My Thoughts on the Bambu Lab H3C After Using it for a Real Project — Transcript
Full transcript
- 0:00For the last 5 years, I've been working
- 0:01on a simple idea, an electronics
- 0:04enclosure with a numeric display and
- 0:06flexible buttons 3D printed using
- 0:08multiple materials in one shot in a
- 0:10single print job. In 2021, I bought a
- 0:14Prusa XL to try it, but after working
- 0:17with Prusa support for almost a year
- 0:18without success, I finally just gave up
- 0:20and sold it. I was hopeful when Bambu
- 0:23Lab released the H2C, but it couldn't do
- 0:25what I wanted either. Snapmaker even
- 0:28sent me a U1, and that machine turned
- 0:30out to be a disappointment that we'll
- 0:32talk about on another day, but this
- 0:34machine
- 0:35this machine finally pulled it off.
- 0:38>> [music]
- 0:41>> Welcome back to Cloud 42. I'm James.
- 0:44This is the H3C from Bambu Lab. This is
- 0:47finally the printer that I, and to be
- 0:49fair, a lot of other people have been
- 0:51asking them to make for a long time. For
- 0:55several years now, I've been asking them
- 0:56when they'll have a tool changer, and
- 0:58the answer has always been the same. Not
- 1:01this year. Well,
- 1:02>> [snorts]
- 1:03>> it's finally here, and not to bury the
- 1:05lead, it is freaking awesome. This is
- 1:08finally the machine that unlocks all of
- 1:10the crazy things that I personally want
- 1:12to do with a 3D printer.
- 1:18When I built my electronic lead screw, I
- 1:20used an off-the-shelf LED display with
- 1:22tactile buttons, and I machined and 3D
- 1:25printed an enclosure around it. But, in
- 1:28the back of my mind, I was thinking why
- 1:30couldn't the entire thing, including the
- 1:32display and the buttons, be 3D printed?
- 1:36Now, the obvious answer is because
- 1:38mixing all of the required materials in
- 1:40a single 3D print is hard. To do what I
- 1:43envisioned, I would need nine different
- 1:44materials. A rigid material for the
- 1:46frame, three different transparent
- 1:49materials for the displays, two flexible
- 1:51materials for the buttons, a conductive
- 1:53flexible material for the key contacts,
- 1:56and two different types of support
- 1:58material. One for the rigid, and one for
- 2:00the flex. Now, I did have some success
- 2:03printing just the displays on the A1
- 2:06Mini back in the day, but this machine
- 2:08can handle all nine materials in a
- 2:10single job. And
- 2:13here's the result. This calculator is
- 2:15just a proof of concept, but it was a
- 2:17good test bed to work through the
- 2:19challenges of mixing all of these
- 2:21different materials together. The
- 2:23printer can handle them all just fine,
- 2:24but physics is physics and the materials
- 2:27themselves have to coexist. They have to
- 2:29work in the same thermal print
- 2:31environment. They have to stick to the
- 2:32bed and each other when I want them to
- 2:34and release when I want them to. And
- 2:36there is definitely a learning curve to
- 2:39doing that. I'll do a full video on this
- 2:41project, but for now, let's open it up
- 2:43and I will show you the highlights.
- 2:47The black frame is printed in Bambu Lab
- 2:49PLA carbon fiber, and that also forms
- 2:53the partitions that keep the light from
- 2:55the individual display LEDs from
- 2:56spilling into neighboring segments.
- 2:59There's a single layer of white PLA
- 3:01basic down at the bottom of each light
- 3:03channel to act as a diffuser. Behind
- 3:06that, there are four layers of Polymaker
- 3:08Pan Chroma transparent red PLA to color
- 3:11the light. And over the entire front of
- 3:13the display, there are two layers of
- 3:15Toxen 3D transparent black PLA to hide
- 3:18the segments when they're not lit and
- 3:20increase the contrast. I also had a
- 3:23version of the display that used PETG,
- 3:25but I liked the red color of the Pan
- 3:27Chroma PLA better, so that's what I
- 3:30ended up using. The keys are printed in
- 3:32Bambu Lab TPU 75A.
- 3:35That's a new ultra-soft TPU that's being
- 3:37introduced with this printer, and we'll
- 3:39talk more about that in a little bit.
- 3:42The embedded labels on the keys are
- 3:43printed in Bambu Lab TPU 85A. Each key
- 3:47is supported on a thin flexure that
- 3:49gives it a tactile feel. That flexure is
- 3:52a single wall of soft TPU 75A, 0.3 mm
- 3:56thick, printed unsupported at a 45°
- 3:59overhang. Now, let that sink in for a
- 4:02minute. 75 ATPU, 0.3 mm wall thickness,
- 4:0745° overhang. It is just crazy that you
- 4:11can do that on an FDM printer.
- 4:13On the back of each key is a conductive
- 4:15pad printed in Recreus conductive
- 4:18Fillaflex. That's a 92 ATPU with
- 4:21graphene in it to make it conductive so
- 4:23that it can make contact with the
- 4:25printed circuit board.
- 4:27The conductive pads are printed on top
- 4:28of little flexible posts so they can
- 4:30flatten evenly against the circuit board
- 4:32and I used Bambu Lab matte PLA as
- 4:36support material because it separates
- 4:37easily from the TPU. And to allow the
- 4:40buttons to protrude out the front, I
- 4:42supported the entire front surface of
- 4:44the enclosure with Bambu Lab support for
- 4:47PLA. Now, you're seeing clips with a
- 4:49bunch of different filament colors
- 4:51because I printed several prototypes
- 4:53while I was iterating on the design, but
- 4:56in the end, it all came together and it
- 4:59works. This machine is a game-changer.
- 5:02It is going to open up all kinds of
- 5:04capabilities that you will definitely be
- 5:06seeing in future videos.
- 5:08Let's take a closer look at it and I'll
- 5:10share my thoughts now that I've used it
- 5:12for a few weeks.
- 5:17But before we do that, you should know
- 5:18that I didn't pay for the printer. Bambu
- 5:21Lab sent it to me for testing and
- 5:23feedback. They're not paying me anything
- 5:25to make the video. They have no control
- 5:27over what I decide to show or what I say
- 5:30and they are seeing this the same time
- 5:32that you are. Now that said, there is a
- 5:35link down in the video description and
- 5:37if you use it to make a purchase,
- 5:39I earn a commission that helps support
- 5:40the channel. If that's something that
- 5:42you want to do, I appreciate your
- 5:44support, but if not, that is cool, too.
- 5:46I still hope you find this helpful.
- 5:52So what exactly is this thing? This is
- 5:55the H3C,
- 5:56the latest in the large H series of
- 5:58printers from Bambu Lab. It has all of
- 6:01the specs that you've come to expect in
- 6:03these machines, A fully enclosed 330 by
- 6:06320 by 325 mm build volume, 350°
- 6:11hardened steel nozzles, 120° heated bed,
- 6:1465° actively heated chamber, and HEPA
- 6:17and activated carbon air filtration.
- 6:20What's new is the Swap-Tech tool
- 6:23changer. The machine has a rack of nine
- 6:25tools, each with its own independent
- 6:27filament feed path, heater, and
- 6:29temperature sensor. This allows you to
- 6:31print with multiple filaments in any you
- 6:34like. You load them once and they stay
- 6:36loaded, each in their own dedicated
- 6:38tool, and the print head picks them up
- 6:40as needed during the job.
- 6:42Also new in the H3C is support for 2.85
- 6:46mm filament.
- 6:48I know, I thought it was weird, too, but
- 6:51the larger filament diameter makes it
- 6:53easier to handle soft materials. They
- 6:56have a new line of 75A TPU in that
- 6:59larger diameter filament, and it prints
- 7:01at 10 cubic millimeters per second,
- 7:04which is just
- 7:05crazy. I will show some prints a little
- 7:08later and I think you'll be impressed. I
- 7:10certainly was.
- 7:12You can't, of course, mix different
- 7:13filament diameters in the same job.
- 7:16The machine has a nine-bay integrated
- 7:18AMS built into the top of the enclosure
- 7:21with one bay feeding each tool. There
- 7:24are ports to bypass the tubes for
- 7:26printing flexibles, and there are ports
- 7:28in the back of the machine to print from
- 7:30external spools. The AMS also supports
- 7:33drying up to 85° C, but only in two of
- 7:37the nine bays. Not two at a time,
- 7:39there's just two bays that have the
- 7:41drying functionality.
- 7:43This is also the first Bambu Lab printer
- 7:45to support mixing nozzle sizes. You can
- 7:48finally use a large nozzle to lay
- 7:50material down quickly in a print and
- 7:52then switch to a different tool with a
- 7:54smaller nozzle to embed fine details.
- 7:57I'll show some examples of that later,
- 7:59as well. The printer supports
- 8:00networking, but it is Wi-Fi only. It
- 8:04doesn't have an Ethernet port. Now, as
- 8:06of today, this printer is selling for
- 8:08$24.99 in the US equipped exactly as you
- 8:12see it here with nine tools and the
- 8:14integrated nine-bay AMS.
- 8:21There is no getting around the fact that
- 8:23this machine is big. The footprint is
- 8:26about the same width as the previous H2
- 8:28series printers, but it's about 60 mm
- 8:31deeper to accommodate the tool rack in
- 8:34the front and it is
- 8:36tall. Compared to the other H series
- 8:38printers with an AMS stacked on top, the
- 8:40difference is only a few inches, but
- 8:42because of the size and depth of the
- 8:44AMS, most people are going to struggle
- 8:47to reach the back row of spools with it
- 8:49sitting on a normal height workbench. I
- 8:52have it on an adjustable bench here
- 8:54lowered down quite a bit and I find it
- 8:56pretty comfortable to access at this
- 8:58height, but I'm 6-ft or 183 cm tall, so
- 9:02not everybody's going to have that same
- 9:03experience. I think Bambu Lab is going
- 9:06to offer a low storage stand that brings
- 9:08it to a more comfortable height, but I
- 9:10haven't seen it yet and I'm perfectly
- 9:13comfortable using it like this.
- 9:15I do think the decision to put the AMS
- 9:17on top was a good one. I really like
- 9:20that all the spools have a place to
- 9:22live. They're in a sealed environment,
- 9:23so they don't pick up moisture from the
- 9:25air as quickly and they're not hanging
- 9:27off the sides of the machine, so you
- 9:29don't need to leave extra space beside
- 9:32the printer.
- 9:33Now, the rack does block visibility from
- 9:36the front, which is a little bit
- 9:38frustrating, but the rack does move up
- 9:40and down during a print and it allows
- 9:42you to get a decent view of the nozzle
- 9:45laying down the first layer from the
- 9:47front looking underneath the rack. I
- 9:49find it much easier though to look
- 9:51through the window on the front of the
- 9:53AMS. With the printer on a low surface
- 9:55like I have it here, that window is just
- 9:57about at the right height for me just to
- 9:59walk up and get a good comfortable view
- 10:02of the print. If you're shorter, that
- 10:04might be more of a struggle and if you
- 10:06wanted to keep a small step stool by the
- 10:08machine, I would not judge you for that.
- 10:11The other thing that I noticed right
- 10:13away is that they changed the little
- 10:15musical tones that play at the end of
- 10:17the print job. They lowered the last
- 10:19note a half step and it really bothers
- 10:22me. Just listen to this.
- 10:27I want the end of my print to sound
- 10:28happy and that does not sound happy to
- 10:31me. Bambu Lab, if you are watching this,
- 10:34please put the major third interval back
- 10:35so that it sounds happy. Is this just
- 10:37me?
- 10:39Maybe it's just me.
- 10:44Bambu Lab calls their tool changer
- 10:46system SwapTech. It consists of a rack
- 10:49of nine tools at the front of the
- 10:50printer and a head that picks them up
- 10:52one at a time during the print. The
- 10:55extruder motor and feed rollers are in
- 10:57the head, so the individual tools just
- 11:00have the nozzles, heaters, sensors, and
- 11:02a status LED.
- 11:04Unlike the Prusa Index system, Bambu Lab
- 11:07decided to include a hardwired
- 11:09electrical connection to each tool
- 11:11alongside the feed tube so that they can
- 11:13use a contact temperature sensor and
- 11:16they can preheat the nozzle before
- 11:17picking it up.
- 11:19The tools themselves are specific to the
- 11:21filament diameter. My printer came with
- 11:24eight 1.75 mm tools and one 2.85 mm. And
- 11:30I think that's what the production units
- 11:32will have, but you can always buy more
- 11:34tools and install whatever combination
- 11:36you need. The head itself has a
- 11:39motorized cam system so that it can grip
- 11:41either filament diameter without
- 11:43changing any parts.
- 11:45Swapping out a tool is just a matter of
- 11:47one electrical connector and one PTFE
- 11:50push-to-connect fitting, but there are a
- 11:52lot of tubes and wires and springs in
- 11:55the way and it is a little bit fiddly.
- 11:57You can lift the AMS off the top of the
- 11:59machine to get easier easier access to
- 12:01all the tubes and wires and I highly
- 12:04recommend doing it that way. Smaller
- 12:06fingers would also be helpful. If you're
- 12:09a small person, you might need a step
- 12:11stool to reach the tools, but you will
- 12:13have an easier time getting your fingers
- 12:15in to actually change them. The hot ends
- 12:18are attached with three screws and can
- 12:20be swapped separately. You just take out
- 12:22the screws, unplug the hot end, and plug
- 12:24in a new one to switch nozzle diameters.
- 12:27Now, in theory, Bambu Lab will also
- 12:30offer a blade cutter and other tools for
- 12:32the Swap Tech system, but I haven't seen
- 12:35any of those yet.
- 12:40The new AMS 3H is integrated directly
- 12:43into the top of the printer. It comes in
- 12:45a separate box, but it mounts on the
- 12:47printer with four screws. It's a fully
- 12:50sealed system with nine bays and a
- 12:52cam-driven feed roller to pull the
- 12:55filament down to the tool when loading,
- 12:58but it doesn't actually have a spool
- 12:59drive, so when unloading a filament, you
- 13:02have to physically pull the filament out
- 13:04yourself. It doesn't rewind on its own.
- 13:06Now, that's not a big deal, of course,
- 13:08because it's a tool changer and it
- 13:10doesn't have to unload and reload during
- 13:12a job. It does mean, though, that you'll
- 13:15have to manually unload PLA before you
- 13:18use the heated chamber. This wasn't an
- 13:21issue with previous printers because the
- 13:22AMS pulled the filament back at the end
- 13:25of the print, but because this is a tool
- 13:27changer and the filament stays loaded,
- 13:29PLA left in the tool will be exposed to
- 13:32the elevated chamber temperatures when
- 13:34printing something like nylon, and it
- 13:36will soften and cause issues if you
- 13:38don't unload it.
- 13:40Now, most of the bays in the AMS have a
- 13:42little bypass port that opens in the
- 13:44bottom of the bay so that you can bypass
- 13:47the tube for printing flexibles. There
- 13:50are little loops on the cable clips to
- 13:52guide the filament so that it stays with
- 13:53the tool, and it reduces the feed
- 13:56resistance, which would otherwise be a
- 13:57big problem for very soft materials.
- 14:01Now, bays six and eight support drying
- 14:03up to 85° C.
- 14:06There's a little chamber within a
- 14:08chamber to contain the heat in just
- 14:10those two bays.
- 14:12The electrical and thermal challenges
- 14:13become a much bigger deal with nine
- 14:15bays, so I understand why they limited
- 14:18drying to just two, but
- 14:20it really is pretty inconvenient,
- 14:21especially since you have to manually
- 14:23unload the filament to move spools
- 14:25between the bays. I find myself doing
- 14:28most of my drying outside of the printer
- 14:30because of that.
- 14:36They sent me a few rolls of the new 75A
- 14:38TPU in the larger 2.85 mm diameter, and
- 14:43it might be my new favorite filament.
- 14:46Soft filament is hard to manage in a 3D
- 14:48printer, but the larger 2.85 mm diameter
- 14:52makes it considerably
- 14:54easier. This stuff prints fast and
- 14:57clean. I got clean prints using the
- 15:00default settings with absolutely zero
- 15:02stringing, and it prints at a crazy 10
- 15:05cubic millimeters per second, which is
- 15:06just insane. That's more than three
- 15:09times faster than their harder 85A TPU.
- 15:14The larger filament diameter really
- 15:15makes that much of a difference.
- 15:18In my testing, I had good results using
- 15:20ordinary PLA basic for support material
- 15:23for the TPU 75A. It sticks well, and it
- 15:27comes off cleanly if you peel it. Now,
- 15:30believing a 0.1 mm gap makes it come off
- 15:33easier, but at the expense of a worse
- 15:35finish on the supported surfaces.
- 15:38If you need to mix soft TPU with
- 15:40materials like PETG that require a
- 15:42hotter bed, I would recommend using
- 15:45liquid glue on the bed sheet as a
- 15:47separation layer to keep the TPU from
- 15:50fusing to the hot sheet permanently.
- 15:53With the glue layer, it will peel off
- 15:55cleanly, and you can always add a few
- 15:56drops of isopropyl alcohol to release it
- 15:59if it doesn't come off on its own. And
- 16:01don't forget to apply glue for the prime
- 16:03tower, too. The thin initial layers of
- 16:06the tower can be pretty hard to get off
- 16:08the bed.
- 16:13Let's take a look at some test prints.
- 16:15This is a smoke collection hood for my
- 16:18fiber laser. It's printed in bamboo lab
- 16:22PLA basic with just the default settings
- 16:24and it's pretty clean. PLA basic tends
- 16:28to show every little imperfection in
- 16:30layer alignment and flow calibration and
- 16:34this looks credible. No notes.
- 16:37Jumping to my previous favorite filament
- 16:40ABS glass fiber, this is the drill press
- 16:43control housing that I've shown in
- 16:45pretty much every 3D printer review I've
- 16:47ever done. It's printed in black and
- 16:50blue ABS glass fiber with bamboo lab
- 16:53support for ABS as a breakaway support
- 16:57interface. The printer did a beautiful
- 16:59job with this material. All their
- 17:01printers do a beautiful job with this
- 17:02material. The walls are clean, the tool
- 17:05alignment looks good and the support
- 17:07material peeled away cleanly.
- 17:10This is the control panel for my surface
- 17:12grinder, also printed in ABS glass
- 17:14fiber. This is a large part and it
- 17:17really requires a heated chamber to stay
- 17:19flat and the H3C did a great job with
- 17:23it. Now in the past, I have always made
- 17:25the markings on the front with a fiber
- 17:28laser after printing, but since the H3C
- 17:31supports mixed nozzle sizes, I used a
- 17:340.4 mm nozzle for the body and a 0.2 mm
- 17:38nozzle to print the text and the
- 17:40markings using white ABS. And as you can
- 17:43see, it's it's it's gorgeous. I would
- 17:47not have thought it would be possible to
- 17:49print this clearly at this size, but
- 17:52you're you're looking at it. I will
- 17:54definitely be skipping the laser in the
- 17:56future if I can get results that look
- 17:58like this.
- 18:00I also used the multiple nozzle
- 18:02diameters to print these labels for
- 18:04terminal blocks in an industrial control
- 18:07cabinet. These are white PLA printed
- 18:09with a 0.4 mm nozzle and black text
- 18:13printed with a 0.2.
- 18:15These were printed standing up on the
- 18:16bed and while they turned out better
- 18:18than they would have with a larger
- 18:20nozzle. I think the benefit isn't as
- 18:22pronounced in this orientation. When you
- 18:25mix nozzle sizes, the slicer also mixes
- 18:28layer thicknesses, and while it works
- 18:31well, I think the benefits of the
- 18:33smaller nozzles are more apparent with
- 18:36details printed flat on the bed or on
- 18:38top surfaces of the print.
- 18:41I did try the new color mixing feature
- 18:43in the slicer. This part was printed in
- 18:46CMYK W with cyan, and magenta, yellow,
- 18:50black, and white PLA. All of the other
- 18:53colors that you see here are the result
- 18:55of mixing those filaments in the print.
- 18:58I also tried to print a gradient from
- 18:59black to gray across the width of the
- 19:01wheel, and it's
- 19:03okay, not great. This isn't really
- 19:06something that I see myself using much,
- 19:08but I thought I'd give it a try. I'm
- 19:10sure other people with more experience
- 19:12and taste could get better results.
- 19:16Let's jump up to the new 75 A TPU. This
- 19:19is a set of flexible grommets for a
- 19:21cable entry system on an industrial
- 19:24control cabinet. This is the very first
- 19:26thing I attempted to print in this
- 19:28material, and it came out essentially
- 19:31perfect on the first try. I expected to
- 19:33have problems with stringing in a
- 19:35material this soft, but it's just clean.
- 19:39This stuff is way softer than any TPU
- 19:42I've ever used, but it prints with less
- 19:44drama, and that opens up all kinds of
- 19:46possibilities. My only real complaint
- 19:49about it is that they didn't send me any
- 19:51black.
- 19:52I also printed a full bed of button
- 19:54inserts for my ELS project. I normally
- 19:57print these in something much harder,
- 19:59like Ninjatek Cheetah, which is around
- 20:0295 A, and 50 sets of inserts usually
- 20:06take 10 to 12 hours.
- 20:09This job finished in just over 5 hours,
- 20:11and they turned out great.
- 20:13This job is really the worst case for
- 20:16stringing because starting at layer 15,
- 20:19there are literally 350 tiny islands in
- 20:22every layer of the print. There's nearly
- 20:25always at least a few little hairs
- 20:27between the buttons, but I don't see any
- 20:30in this print. This is really good.
- 20:33And here of course is the calculator I
- 20:35showed earlier. This was a challenging
- 20:37print just because it mixes so many
- 20:40different filaments and it has a bunch
- 20:42of weird requirements. I need the
- 20:44buttons to flex, but I want them to have
- 20:46a very specific tactile feel. I'm mixing
- 20:50different colors of soft TPU for the
- 20:52keycaps and I need the markings to be
- 20:54clean. I need to control the optical
- 20:56properties of the smoked lens over the
- 20:58front and the red lenses inside the
- 21:01display segments and I need the
- 21:03translucent white diffusion layers to be
- 21:05smooth and consistent. I think the real
- 21:08standout feature here though is the
- 21:10super thin walls of the button flexures.
- 21:13They're 0.3 mm thick built up in free
- 21:16space at a 45° angle in 75 A TPU. I know
- 21:19I already mentioned that. I did have to
- 21:22fiddle with the settings a little bit to
- 21:23get them perfect, but the end result is
- 21:26pretty much just that. Perfect. There
- 21:29are no lumps or holes and they flex
- 21:31exactly the way I wanted them to. If I
- 21:35had to quibble with anything here, it
- 21:37would be the Recreus conductive
- 21:39Filaflex. It's a little bit underfilled,
- 21:41but I mean I just used a generic TPU
- 21:44profile and I did zero tuning aside from
- 21:46setting the recommended temperature. So,
- 21:49that's kind of on me.
- 21:51The last test print I have here is the
- 21:53enclosure for the power testing fixture
- 21:55I showed in my previous video. This is
- 21:59printed in Bambu Lab PAHT carbon fiber
- 22:02and while most of the print is
- 22:04exceptional as usual, there is something
- 22:07going on with the first layer. It felt
- 22:09just a little bit crowded to me just
- 22:11watching the print and you can see some
- 22:13wrinkles in the surface finish as a
- 22:15result. Now, no other Bambu Lab printer
- 22:18I've ever used has had any trouble with
- 22:20this material and no other material I've
- 22:23used in this printer has had any
- 22:25problems. ABS glass fiber lays down
- 22:27perfectly.
- 22:28I reported this to Bambu Lab and they're
- 22:31working on it. The latest beta version
- 22:33of the slicer is a little bit better and
- 22:35I expect they'll fix it completely at or
- 22:37soon after release, but this is kind of
- 22:40a rare miss for them.
- 22:46This is a big printer with a big bed and
- 22:48a fast, powerful heater and it can pull
- 22:51a lot of power. The spec sheet says that
- 22:54it can pull a peak of 1,800 W, which is
- 22:58going to pretty much max out a 15 amp
- 23:00110 V circuit here in the US. It only
- 23:03draws that for a few minutes during warm
- 23:05up and then it settles down to 200 to
- 23:07500 W depending on the print
- 23:10temperatures. The printer does have a
- 23:12low power mode that caps that at 1,050 W
- 23:16for better compatibility with
- 23:18residential outlets.
- 23:20I took a look with my power analysis
- 23:21gear and can confirm that those numbers
- 23:24are accurate, but there are also some
- 23:26neat things that they're doing with the
- 23:28power modulation that I wanted to show
- 23:30you.
- 23:31This is the power waveform while heating
- 23:32both the bed and the chamber to high
- 23:35temperature for printing nylon. I never
- 23:38saw the average power exceed about 1.3
- 23:40kW,
- 23:41so I assume the 1.8 kW they quote is a
- 23:45worst-case number that includes the
- 23:47filament dryer. But what's really
- 23:49interesting is what happens when the
- 23:51printer reaches temperature and starts
- 23:53modulating the power to the heaters.
- 23:56Previous printers used phase angle
- 23:58modulation, meaning that they switch the
- 24:00power on to the heaters partway through
- 24:03the AC cycle, putting a nice square edge
- 24:06on the waveform. But the H3C is using a
- 24:09zero-crossing switching algorithm that
- 24:11turns the power on and off for entire
- 24:14half cycles of the AC waveform,
- 24:17eliminating those sharp edges. This
- 24:20keeps the power factor high, but it also
- 24:22dramatically reduces current harmonics.
- 24:25In my testing, this machine comes in way
- 24:28under the IEC standard limits.
- 24:30But, unlike the early versions of the
- 24:33H2D before they switched to phase angle
- 24:35modulation, it [snorts] isn't using
- 24:37burst fire. It's not running multiple
- 24:39cycles together. It's switching the
- 24:41heaters on and off for individual half
- 24:44waveforms. And it's dithering those half
- 24:46waveforms across the positive and
- 24:48negative halves of the AC cycle. And
- 24:51that is going to reduce the tendency for
- 24:54the heater switching to cause LED lights
- 24:56to flicker and should help balance
- 24:59neutral currents in a multi-printer
- 25:00environment. These are all well-known
- 25:03power modulation techniques, but I have
- 25:05never seen them implemented this well
- 25:08in a consumer device before. This is
- 25:10actually pretty sophisticated.
- 25:12With a high-temperature nylon job, the
- 25:14average power settled down to around 450
- 25:18W steady state during the print. And
- 25:21with a PLA job, it settled down to
- 25:22around 250 W. So, that major power spike
- 25:26is really just during the initial
- 25:28warm-up. And when the printer isn't
- 25:30printing anything, it goes into sleep
- 25:32mode and turns off the LED lights unless
- 25:34you have them configured to stay on. And
- 25:36the total power consumption drops below
- 25:38about 15 W. I
- 25:40I also tested the printer in the
- 25:42low-power mode. The spec sheet says that
- 25:45it should limit peak power to 1,050 W.
- 25:48And I measured it around 1,030.
- 25:51It appears to be accomplishing this by
- 25:53heating the bed first and then only
- 25:55turning on the chamber heater after the
- 25:57bed reaches steady state and starts
- 25:59modulating. After that, the two heaters
- 26:02are dithered together and settle down to
- 26:04the same 200 to 500 W during the print.
- 26:08Low-power mode does slow down the
- 26:10warm-up cycle, but not by very much. In
- 26:12my test, it only added a couple of
- 26:14minutes. Now, I have 20 amp circuits for
- 26:17my printers, but if you just have
- 26:18typical 15 amp 120 V residential
- 26:22circuits in the US, you might just want
- 26:24to leave low-power mode turned on all
- 26:26the time.
- 26:31The H3C is a part of the standard Bambu
- 26:34Lab software ecosystem with Bambu Studio
- 26:37and the Bambu Handy mobile app. The
- 26:40slicer Bambu Studio synchronizes with
- 26:42the printer so that you don't have to
- 26:44configure which filaments, filament
- 26:46diameters, and nozzle sizes are
- 26:48installed and loaded. It just knows, and
- 26:51you can just select the options you want
- 26:53to use in a print. If you've been in the
- 26:55Bambu Lab ecosystem for a while, you
- 26:57probably just take this for granted, but
- 26:59not all printers have that. And once you
- 27:02get to the level of complexity a true
- 27:04tool changer like this offers,
- 27:07that's a real differentiator.
- 27:09Now, slicing isn't what it used to be.
- 27:12The overall complexity is getting
- 27:14really, really high. For this printer,
- 27:17the slicer has to manage nine tools,
- 27:20nine loaded filaments, multiple filament
- 27:22diameters, multiple nozzle diameters,
- 27:25variable layer heights, color mixing,
- 27:27multiple wall generators, and a hundred
- 27:29other minor features that all interact
- 27:32with one another. The test matrix is
- 27:34insane.
- 27:36As a software developer myself, I
- 27:38appreciate the complexity of the
- 27:39challenge, but I also have to say that
- 27:42the pre-release software that they
- 27:43provided with the printer for those of
- 27:45us who were testing it was a little more
- 27:47wobbly than it has been in previous
- 27:49releases. I found several bugs and
- 27:52reported them, and to their credit,
- 27:54they've already fixed many of them, and
- 27:56more fixes are promised in the final
- 27:58release. I'll talk more about the
- 28:00specific issues in a little bit.
- 28:07This is a cloud-connected printer, and
- 28:09it turns out some people have feelings
- 28:11about that. Now, for me and my usage of
- 28:14the machine,
- 28:15I love it. I love having my settings
- 28:17synchronized across multiple computers.
- 28:19I love being able to start and monitor
- 28:21prints from anywhere, and I love the
- 28:23fact that the ecosystem is constantly
- 28:26improving. Every Bambu Lab printer I've
- 28:29owned has gotten better over time as
- 28:32they incorporate feedback and improve
- 28:34the software and the firmware. I know
- 28:37that this also means that they sometimes
- 28:39change things that upset people and
- 28:41there have been some recent examples,
- 28:44but it also means that the ecosystem can
- 28:46respond to emerging security threats.
- 28:49And boy, have there been a lot of
- 28:52emerging security threats lately.
- 28:55That trend is not going to change. It is
- 28:57only going to accelerate. I know that
- 29:00there are people who advocate never
- 29:01updating your software or firmware and
- 29:03maybe you'll get away with that if
- 29:05you're not connected to a network
- 29:07at all, any network, even a private
- 29:10network at your house,
- 29:12but if you're connected to a network,
- 29:13that is terrible advice in today's
- 29:16world. Just understand that this is a
- 29:18connected device and if you want to live
- 29:20in a disconnected world where nothing
- 29:22ever changes, this might not be the best
- 29:25fit for you. Though, if you're an
- 29:27enterprise user and you need to run in
- 29:28an isolated network environment, check
- 29:30out Bambu Lab Fleet Hub.
- 29:36Let's go through some of the issues I
- 29:37encountered and some solutions that
- 29:39might help you if you run into the same
- 29:41things. The very first thing that I ran
- 29:43into is that I could not find the screws
- 29:46to mount the AMS on the top of the
- 29:48printer. Turns out they're in a little
- 29:49plastic bag taped to the package of
- 29:52desiccant that you're also supposed to
- 29:54put in the AMS, but when I was unpacking
- 29:56it and I saw that, I thought it was just
- 29:58a desiccant pack for shipping and I
- 30:00threw it away. Don't do that. The screws
- 30:03you need are in there.
- 30:05I already mentioned the issues I had
- 30:07with the first layer of PAHT-CF
- 30:10being a little bit too crowded into the
- 30:12bed. I reported that and they sent me
- 30:15some start G-code that made it a little
- 30:17bit better. Now, other high temperature
- 30:19filaments don't have this problem, so
- 30:21it's probably just a tuning issue that
- 30:23they will eventually work out. In the
- 30:26pre-release version of Bambu Studio that
- 30:28I've been using for my testing, the
- 30:30slicer has a bug when mixing .4 mm and
- 30:34.2 mm nozzles and using dedicated
- 30:37support material. If you have the slicer
- 30:40in multi-diameter mode, it won't
- 30:43actually allow you to set the support
- 30:45interface gap to zero. It always leaves
- 30:48a .2 mm Z gap. I reported this, and they
- 30:51said it is a confirmed bug, and it will
- 30:53be fixed in the final release of the
- 30:55slicer, but I of course haven't seen
- 30:57that, because it's going to be releasing
- 30:59about the same time as this video. I
- 31:02also had a lot of trouble with nozzle
- 31:04offsets, or maybe it was with bed
- 31:07leveling when using support material. My
- 31:10support interface layers were either not
- 31:12sticking to the material below, or the
- 31:14print material was not sticking to the
- 31:16top of the support. It appears that
- 31:19there's some kind of issue with the
- 31:20automatic mode that determines when to
- 31:23recalibrate nozzle offsets. I ran the
- 31:26printer calibration cycle, and switched
- 31:28the nozzle offset calibration and bed
- 31:31leveling calibration to on when sending
- 31:34a job to the printer, and the problem
- 31:36went away. I reported this to Bambu Lab,
- 31:39and the working theory is that there's
- 31:40something wrong with the auto mode that
- 31:42decides when recalibration is needed,
- 31:44and just switching it to on bypasses
- 31:47that entirely.
- 31:49There also appears to be a similar issue
- 31:50with flow dynamics calibration and the
- 31:5375A TPU. The printer should never
- 31:56attempt to calibrate the flow dynamics
- 31:58for the 75A material, because it's
- 32:01pre-calibrated from the factory. But, if
- 32:03you set the flow dynamics calibration to
- 32:06on when sending a print job, it tries to
- 32:08do it anyway and causes print quality
- 32:11issues. The solution for now is just to
- 32:13leave it set to auto. I I only changed
- 32:15it because I was trying to solve the
- 32:17support adhesion issue I just talked
- 32:19about, but the default works fine. On
- 32:22mode should also work though, and
- 32:23they're looking at it.
- 32:25I'll also mention that when you replace
- 32:27the tools in the rack, you need to
- 32:29insert them straight in, and then they
- 32:31drop down onto the nozzle pads. At
- 32:34first, I thought you were supposed to
- 32:35just hook the top and then swing the
- 32:37bottom in, but if you do that, it's easy
- 32:39to not get it fully seated, and it can
- 32:41fall out of the rack during the print.
- 32:44Ask me how I know.
- 32:46Once I started inserting them correctly,
- 32:47I didn't have any more issues.
- 32:50I also ran into a weird bug when
- 32:52performing a cold pull. The cold pull
- 32:54works great, but the printer lowers the
- 32:56bed a little bit during the process. I
- 32:59did several cold pulls in a row, and the
- 33:02bed collided with the bottom of the
- 33:03chamber and made a loud noise that did
- 33:06not make me happy, though it it doesn't
- 33:08appear that it hurt anything. I reported
- 33:10it, and I haven't seen the issue again
- 33:12after a firmware update. Maybe it was a
- 33:15weird corner case, or maybe they fixed
- 33:17it. I'm really not sure. And as long as
- 33:20I'm mentioning issues, I will mention
- 33:22the print job end music again.
- 33:25That's a bug, right?
- 33:28Right?
- 33:34So, what's the bottom line?
- 33:36This printer is a tour de force, if
- 33:38you'll excuse my French. It handles
- 33:40high-temperature engineering materials
- 33:42with a heated chamber. It handles super
- 33:44soft flexibles with the 2.85 mm filament
- 33:47support. It allows you to mix multiple
- 33:49flexible materials, multiple filament
- 33:51diameters, multiple nozzle diameters in
- 33:54a single print. And in my testing, I was
- 33:56able to freely mix different brands of
- 33:58filament and even weird stuff like
- 34:00graphene-filled TPU without any issues
- 34:03at all.
- 34:04This printer finally does all of the
- 34:07crazy things that I have wanted to do
- 34:09with a 3D printer since I built my first
- 34:11dual extruder machine back in 2014.
- 34:14And I think it bears pointing out that
- 34:16this isn't an experiment or a tech demo.
- 34:19This is a refined, well-engineered,
- 34:21mass-produced product with all the
- 34:24polish that we've come to expect from
- 34:25Bambu Lab.
- 34:27Now, the selling price of $24.99 in the
- 34:30US is incredibly aggressive. There is
- 34:33not another machine on the market that
- 34:35combines a high-count tool changer, a
- 34:37large build volume, an actively heated
- 34:39chamber, an enclosed filament management
- 34:41system, and all of the rest of these
- 34:43capabilities at
- 34:45really any price, let alone under
- 34:47$2,500.
- 34:49This is going to be a market disruptor.
- 34:52It's going to be very challenging for
- 34:54any other manufacturer in the market
- 34:56today to respond. I hope they do because
- 34:59competition benefits everyone, but this
- 35:02is going to be a hard act to follow.
- 35:04Now, to be fair, if your needs are
- 35:06simple and you just want to print
- 35:07multiple colors of PLA and PETG, there
- 35:10are plenty of other machines on the
- 35:12market that can do that at a much lower
- 35:14price point. But, if you routinely use
- 35:17higher temperature materials that need a
- 35:19heated chamber, and if you need to mix
- 35:21multiple flexibles in a single print,
- 35:23this might be the machine you've been
- 35:25waiting for.
- 35:26I know it's the machine that I have been
- 35:28waiting for.
- 35:29Like I said at the beginning, there is a
- 35:31link down in the video description, and
- 35:33if you found this useful, that's a great
- 35:35way to help support the channel.
- 35:37But, as always, if that's not something
- 35:38you want to do, I still hope you found
- 35:40this useful.
- 35:41Thank you for watching.
- 35:44>> [music]
- 35:53[music]
About this transcript
This page contains the full transcript of My Thoughts on the Bambu Lab H3C After Using it for a Real Project by Clough42, generated from the public captions YouTube serves with the video. The transcript has 6,054 words across 893 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.