The woes of owning an X2D

I bought the printer close to its release date. While it took me a couple of weeks before I had time to unpack it, I have now been printing with it for two to three months, long enough to get a solid impression of its strengths and weaknesses.

Background

I’ve been 3D printing for three to four years now. I cut my teeth on low-end Ender machines, which made you feel more like a mechanic than a maker. Eventually, I graduated to an early P1P right when it was released, added an AMS, and upgraded it to a P1S over time. That machine ultimately logged over 5,000 hours, pushing me into the gritty side of maintenance, replacing various components and culminating in a full timing belt swap that took an entire weekend. Finally, I upgraded to the X2D a couple of months ago.

In general, working with Ender printers felt like being a mechanic who occasionally got to print a model. The P1P, by contrast, felt like using a microwave: you start it, and it just works. Adding the AMS did increased the failure rate tenfold however. That is somewhat expected, the more complex and longer the mechanical pathways, the higher the chance of failure. To be clear, single-color prints (even through the AMS) still “just work”. Most print issues only arise during multi-color jobs.

The New Flagship: Bambu Lab X2D

I mostly print multi-color items: toys, figurines, and creatures, primarily for fun. I currently have three AMS units attached to the printer, allowing for up to 12-color prints. Occasionally, I charge people when they ask for large, complex, or custom pieces, so I consider myself a “hobbyist-plus.” My prints are usually not very large, but due to their multi-color nature, they take a long time and typically generate significantly more waste than the weight of the actual model.

I originally intended to buy a P2S, but when the X2D dropped, it was enticing. It was only $100 more and featured many more sensors to reduce mid-print failures, a crucial factor given the duration and complexity of my jobs.

As for time and filament savings: I didn’t go in blind. I watched every hands-on review I could find, and they all made the same promises: massive time savings (up to 50% faster multi-color prints) and significant reductions in purged filament. It seemed like a win-win-win.

The reality couldn’t be further from the truth. I saw a post here claiming the X2D is a well-designed printer held back by bad firmware. I disagree. It feels like a poorly designed printer where the firmware is desperately trying to compensate.

The Initial Promise vs. Reality

The core promise from Bambu (and the many YouTubers promoting the X2D) was simple: drastically lower print times and far less waste.

Let’s dispel those claims. As I detailed in a forum post the X2D frequently takes the same time, or longer, to print while offering only marginal filament savings. In the example I posted, a two-color multi-nozzle print took three hours (15%) longer while saving a whopping 20 grams (3%) of filament.

This isn’t a bug; it’s by design. The secondary nozzle is inherently slower, so anything it prints increases total time.

The waste reductions are equally misleading. Model waste falls into three categories: supports, the wipe tower, and purge material.

  • Supports: Cannot be eliminated.

  • Purge: The X2D reduces purge volume, but that depends on color count. If you use more than two colors, you still require color swaps and purges, even if reduced.

  • Wipe Tower: This was the biggest surprise. The tower exists to stabilize extrusion pressure when starting a new filament. I expected it wouldn’t be needed once filaments were loaded into both nozzles (e.g., in a two-color print). Yet the X2D still prints a wipe tower, and actually uses more filament on it. Bambu claims this stabilizes internal nozzle pressure, but competitors like Prusa achieve the same result using a micro-purge of approximately 13 milligrams (0.013 g).

The Filament Switch

For the X2D to be genuinely useful, you need the Filament Track Switch. It shouldn’t be optional; it should have been included by default.

The main problem is that the X2D is already a mechanically complex machine. Adding the filament switch makes it even worse. Compare the rear profile of the X1C to the X2D.

X1C

X2D

Every extra mechanical component or PTFE tube represents another potential failure point that requires clearing, repairing, or replacing.

Furthermore, the filament switch does not support Silk PLA, a basic filament type that existed long before Bambu started making printers, rather than a niche new material.

Maintenance Pains

Remember how Bambu printers initially felt like microwaves compared to old Enders? That is no longer true with the X2D.

One of my first issues was a “clogged extruder” error message, which turned out to be a clogged nozzle. But I tried to follow the (Bambu-provided) instructions to actually disassemble & clean the extruder as the manual suggest. The toolhead is extremely complex, not-well-documented (looking at you filament cutters), and full of delicate connectors that must fit precisely. Parts are easy to assemble incorrectly and fragile enough to break during maintenance.

On a P1P or P1S, replacing the entire extruder takes about 30 minutes once you know what you are doing. On the X2D, replacing a single component of the extruder, without even fully disassembling it, took me over two hours.

The difference shows in Bambu’s official instruction videos: the P1S extruder replacement video is two minutes long (granted it skips most of the process), whereas the X2D video takes 12 minutes (again, most of the process skipped).

The Secondary Extruder

The known trade-offs with the secondary extruder are slower speed and slightly lower print quality. Most online comparison videos struggle to show quality differences because they use small test models. The issues with the secondary extruder take time to manifest.

Once you attempt a significant multi-hour print using the secondary extruder, the flaws become obvious:

  1. Under-extrusion & Misalignment: You initially notice minor under-extrusion, occasional missing filament, and slight layer misalignment.

  2. Structural Failures: Over long prints, under-extrusion creates tiny gaps and compromises layer adhesion.

  3. Cascading Print Failures: Weakened layer adhesion causes sections of the model or support structures to break mid-print. Because supports are structurally thinner by design, they break easily, ruining the entire job.

Since Bambu Studio does not allow you to specify which nozzle prints support structures, you cannot mitigate this risk. Once you slice a large multi-color model, chances are some support layers will be assigned to the secondary nozzle, and those are precisely the layers that fail.

It is ironic that Bambu advertises the secondary nozzle as ideal for printing supports, when using it for supports might be the worst thing you can do.

even full tool changers use a prime tower (wipe tower) (edit new slicers are starting to remove much of the lower parts of the prime tower, in cases where the colur changes only occur higher up)

The X2D is a bit weird, that second nozzle is like maybe not the best thing. I think it’s OK if you link an AMS to main nozzle and use the secondary for support only. To be honest I think that is how it was advertised.

I think the main use of the filament track switch is that you don’t need to reload your AMSs before starting a print. During a print, it does basically nothing because there is only 1 outlet on each AMS

I agree. However:

  1. There is no explanation why a purge-less print needs a larger wipe tower than a regular print
  2. There are some printers (like the Prusa Core One) that no longer need the wipe tower.

Even if you use the secondary nozzle for support only (as it was advertised to work well) - that would still be a huge mistake. As I explained in my last paragraph of the post, printing supports using the secondary nozzle would make the supports (that are weaker than the model by default) even more brittle due to under-extrusion and bad layer adhesion. So any supports printed with the secondary nozzle are prone to break mid-print and fail the entire print job.

No, that’s not quite the case.

If you have 1 AMS - it allows you to use filament from that AMS on both nozzles.

And if you have 2 AMSes it’s helps you increase speed & quality by allowing you to print any filament from any AMS at any nozzle at any time. So for example if you have a print that has the bottom half mostly red, but then the top half mostly yellow with red dots. The printer can use the main nozzle to print the red of the bottom half. And then switch nozzles and print the yellow via the main nozzle, while the red dots will be printed by the secondary nozzle.

Mutlicolor/Multifilament prints are the main reason to use the Filament Switch. If you’re printing with a single color/filament, it doesn’t do you any good.

OK this is true, but then you’re printing non-support material with the derpy print head, which I do not think should be done. To be honest I do not blame the user for trying it. Bambu’s marketing is not very clear. It starts out great saying the 2nd nozzle is for support, then later on it talks about using it for multi-colour.

OK I had no idea the indx was actually shipping

As a sidebar has the INDX been demonstrated working in the field (it always works for content creators in their labs somehow)? They generally have problems with launch, the XL took 4 years to work, the Core One had VFAs for 6 months until they worked out how to fix the belts, the MMU is still a meme.

Then again, the X2D also has many problems,

It is ironic that Bambu advertises the secondary nozzle as ideal for printing supports, when using it for supports might be the worst thing you can do.

Even if you use the secondary nozzle for support only (as it was advertised to work well) - that would still be a huge mistake. As I explained in my last paragraph of the post, printing supports using the secondary nozzle would make the supports (that are weaker than the model by default) even more brittle due to under-extrusion and bad layer adhesion. So any supports printed with the secondary nozzle are prone to break mid-print and fail the entire print job.

You know what they say: anecdotal evidence is the evidence of an anecdote.

My experience with the X2D (which I would not call a flagship by any metric) is the opposite. I have been printing highly complex PETG parts with PLA supports. Parts that should be perfect as they are used to test assemblies that would be machined. Even the smallest DIN 5480 splines came out perfectly and mated together. In the extensive hours of printing all these parts, I have had no failures on the supports. In fact all failures were creations of my own doing, like printing supports that were so strong they were impossible to break away.

So with two contradicting anecdotes, where does this leave us?

This is definitely supported, and even advertised.

It’s actually part of the Bambu Slicer. Once you slice a model for X2D you have an option to Manually assign filaments, to have a “Quality” assignment (i.e. all filaments assigned to the main nozzle) or to have a “Filament-saving” assignment - which means the filaments will be auto-assigned to nozzles and swapped mid-print, in order to produce the most filament saving result.

100% agree on this. I generally have no issues with the printer other than the toolhead. It is a nightmare to work on. I had to take it apart even before I could start my first print as it came from the factory with a piece of filament stuck in the filament sensor. WTF?

The front of the toolhead is nicely designed as it is held on with magnets and the wiring harness detaches with a press of the clip. The nozzles are also easily removed. So far, an A+ design. Then the rest gets a solid F. Having to remove 8 tiny screws to remove the cover is nuts. I am surprised I have not lost any yet. It should be held on with clips or magnets, not with screws. Then every other connector is tiny and not easy to remove or seat without the possibility of damage. More tiny screws than I can count and all different sizes.

Having to clear stuck filament is a nightmare. I recently had to take it apart to clear out stuck filament and I somehow broke the left eddy sensor in the process. So had to take it all apart again and wait on a part. Replaced the sensor. Now the cooling fan no longer works. Currently waiting on that part and not looking forward to having to take it all apart again. Printer will likely be down 3+ weeks and require extra 2 parts due to the complexity of maintenance on the toolhead.

I realize the toolhead is a tremendous piece of engineering from a technical perspective. Fitting all those components in that small space is extraordinary. But they need to consider that stuck filament and part replacement is commonly required and design for that as well.

As I mentioned in my post “Most online comparison videos struggle to show quality differences because they use small test models.”
It all boils down to how big is the model you’re printing.
If you’re printing a model that takes a couple of hours - makes sense everything will look fine.
But if you’re printing a model that takes dozens (or hundreds) of hours - I suspect you would encounter the same issues as me.

“Most online comparison videos struggle to show quality differences because they use small test models.”

Given the current landscape of content creators, sorry I mean influencers/marketing agencies that’s a really low bar. But that’s a discussion for another day.

It all boils down to how big is the model you’re printing.

Are prints approx. 200x200x200 big enough, or prints going over 26hrs long enough for you?
The answer is still no, I have not encountered the same issues as you.

And hundreds of hours for a single print? on a 235x256x256 build volume? I want to see that.

It might not be as large as 200x200x200, but it still takes over 100 hours to print

dear god

A full kilogram of plastic in the bin

I don’t know, which reviews you watched. The ones I watched, all rated the X2D as the perfect budget printer for engineering filaments. For multicolor, it was never ideal. It also isn’t a flagship printer. that’s the H2x series.

So I think you just bought the wrong printer for your use case.

Honestly, I would be pretty stupid to buy an H2C purely to reduce multicolor waste.

P2S Combo costs $750. H2C Combo costs $2,200. That’s a $1,450 difference.

Considering average price of 1kg of filament at $20, that’s 72.5 kilograms of poop you need to save, in order for this to be economical.

It’s MUCH cheaper to have a P2S and let it poop all day long. Or even buy 2, for x2 the print speed. Than buying a single H2C for multicolor “poop” economy.

P.S. If you want an example of a review: https://youtu.be/Vm0biF0_-jw?t=671

Just referring to your claim that the X2D was advertised and tested as a great multicolor printer. It is just marginally better at that than the P2S or older P/X series printers. The reviews I saw clearly stated that.
For 2-color I find it great, but I never expected it to save much time or waste for 4 or even more color prints. At time of release, the only good BambuLab option for that was the H2C.
At that time, there was not much choice of multi-nozzle printers under 1000 $/€. I think Snapmaker U1 was the only relevant option available (or on the horizon?).

I just wanted to point out, that your frustration seems to me more a case of wrong expectations, than a bad or wrongly advertised product.

I mentioned the H2x series as the actual flagship series. the X2D is not as you claimed. It is a budget engineering material printer. And phenomenal at that.

And regarding price of H2C for multicolor PLA: if you directly compare it like you do, then yes, probably P2S would be the more reasonable choice for you. Other factors besides filament cost could be print time per project, which is a fraction of a single or dual nozzle printer. Bigger build volume would be another benefit.

If your only interest is reducing cost by less waste, then the best option at that time probably would have been the U1. With your previous background, it should have been clear to you, that the X2D is not the best printer for that task.

I’m speaking to just the “poop” here and not the prime towers. Those towers can be minimized but are generally necessary for the stuff I print.
Some would be surprised at just how much poop waste a dual nozzle printer can put out. I noticed my buckets filling up all too often for my H2D and decided to monitor each print. I do the normal household stuff and lots of fun toys and whatnots. An example - I printed one multicolor model in PLA on my H2D. It was for a gift. It wasted 3/4 of a spool (pushing 800g including the tower according to my scale). I ordered my H2C that week. Haven’t had anywhere near that much waste in total for all the printing I’ve done since - not one full bucket from either printer in over six months. For a hobbyist like me, the H2C is worth its weight (a bit on the nose). I love that I can easily run a dozen or more colors through it and have minimal waste. It fits my workflow. And for the record, those single and dual nozzle printers can generate some SERIOUS waste if you print the kinda stuff I do - lots of colors and such. That 75kg’s could be surpassed relatively quickly.

Let’s be a little kinder here and not blame the user for the product not meeting expectations set by the manufacturer.

I just re-read the entire X2D product page to see what it says about multi-color printing. While it refers several times to using the dual-nozzles for supports, nowhere on that page does it say “Don’t use it for multi-color prints”, or anything that even hints the second nozzle may not work as well as the primary.

Those of us who like to nerd out and go deep into product research know the X2D is designed for functional prints, not aesthetic multi-color prints. But a customer relying on Bambu Lab’s own product representation is not going to know that.

I don’t think it’s appropriate to blame @AlexM for buying the printer for his purposes. Bambu Lab is the entity to blame for not clearly communicating the product’s limitations and appropriate uses. Clearly it’s not the best printer for AlexM, but if there is a failure in this story, it is Bambu Lab’s product marketing, not the customer.

(I suspect AlexM would be well served by an H2D series printer – it is meant for just the kind of multi-color printing that he needs. But that’s an expensive printer that not everyone can afford, and also, based on the reports of rapid lead screw wear, I’d be hesitant to buy it.)

Now you’re making me feel stupid, I’m even paying to dispose of waste (some of which costs a lot per kg) when I could just dump it in the ditch for free… economical :melting_face:

You are a very unpleasant person to converse with. I notice this both in this thread, and the other thread where you replied to my message.

In this case separating the weapon from the hand was not an option, because the person I’m printing this for explicitly asked for the rabbit to be printed whole. Without cutting it to pieces. And even cutting it to pieces, I would not gain much. The blue of the gun is only 1 color. The Rabbit’s head alone has like 6 colors. The gun has 3. The body has 5.

But you took no effort in asking me why I decided to print it this way, before accusing me of lack of basic 3D printing skills. What does that say of you?

The orientation you offered is not a good idea. Not sure why you think it is.

  1. You would only fit 3 rabbits, because you need room for the wipe tower
  2. Supports usually cause imperfections to remain on the model once removed. So instead of 4-5 small patches being supported, the whole back of the rabbit would remain uneven due to removed supports. The alternative of using a PLA-supports/PETG interface on the supports - will significantly increase print time.
  3. Instead of having a lot of layers with few color swaps per layer - you would now have a few layers with a lot of swaps per layer. What exactly are we gaining here?

Instead of 5 rabbits printed in 5 days, we get 3 rabbits printed in 3 days.

For example: https://youtu.be/Vm0biF0_-jw?t=671

And here’s another one, with same results: https://youtu.be/IsclJeIuvME?t=207

And there are plenty of others…

Sharky does have some rough edges, and tends to be blunt. Despite this, I find them worth engaging with; @RetroSharky is extremely knowledgeable, willing to share what they’ve learned, and in my experience participates in good faith and trying to help. I feel pretty lucky to have them around, even if I don’t always agree with them :slightly_smiling_face: