120V machine - I thought it looked pretty good, too, and I think shows that temperature variation isn’t the cause of the slight adhesion artefacts I’m seeing.
Bambu Tech Support say they suspect a build plate issue so I’m experimenting further.
2 Likes
I’ve been VERY happy with bed adhesion and temp uniformity. I can’t see what the YouTubers were talking about. I’m in the boat that they had oily fingerprints on them. Obviously I can’t confirm, and I would HOPE they know better, but the time I’ve spent printing on it, has completely removed any doubt.
I’m using both the Smooth PEI and the Textured PEI, and things have been great. I think I’ve had two pieces pull off the bed and both are prints I should have used a brim on (thin surface area where it meets the bed).
I’m finding that the temps are really uniform (1-2C) by the time a print starts (from a dead cold/room temp starting point).
3 Likes
Received H2D today and am currently printing a large pla part covering almost the entire bed. No adhesion issues so far using stock settings and from my rudimentary testing temperature seems to be pretty uniform. Never seen a bed heat so fast.
It seems that the bed is made from a relatively thick piece of aluminum, which requires a bit of time to evenly distribute heat. The video on the youtube channel from Mike Snyder (I can’t post any links apparently) demonstrates that after a few minutes of warming up, the temperature distribution becomes quite uniform.
2 Likes
I’ve watched a couple of videos were people have unboxed handled the plate and just sent a print. I’m therefore not convinced how much prep was done correctly.
Segmented Heating is the solution for this case, every printer in the market with large bed has the same issue.
Only the prusa XL has made it correctly with with 4x4 heaters to compensate the time and uneven heat.
Here is a photo to cheat of:
3 Likes
for all its worth, while i have the gold pei plate and have had no warping on PLA, PETG or ABS, ive printed some large parts with PLA too and no issues.
3 Likes
I ordered H2D on day one and immediately had warping problems on the models that I normally print with P1S, A1Mini or X1C. I read the entire thread in these first days of use, I checked the temperatures of the bed with the non-professional tools that I have but honestly I still haven’t solved it, I find a deviation of about 8/10 degrees compared to the set temperature, but it could depend on the tools. Printing a single layer for the entire size of the bed I get a perfect sheet of PLA, while if I throw the supplied benchy into the printer after 6 or 7 layers it comes off the bed. The prints of my models that I have been publishing on makerworld for a year have all had warping defects. I cleaned the bed as per the wiki, dish soap and brush, as I do with the plates of my other bambu. I’m waiting for a response from support, I’m a bit demoralized!
2 Likes
First bed adhesion issue here. lost a small tree support on the back of a model. however,
This is the first time I printed with essentially no preheat time and did not increase the bed temp by 10c. I was also on sport mode.
This essentially confirms to me that this large of a bed requires more heat soak time at a higher temperature then is currently set as the default PLA profile, at the very least.
I would HIGHLY recommend preheating for at least 10 minutes before starting a print as well as increasing the bed temperature by 5-10c.
Interestingly I did not have any adhesion issues with the larger ABS parts I printed. I suspect the extended time of preheating the bed as well as the chamber means it is not a problem for the ABS profile.
2 Likes
CNCKitchen asked Tao why Bambu didn’t use a PCB heated bed instead, and the answer given was that Tao wasn’t comfortable with the fire hazard potential of having traces covered by what he called a thin layer of paint. I’m forgetting why now, but Tao also wasn’t comfortable with a silcone heater matt stuck underneath an aluminum bed either, as some other printers use. Their discussion danced around the topic and didn’t shed any light that I remember.
Summary of where things stand
Not everyone seems to be having problems, but some are reporting difficulty. For those people, the only solutions I’ve seen recommended are:
- Ensuring the build plate is clean
- Increasing build plate temperature by 10C.
- Heat soaking prior to printing
Some have suggested using a blue build plate. Not sure if those are yet available. Other suggestions were to use a glass build plate, or possibly a garolite build plate. A few suggested adding insulation underneath the heated bed. None of these are official recommendations, though, at least not yet.
The official Bambu position seems to be that there is no problem, while also saying a future firmware may help improve temperature uniformity.
2 Likes
Yeah, the other thing that struck me was no mention of 2 bed heat stages. Again, copying the H2D FAQ:
The bed has two heating elements. The high-power heating element operates during the heating phase to ensure a fast heating rate. It has a high heating power, resulting in a larger temperature difference. After reaching the target temperature, the smaller heating element takes over. As the high-power element turns off, the temperature becomes more uniform.
so as far as I have seen direct evidence of there is only the snake heater. The FAQ states there are 2. Is it just a 2 stage single snake element? A snake and flat bed? Which is it?
Edit:
Second Bed adhesion issue. This is more of a torture test this time in PETG with small tree supports near edge of bed. All but one (of 7) stayed successfully and the print finished with no deformation. However this test was done with preheating the bed to 75c and even the chamber to 45c awhile (20+ min) before printing. Going to redo with bed at 80c and see if its better.
1 Like
I’m surprised that there’s no mention of using adhesive on the build plate.
2 Likes
Good point. Also, no one (or at least almost no one) mentions exactly which PLA filament they are using when they are having problems. Who knows? Maybe that is a factor as well.
So, anyone who is reading this: please report which brand and model of filament you are using. That way we can look for patterns.
Well, on a personal note, I finally got back from travelling, and I just now printed my first high speed benchy (the one built into the H2D) using Sunlu HS-PLA and telling the printer it was generic HS-PLA. Upon closer inspection after pullling it off the plate, I see it printed not perfect, but acceptable. Actual print time, from start of purge line to done with smokestack was 15 minutes 37 seconds, according to my stopwatch. I haven’t gotten into challenging prints yet, but so far it definitely gives the impression of being both stout and next level. 
3 Likes
Now that I’m trying two color prints (so far still the pre-sliced built-in ones) is that it appears to start printing as much as 5C below the target temperature, and it can then overshoot by around 2C before reaching the target. Probably doesn’t matter on small HS-PLA prints, but I can imagine this may be less than ideal for a warp prone filament or a warp prone print. I say this based solely on the nozzle temperatures reported on the console screen. I don’t know whether it is or isn’t a factor in the problems that some people are reporting, but those people may want to observe and report on this as well.
If this turns out to be a factor, I would think it should be relatively easy to correct either in firmware or through added slicer commands, at the expense of prints taking a bit longer to complete.
1 Like
One final thought for tonight: noticing that the default sparse infill pattern is still grid:
I’m not going to comment on the sheer idiocy of that still being the slicer default. 
Nonetheless, I wouldn’t be surprised if even just that is what could be what’s causing detachments for some people who may be blithely printing with the default print profile for their early test prints. Not their fault, really.
This could turn out to be a true BFO: Blinding Flash of the Obvious. 
Hi, for now I have only used PLA Bambu, matte or basic, both with standard profile and with modified profile in terms of cooling, with the first 3 layers without cooling and Aux fan off. I also tested with filaments, always PLA, azurefilm or esun pla+ and always with standard profiles or with the modifications I told you about above.
To also respond to what you wrote two posts below, I always use gyroid filling, in fact the detachment of the models occurs because the nozzle hits the warped corner and not the infill.
1 Like
@mteomarch
If you’re willing, may I suggest you post a .3mf file for one or more of your most likely to fail prints? I’ll try printing it and see if I get the same failure. I’m looking for a good test print anyway, so it could be a win-win for both of us.
At the moment I’m finishing up my first sliced print on the H2D using yellow Sunlu HS-PLA from amazon. So far all I did was pre-emptively increase the bed temperature to 70C, but aside from that I’m running straight stock print profile, including the grid infill because I overlooked it before launching the print. I did zero filament tuning, which I ordinarily would do whenever onboarding a new filament.
So far no detachment, but it could also be that I was simply lucky regarding its placement on the heated bed.
Napoletane-Modiano-Borbone.3mf (833,9 KB)
Here is the file, as you will see I changed 3 values in the filament profile, fan off for the first 3 layers, aux off and textured PEI temperature 65 - 60.
Also height first layer 0.2 and line width initial layer 0.3. These are all settings that I have been using for over a year on p1s and a1 mini!
In the photo you can see the defect!
The filling is not changed in gyroid, but in practice it is absent!
Thanks for your help!
@mteomarch I’ll give it a shot tomorrow when I’m more awake and can keep an eye on it. Whether it fails or succeeds I’ll post the result either way after printing it. Luckily I’m stocked to the gills on filament, and this will be a good chance to burn off some pink filament that I don’t otherwise have much use for. For testing purposes, color probably won’t matter.
1 Like
Any color will do, that’s not the problem! Thanks