, but they came loose from the bed.

For these other three... well, these two are clearly warped. This one is flat, so... well,

I guess this one must be the H2S, and it is. These two, though, I don't know. Both of them are warped pretty heavily. This one has a little more warp than this guy, and that is the P2S, and the Core

pretty heavily. This one has a little more warp than this guy, and that is the P2S, and the Core One actually held on just a little bit better, but again, not really a good, usable print.

So, how did these printers do?

The H2S was really consistent here, it did have a spot where it was going too fast around a corner, but otherwise, yeah, good job. At least for XT and ABS, the Core One L also did a pretty good job - I should mention that I used the satin PEI sheet on the Prusas,

for these materials, that’s what they recommend. Bambu was on standard textured PEI, which they greenlight for pretty much all materials. Between the Core One and the P2S, pretty much all the parts had some noticeable warp to them, which makes sense, without any sort

of active chamber heating, the best you can hope for is around 40, 45°C in the chamber, you know, not nothing, but pretty far off from the 60-ish degrees of the two heated printers.

The Core One did a little better than the P2S, but neither one was great, you’re definitely looking at printing with a brim and sticking to much smaller parts than these if you want them to stay usable with high-temp materials. For some reason, the XT print on the

P2S looked like it was printed way too cold, even though it used the same profiles as on the H2S.

Of course, the one that's standing out here is the Prusa Core One L, and not in a good way.

This was the only printer that I couldn't handle printing polycarbonate at all, and it should not be this printer that has issues here. Of course, the Core One L has a much more powerful bed now,

the first Prusa machine to use a 230-volt heater, and it uses that heater to actively transfer heat from the underside of the bed into the chamber air with two fans underneath the bed, but it doesn't look like that was actually helping. There were two issues that I think were compounding to cause its poor performance here. The first one was these dots on the print bed. From what I can tell,

these are grease drops from the lubrication that the factory applied on the linear rail of the tool head. As the chamber gets hot and that rail gets hot as well, the grease gets soft and liquid,

tool head. As the chamber gets hot and that rail gets hot as well, the grease gets soft and liquid, and it just ends up dropping onto the bed, which obviously is not a thing that ends up improving bed adhesion. It also seems to be going pretty much everywhere. After a couple of these

failed polycarbonate prints, I was noticing that the front door was super smeary and basically coated in a grease film. I tried cleaning the bed, wiping it down with ethanol again, but it did not improve things. The other thing I was suspecting was that the new heated bed construction may not be optimal for actually keeping the bed surface at temperature.

So, to get some comparative numbers, I added some super-matte gaffers tape to the print beds of all of these machines, because thermal cameras need a consistent surface to measure,

and I started the print cycle and let the machines heat up to their set bed and chamber temperatures.

The two Bambu machines were pretty hot, with the H2S landing at exactly the 110°C set temperature, while the P2S was displaying 100°C, but was actually reading 105°C on the surface. The

Core One was displaying 100°C, but actually was sitting around 95°C, 10°C colder than the P2S, and the Core One L was displaying 110°C, but was sitting at 96 to 98°C. Notably,

though, the Core One L had visible cold spots where the two bed fans or chamber heating fans were blowing against the underside of the bed, and the sides of the bed were 5 to 10° hotter.

I don't know if the poor adhesion was just a combination of the grease fouling and the over-reported bed temperature, or whether the bed fans also caused some sort of unfortunate draft that was cooling down the part faster than it would otherwise.

The unfortunate correlation here is that the more the chamber heating needs to work, either because you want to have a higher chamber temperature or because a colder ambient temperature means the printer is losing more heat through its shell, the more energy the fans will need to extract out of the heated bed. So, with the hybrid bed heater,

you trade chamber temperature against consistent bed temperature. And of course, the Core One L's side panels are now made of aluminum instead of steel, as on the original Core One - so the Core One L ends up losing quite a bit more heat through that more thermally conductive shell.

But knowing that these cold spots exist, I retried the print in the center of the bed, with the ends oriented towards the hottest part of the bed, which are in the center, to the left and right. I also manually adjusted the bed and chamber temperatures to their maximum, which is 120 degrees and 60 degrees, respectively. The bed had no problem at all getting to temp, but

the chamber really slowed down in heating past 56, 57°. My basement studio was at a controlled 18° for all these tests, but if you had your room temperature at 22 degrees or more, which is quite cozy warm, it should make it to 60°. But when the bed is set to anything lower than its maximum 120

degrees, the maximum temperature that the chamber can reach and maintain also drops proportionally.

Maybe, you know, wrap the Core One L up in a blanket, but do keep the electronics vents open.

Either way, this print worked and it compares pretty well.

Of course, that brings us to the age-old discussion of whether print profiles that give you predictable results should be considered a part of the product, and my stance for the longest time now has been that yes, they definitely are part of what you are buying and paying

for in a printer. Cranking up the temperatures like I did is a rather small correction; it's certainly one that is very dependent on the exact filament you choose, but considering that all the other printers did a much better job printing polycarbonate, with their generic, stock settings, and even using technically inferior setups, I just have to wonder why the Core One L didn't

just add a dedicated chamber heater. These are not that expensive; For reference, you can get a full, little, plug-in PTC heater with a fan and temperature regulation for less than 10 bucks.

And I don't see how these would be difficult to integrate into the printer. Everyone else is doing it, and the Core One L already has extra AC safety cutoffs that the heater could share with the bed.

This whole situation reminds me of the lil mini heater beds that the Prusa XL uses, where I was already questioning how it was at all viable to use 16 individually controlled heated beds, and Prusa was like, "No, no, this isn’t making the printer more expensive", but then the Core One L, in fact, did not use heater tiles in for example a smaller 3x3 grid,

but went with a standard silicone heater mat, and the Prusa XL ended up being the only printer where they ever made use of their heater tiles. Now, the Core One L is the first printer from Prusa where I think you notice that they were definitely optimizing things for cost, but still, these fans still feel more like an afterthought than anything.

To get back on topic of which of these printers can actually handle high-temperature materials: they all did a surprisingly good job, in my opinion. Compared to the benchmark of having open-frame machines that are just running in ambient air, these all did a fantastic job printing materials like ABS and polycarbonate.

I didn't print these same test prints on something like a MK4S, but speaking from experience, I don't think any open-frame machine would handle printing this part in any of these materials.

Layer adhesion on ABS has always been a problem, and here it is, too, but with a slightly higher

nozzle temperature, this might be salvageable on any of these machines. I think both the jump from the P2S to the H2S and the Core One to the Core One L, where the larger machine adds some sort

of active chamber heating, makes sense because the small machines printing small parts do an okay job here, but larger parts also like to warp more, so you need to pay even more attention to parameters like controlling the chamber temperature. And I really do like seeing that even pure polycarbonate

is totally printable now - this stuff is pretty cool, of course, it does high-temp, but it’s also super stiff, so mechanical stuff, you know, robot arm joints with motor mounts,

this might be a really good option. All the links below, as always, thanks for watching, massive thanks to everyone who already is subscribed or even supporting the channel on YouTube memberships or Patreon, and to all of you - keep on making, and I’ll see you in the next one.