[GUIDE] Mecha's Comprehensive Guide to PVA Supports for the H2D

Mecha’s Comprehensive Guide to PVA Supports for the H2D

Updated: 7/21/2025

Hello everyone,

Over the past few weeks, I’ve been bashing my head against a wall trying to perfect the printing process for using PVA supports on organic, rounded miniatures—and I’m happy to say I’ve finally developed a repeatable, functional, and clean workflow that solves many of the core issues with Bambu’s H2D presets.

In this post, I’ll cover:

  • The problem that led me to use PVA in the first place
  • Common issues and community advice
  • My final working solution, settings, and tips

:gear: If you’re just here for the settings, you can find the 3MF file at the bottom of this post.


The Problem

I wanted to print this highly detailed model of Tracer from Overwatch in PLA Silk / Galaxy. These specialty filaments rely on a perfect surface finish, so post-processing like sanding or filling is not an option—it ruins the sheen. While some careful heat-treating can restore a bit of luster, it’s nowhere near the clean finish that a pristine print provides.

If I were printing with standard PLA, I could’ve used tree supports and cleaned it up afterward. But even then, the bottom-facing surfaces would suffer. Soluble supports like Bambu’s PVA seemed like the ideal solution.

“Luckily, I have a Bambu dual-extrusion printer, so this should be easy!”
— said a much more naive version of me.

What followed was weeks of print failures, clogs, motor overloads, and forum-diving hell. Most discussions ended with “PVA sucks, don’t use it.” Still, I was determined—others had made it work, so I knew it was possible.


Drying

You’ve heard it a thousand times, but it bears repeating:

:arrow_right: DRY. YOUR. FILAMENT.

PVA is extremely hygroscopic—even more so than the desiccants you store it with. My local humidity hovers around 55% RH, and just a few minutes of exposure caused bubbling and stringing. I originally followed Bambu’s drying instructions using a Creality Space Pi 2 dryer, but its tiny vent holes allowed the PVA to reabsorb moisture shortly after drying.

Now, I keep my PVA drying at 60°C continuously—even during printing.
If your dry box is well-sealed, you might get away with transferring it quickly after drying, but PVA is fickle enough that it’s worth the constant heat. Just don’t go much higher than 60°C, or you risk making it too soft, which can cause jams or slipping.

Also: make sure your filament path from the dryer/dry box to the extruder is fully sealed with PTFE tubing.

Crackling? Bubbling? Stringing? Poor print quality?
Your PVA probably needs more drying.


“Extruder Motor Overloaded” / “Filament Not Extruding Properly”

This one nearly broke me.

Nearly every extruder switch would trigger the dreaded beeping, followed by:

  • Extruder motor overloaded
  • Filament not extruding properly

I’d remove the filament, snip the end, maybe disassemble the extruder, and reattempt the print—only to fail again. Sometimes the PVA would ooze so much that it caused toolhead switch failures.

Eventually, I realized these errors only happened during nozzle switches, which pointed to a temperature and retraction issue—specifically, PVA sitting too long in a hot nozzle. Left idle, PVA becomes a brown, hardened tar that clogs everything. It also softens drastically, causing the extruder gear to slip or deform it into a clog-inducing blob.

At first, I thought active drying during printing was the issue. It wasn’t.
The real culprit? Nozzle temperature.

Bambu’s H2D defaulted to 240°C for PVA—way too high. Most X1C profiles use 220°C, and this GitHub issue report confirmed my suspicions.

Fixes:

  • :white_check_mark: Lowered PVA temp to 220°C
  • :white_check_mark: Moved PVA to the right (fixed) extruder

Using the fixed extruder prevents soft filament jams and follows Bambu’s own TPU guidance, which oddly doesn’t apply the same caution to PVA—but should.

Once I made these changes, I never saw an overload or feed error again.


Adhesion Issues: PVA and PLA Not Sticking

Some guides suggest using PVA only for the interface layer, but on complex organic models, I found it more reliable to use PVA for the entire support.

When I tried hybrid supports (PLA support + PVA interface), they often failed:

  • PVA wouldn’t adhere to PLA
  • PLA wouldn’t adhere to PVA
  • Supports would spaghetti mid-print

This was made worse by PLA Silk/Galaxy’s slick surface. Fortunately, forum user cbath shared this post, which became the foundation for my solution.

He advocated building a PVA block “mold” around the model, ensuring consistent adhesion PVA-to-PVA, not PLA-to-PVA.

I tweaked his method slightly and ended up with this reliable configuration:


Final PVA Settings (highlighted in white in screenshot)

  • Support Type: Normal
    Threshold Angle: 15°
    → Maximizes surface coverage and stability.

  • Raft Layers: 2
    → Prevents floating PVA islands.

  • Support Base + Interface: PVA
    → Full PVA provides better consistency.

  • Initial + Support Layer Expansion: 6mm / 5mm
    → Builds the surrounding “mold”; tweak as needed if contact looks weak.

  • Top/Bottom/XY Distance: 0mm
    → Ensures tight, clean surface contact.

  • Base Pattern Spacing: 1mm
    → Increases grid density; you can try 2mm to save filament but expect tradeoffs.

  • Interface Pattern: Rectilinear Grid
    → Maximizes surface contact; reduces floating islands.

  • Independent Support Layer Height: OFF
    → Keeps everything at consistent height.

:warning: Lower all support print speeds to 10–30mm/s to reduce failures.


The “Island” Problem

Even with reliable supports, I ran into a big issue with tiny PLA “islands” that form in early layers of curved organic models.


These islands are prone to detaching, ruining the print. Lower layer heights (required for detail) and slippery filaments make it worse.

Fixes:

  • :white_check_mark: Slowed all PLA and PVA print speeds to 10–30mm/s, including infill

  • :white_check_mark: Disabled extra Z-hops and wiping moves, such as:

    • Reduce Infill Retraction
    • Flush into object infill
    • Wipe when retracting
  • :white_check_mark: Increased PLA temp by 10°C to improve first-layer adhesion

  • :white_check_mark: Used Gyroid infill, which seemed more stable than Monotonic

  • :white_check_mark: Oriented the print for more surface contact when the island first forms

These changes significantly reduced movement-induced failures.


Conclusion

PVA is still a pain to work with and requires many compromises—but if you’re stuck (like I was), I hope this guide gives you the clarity and working baseline you need.

The surface finish I achieved simply would not have been possible with traditional supports.

:point_right: I highly recommend importing the 3MF file rather than copying settings manually—there may be additional tweaks not covered here.

Happy to answer questions or troubleshoot with anyone going through the same ordeal.

Like another PVA warrior once told me:
Peace and good luck! :v:


22 Likes

Nice post! I was just curious—was there something specific in Bambu Lab’s PVA instructions that didn’t work well for you?

It looks like you decided to go your own route, so I assume you encountered some issues with their guidelines. I’ve printed several kilos of PVA myself—both as support material and even full prints, since I use it for fishing bait boxes. In my experience, the biggest challenge has always been moisture. If the filament isn’t completely dry, it tends to cause all sorts of problems—from the AMS all the way to the nozzle. I also avoid touching it with bare hands, for the same reason you mentioned—PVA absorbs moisture very quickly.

Would be really interesting to hear your thoughts on where Bambu Lab’s guide fell short and how your approach improved things.

I had these same issues trying to use Support for PLA / PETG, also called Support W, that I bought during the anniversary sale recently. I would like to see how many of your tips apply to that as well.

Nice post! Now do PLA and PETG with dual extruder :stuck_out_tongue:

2 Likes

Excellent post, and thanks for the shout out!

On the Island adhesion, yes, pushing the PLA temp as high as you can get away with was very important to my success. I think I went even higher than the 10C you did.

I like your idea with using the right (fixed) extruder as well. I should definitely have thought of that. Will use in the future. It was just easier to route from my Sunlu dryer to the left (free) nozzle.

I eventually switched to using BVOH soluble support instead of PVA and it is a SIGNIFICANT improvement in every way. I think if you can stomach the higher price, combined with all of the best practices you have outlined here, you will be extremely happy.

my conclusion is that Bambu needs to develop their own BVOH filament.

Awesome work.

2 Likes

Thanks! I agree—Bambu does a good job emphasizing the importance of drying, and their guidance around reducing support speed and minimizing Z distance is certainly on the right track. Moisture really is the #1 enemy with PVA, and I also feel the same way about handling it with my bare hands haha.

That said, I found that their PVA guide tends to focus on simpler or more geometrically stable prints, like their showcased encased D20. While that model is complex in shape, it has well-connected surfaces and consistent layers that don’t present the same challenges as organic miniatures with small, floating “islands” and highly curved overhangs.

As a matter of fact, they don’t really display any sort of rounded model with PVA, and they don’t show any models with PVA just as an interface. I suspect that it may be because they ran into the same issues that I did.

Additionally, the recommended nozzle temperature of 240-250°C for PVA on the H2D seems unnecessarily high. Along moving to the right nozzle, that one detail alone made a big difference in reducing motor overloads and oozing on toolhead switches.

I also spent some time experimenting with dedicated support materials—mainly Support for PLA (new version) and some PETG*

The technique can help in certain cases, but I found that the core issue with these materials is their inherent lack of cross-material adhesion. You end up building a kind of multi-layer “sandwich”: first the support material on top of PLA, then PLA back on top of the support. And unfortunately, none of these layers really want to stick to each other.

What you get is a slippery sandwich of doom—and when your “ingredients” are curved, irregular surfaces (like in organic models), rather than flat, stable planes, the whole structure starts falling apart. While this isn’t usually a problem for flat-bottomed models, on organic prints, those tiny isolated blobs of support essentially double the islanding issue.

That’s why I eventually switched to using PVA for the entire support structure. By relying solely on PVA-to-PVA adhesion, I was able to get much more reliable results—fewer detached supports, better surface quality, and significantly more consistency on complex shapes.

No problem! I’m getting some BVOH soon and hearing this feedback from you makes it all the more exciting :grin:

My trick is to pause and manually melt a dot of PETG onto PLA to ensure adhesion. For whatever reason, PETG cannot stick to PLA when using dual extruders.

I’m running a print I’ve had a ton of trouble with, using your .3mf settings @mecha. It’s actually going well, haha. I swore off PVA after several attempts on my H2D. Other than my prime tower needing some PVA clipped off, it’s looking great. I dried the PVA at AMS HT recommended settings for 18 hours. I have it set to 60c currently while running.
Speaking of the H2D prime towers, they are always an absolute mess on my machine. Does anyone else have that issue? Like I have a million strings on the outsides of the prime tower. For the most part it doesn’t affect my print, but it does cause some residual filament to get stuck on the nozzle at times.

I’m happy to hear that it’s working and thanks for trying my settings! The prime tower was another issue that I encountered and it can leave a stringy and bumpy mess that occurs more often with PVA. I think its because the printer is purging the overheated idle filament thats left in the nozzle, and PVA is just prone to oozing.

It’s doing it’s job though which is good, I also believe the default prime tower speed is 90mm/s which is seperate from all filament extrusion speeds, perhaps turning that down to match the other values can stop the stringing?

Anyways, let me know how your print goes and if you had any issues!

1 Like

I disagree with keeping it heated when printing but otherwise think this is a good post.

If the filament is properly sealed then there isn’t enough outside influence and it should reach an equilibrium. Heated filament /leaving the container to enter your printer is warm/ and thus risks picking up moisture more easily as its residual heat will speed up the chemical reaction and could in theory permamently bond water to the PVA and so no amount of drying will save it.

Dry it for the time needed and then don’t continue drying it.

Just keep it sealed and give it a short dry before use if its been a day or two or a longer one if its been a while.

Yeah, there’s definitely a line between overdoing it and doing what’s required. With how volatile PVA is I opted to fully eliminate the variable of wet filament, ensuring that the connection from the dryer all the way to the printer is also sealed.

I didn’t spend too much time optimizing and perfecting these parameters, as I made this when I finally got some reliability coming from what was previously exclusively print failures – so I’m hoping that as we collect results I can refine this process.

Thank you for your insight!

It’s cool! I just ruined a few rolls of petg thinking “well if its always dry…” and then it turned out that the residual water in the air inside its conainer reacted with the filament even faster thanks to the heat so now I’m kinda paranoid.

I continually dry TPU as I print. Have not had a single failure with that setup yet. The AMS HT makes it easy. I wish you could just feed it normally while drying PVA in the HT, but setting as an external spool is pretty quick. Printing and using the H2D quick nozzle swap off Maker world is an absolute necessity. It makes wrangling all these different filaments so much easier.

Interesting. I NEVER dry TPU and get decent prints on my Prusa XL. I have also printed TPU on my Bambu Labs X1C without drying it. I wonder why our experiences are so varied.

Depends on where you live most likely. I’m in SoCal, so our humidity is pretty stable most of the year. I’m sure I could get away without drying, but my results with drying speak for themselves, so I dare not change my workflow :wink:

Thanks for sharing!
252525

Never dry mine either; x1c(2), P1S, HD2 even Kobra Max and past s1 pro. Just make sure the decadent in the vacuum bags is always monitored and fresh.

I leave my NinjaTek Cheetah out for weeks and months. I live in Vermont, which right now is quite humid. Puzzling.