The information is protected (in the sense that it costs money to read it), so I can only quote small passages, but here is some context:
At least one of these requirements must be met:
Reduce the likelihood of ignition: Equipment is so designed that under single fault conditions no part shall have sustained flaming.
Control fire spread: Selection and application of supplementary safeguards for components, wiring, materials and constructional measures that reduce the spread of fire and, where necessary, by the use of a second supplementary safeguard such as a fire enclosure.
So… I would say yes - attention: opinion - because no sustained flaming. The crucial aspect is sustained.
What is the fire enclosure in this case? Is it the plastic shell of the printer?
Here’s where I’m going with this: is it still an enclosure after it’s been melted through, as in some of the photos posted earlier in the thread? Or does that not count? i.e. it only matters that an enclosure existed at some point in the past? There’s no requirement that the enclosure remain enclosed and not be breached?
The standard does not address the issue of whether the enclousure melts - In the strict sense, that’s another story… This is something that the engineer has to calculate.
In the event of an incident, what is important for the standard is that the device does not catch fire (sustained flaming) and the fire does not spread.
From this perspective, the standard assumes that a device can always catch fire. This point is important: the reality that electronics can always burn.
The important thing is that any fire is contained and not spread.
Therefore, in my personal opinion, I consider the standard to be fulfilled. Which brings us back to the beginning of the actual topic, almost back to the first comments: It smolders but doesn’t “sustained flaming”.
There’s no reason to think 3D Musketeers is using this video to increase his membership. I take the warning seriously. I recently sold my A1 but I contacted the new owner about it.
I think A1 owners should do three things:
Yeah I did the math in about message number 10 in this thread, but I deleted it since Avocado downvoted it. His PhD trumps my MSEE. I don’t remember the numbers, mods are free to restore it, no objection.
It has been such a tornado of information that I can’t remember now whether any of the melt-throughs and/or destroyed NTCs happened at 110v. Did your calculations show that “the problem” will only affect those running on 240vac and not any of us running on 110vac?
It’s a guess, but the real fault could be elsewhere in the circuit and the NTC blowing up just the result. Also some small percentage could just be bad parts or some number could have been mishandled (cracked during insertion). What I’m saying is that under normal operation on 120 the NTC gets a lot less hot and a lot less stressed than you might expect. So because of that lower stress if the problem is the result of an underspec’d part or poor design (which seems to be the leading argument) I expect that to show up in the 120 to 240 failure ratio. Not so much if there is something else going on.
Maybe someone with more background about global distribution (even Avocado, I think this is his field) could shed some light on what the ratio of 120 to 220 should be in the deployed printer population. There are more 220 outlets in the world, but the US buys more than their share of printers.
Addendum, I also mentioned pure thermal cycle count as a possible failure mechanism. Certainly they are good for “thousands” of cycles, but I wouldn’t expect 10’s of thousands of cycles out of one. Hard to argue for this mechanism unless it is cooling down every time the heater cylces even during a print.
I’m going to correct myself here, see maybe this is why Avocado thought it was wrong.
It dissipates a lot less power on 120. Whether it gets a lot less hot, well, it needs to get hot to work. I’d imagine there is an equilibrium that drives the temperature to the flatter part of the curve independent of current.
I could write a novel about it, but it’s much easier to refer to Bambu’s data sheet:
Max power: 1300W@220V - 350W@110V
There’s no need to calculate anything for the 110 V system because it’s so far away that it’s not worth the effort.
The heat bed accepts anything from 100V to 240V at 7A. That’s just a theoretical, calculated value of 700W to 1680W. However, since the printer only draws 350W@110V, the calculated current is: 3,2A.
At 1300W@220V = 5,9A
In short: The heating runs on low power on 110V devices. That is a safety margin of almost 50% - as far as ampere / current is concerned (not watts).
Edit: …and I would like to add that all the components we are talking about that were left out were probably originally only intended for 230V. To keep power under control, “start smooth”. With a 110V power supply, the heat bed gets bored anyway when it only gets 3.2A and can actually handle 7A.
Oh, and just as a note: the 3.2A is the calculated value. In reality, it’s even lower because the other components also require power.
Well let’s go through the math again and figure out why Avocado thought I was wrong. I originally was doing it on my mobile using “calculator” in one window while reading pdf’s on the other and there must be a mistake in there.
The datasheet above (which looks typical) shows a residual resistance of .112 for the 5D-15
On 220V an A1 (peak, heating) draws 1300W at 220V or 5.9A
on 110-120V (we’ll use 115) 350W or 3A
So power dissipated by the NTC would be 3.89W for 220V which seems close to the 4 I estimated. But on 115V, it is 1.008W, which is much higher than the 1/3 I estimated. I bet I squared the resistance instead of the current.
So the ridicule and derision Avocado heaped on me was justified and I hang my head shame. My initial instinct was correct to delete it.
Excellent! IIRC maybe one or two people on the thread may have mentioned that “the problem” might not be a problem for people running on 110v, but your conclusion is much stronger and seems to render the issue either entirely moot or close to it for those powering their A1 from 110VAC. With the problem now sorted, at least to that degree, and myself being one of the 110v tribe, I feel as though I can move on now and not look back. I’m outside the problem’s reach.
Thank you! And thank you @mugglesmuggle , and thank you to everyone who contributed to understanding the problem and getting it sorted at least this far. I already feel lighter not having this albatross around my neck.
Have a look at how the topic started and how it’s now almost come to an end, at least as far as we at community can get. I’ve been wrong sometimes, too. Why?
If anyone should be ashamed, it’s Bambu for not being a little more open in their communication and withholding important information.
Well, I’m not saying that there was no encouragement to subscribe. Everyone does that. But surely you don’t think he fabricated the bit about the fire hazard. Or do you?
I’d love to learn that there isn’t a problem. I sold my A1 so I feel somewhat responsible to the new owner, and I still own a PS1 which was also mentioned as possibly having the problem.
I mean that in a nice way, but you really should read some of the previous posts.
Edit: …and by the way, if your sold device is rated for 110V, it should not be affected. I just read that you posted you are: US citizen. We covered this very topic in detail a few posts earlier.
Unless you’re looking at uncertified transformers with dubious ratings intended for intermittent use, they are pushing the hundreds $/€/£. If you have the technical expertise to make one of those safe, you have what it takes to create a schematic for the board and remediate.
The problem is rather that we don’t know how the printer will react to 50Hz vs. 60Hz. You can adjust the voltage, but not simply the Hz. Setting something like that up would cost thousands of euros.
Maybe the printer doesn’t care, maybe it does. It may expect either 220V/50Hz or 110V/60Hz… or it may not care at all. I wouldn’t try it.
A few hours ago, Tom’s Hardware published a detailed report and summarised some of the information. They also asked Bambu for more information. This clarifies a number of other questions.
The statement from Bambu, for those who don’t want to open the link:
Bambu Lab Responds to Allegations
We reached out to Bambu Lab for a statement on the melting A1 printers. A representative was quick to acknowledge the issue, but insisted that failure rates were extremely low (around .052%), directly connected to preventable power surges, and were fixed in Q3 of 2025 when the NTC was removed and the board redesigned. All customers who reported problems with the power board were provided with repairs or replacements.
"While a damaged NTC may generate sufficient heat to deform or melt adjacent plastic, it does not lead to ignition or sustained combustion. As a result, the risk of fire is considered extremely low.”
Bambu Lab said safety issues are taken very seriously and that an engineering team “conducted a detailed analysis of the affected units. We observed that these cases often correlate with regions and time periods experiencing severe thunderstorms, though other sources of grid surge cannot be fully excluded.”
“After reviewing field data and considering that the inrush current of the printer is relatively modest, we determined that the benefits of the NTC-based inrush protection were limited compared to the potential downsides under rare surge conditions. As a result, we implemented a design change in Q3 2025 that removes the NTC from the circuit in newer production units.”
Your argument would have been so good without that sentence. 
I already feel lighter not having this albatross around my neck.
