Warped bed :( seems like a common QC issue

[Matt]

@xi.2dvh

The thing is, they really can just call their suppliers and demand things to be fixed.
Bambu obviously has blueprints for all their parts, doesn’t matter if they’re made in-house or by a 3rd party.

They can, and they probably already have. But that still doesn’t mean things get fixed overnight. Especially if the part has to be redesigned, reengineered, and retested. That means new tooling and new standard operating procedures.

@xi.2dvh

The fact that they’re shipping beds like this tells me one of two things, their flatness tolerances are incredibly high/non existant, or they just don’t care about QC unless the customers make a stink about it.

Or they are still shipping parts because the majority of them do not have this issue. What they are doing is acknowledging there is an issue for some and promising to take care of it. For those that aren’t happy with this, they are giving full refunds. That’s about as good as it gets from a customer service standpoint.

@xi.2dvh

Also, when you’re having QC issues with your product, you don’t keep shipping defective parts. You put a stoppage to production/shipments until you can fix whatever is going on.

Well, that’s the “Toyota Way” to do things. But setting up a manufacturing process like that is hard and takes time, especially if certain components are outsourced. This is still a company that is less than a year old and was basically in a COVID lockdown in China for the first half of its existence.

@xi.2dvh

Believe me, if we can produce molded radio trimplates that have a flatness variation of less than 200 microns, which is something that someone driving a car isn’t going to notice, then bambu should 100% be able to produce flatter beds for a prosumer grade 3D printer.

But how long did it take your company to figure how to get to that point? How many iterations?

A perfect example of how hard it is to scale manufacturing is Tesla. When they were only making the Model S in low numbers, their build quality was excellent. But as soon as they ramped up and started producing in numbers, in multiple locations, their build quality became hot garbage. And Tesla has pretty much unlimited resources. It just goes to show how difficult it can be until you get your processes figured out and find third party manufacturers you can trust to deliver consistent, quality parts.

@xi.2dvh

Before I get called out, I’m not pulling this info out of nowhere, I know this because I work as a manufacturing engineer.

I’m not disagreeing with you, so don’t take this as being called out. Bambu Lab needs to do better on their QC and warped beds are absolutely an issue for some. Bambu Lab does needs to come up with a better solution for their current bed. I just don’t think it’s as simple as stopping production until it’s figured out or calling up a supplier and “demand” things be fixed. It’s just not that simple.

I’m not pulling this info out of nowhere, I know this because I own a manufacturing firm that also manufactures parts for vehicles.

@xi.2dvh

bambu should 100% be able to produce flatter beds for a prosumer grade 3D printer.

They should. And I think they will in due time. I think this because this printer is leaps and bounds ahead of some of the other printer’s I’ve used over the last few years. There is obviously some smart people that designed, engineered, and built these machines. That’s why I’m guessing these QC issues are in the manufacturing/supply chain. I would guess this is the weak link in their experience/skillset

[anasazi37]

Short version: I’m using a simple tape-based workaround for now and it’s giving me…acceptable results. I’ve also told BL, in engineering terms and with an accompanying diagram, that they can’t “software” their way out of this situation. Read on if you’d like to know more.

I discovered this issue with my P1P about two weeks ago and thought I was going crazy. Up until then I was printing stuff raised above the bed on supports, so I didn’t notice that anything was wrong. I recently needed to print two long, large, flat rectangular pieces and oriented them left-to-right on the bed. The pieces are two halves of something larger, so the bottoms should be flush with one another after printing because they are meant to be joined together. I tried all the usual things to resolve the issue, but kept getting warped results. I wasted a lot of time, material, and sanity in the process. I found this thread and became understandably upset that my $700 P1P can’t do what my $400 Creality CR6 SE can do in its sleep.

I submitted a support ticket, including log files and a picture clearly showing the bed warping (at least 0.6mm downward deviation in the center, compared to the sides) and a BL representative got back to me several days later. They said that the deviation was within acceptable limits (it’s not, because my CR6 can easily get to +/- 0.1 or less), but did offer to send me a new bed. I’m not optimistic that the new bed will be any better, though, and communicated that to them. They said a software update with a “more advanced” bed leveling algorithm should be available in a few weeks. I told them that they can’t solve this with software and provided the following diagram to explain why.

Unless I’m missing something, there are really only four options in software that they can try, and none of them are good. The first removes material from the printed object, the second warps the object, the third requires removing excess material after a print, and the fourth requires some kind of variable manipulation of the lead screws to “unwarp” the bed in real time. That last one could put a lot of strain on the bed and potentially make things worse. I haven’t thought through the physics of the last option, but it seems overly complicated and risky.

While I’m waiting for a reply from BL, I decided to try the aluminum tape option discussed some time back in this thread. I carefully applied multiple layers of standard 0.07mm HVAC aluminum tape, starting in the center and checking the level with a metal ruler to make sure I didn’t use too much. I then worked outwards to the left and right, continuing to check with the ruler. Where there were some slight drops between the big strips, I cut the tape into thinner strips and filled them in. Not a perfect solution, but now the deviation across the bed is around +/- 0.1mm. I created a simple test part that is 240 x 50 x 10 mm (last dimension is height) and oriented it lengthwise on the bed to check how good this solution is. I did a crappy job of removing the brim because I was in a hurry, so please excuse that, but you can see that apart from some very slight deviations that correlate with the tape height (again, not a perfect method), the surfaces are pretty flat. More or less good enough for what I need to do in the near term, but I want a long-term fix ASAP, like all of you.

I should not have to do this with a $700 printer that has so much amazing engineering that went into its design. I think this bed warping issue represents significant flaws in both BL’s design and manufacturing processes. My best guess, and this happens on almost every complex engineering project that I’ve been involved in, someone at BL made some technical assumptions about the bed and those assumptions were not adequately tested before and after manufacturing began. The important thing is that they own the mistake and fix it, and it sounds like that is happening. If the people behind the printer are smart enough to make a machine this impressive, they will be able to figure out a long-term solution.

What I don’t want to see happen is that early adopters across the X1 and P1 lines are told that the solution is to get their next model with the new bed design. If that happens, and we don’t get an offer to replace our current printers (best case) or a massive discount (worst case), I won’t do business with BL again. I’d rather stick with my $400 slower printer.

[xi.2dvh]

You hit the nail on the head there.
I see a lot of people here defending Bambu by saying they’re a new company and they’re still learning. But the thing is, there SHOULD be control plans to prevent parts with such poor quality; and even in the case that they got a bad batch of beds after initial kickstarter shipment/website sales, there should also be contingency plans on how to keep shipping printers with good beds (you usually keep a surplus of good beds that you can use to build beds in case things go bad with suppliers so you have some time to fix it.)

It’s like you said, obviously they have some really good engineers behind these projects doing some really complicated stuff. So it baffles me that they can produce a printer that is leaps and bounds beyond anything on the market, but they somehow don’t have what equals to manufacturing 101 quality control.

They need to make this right and stop hiding behind what seems to be non-existent deviation limits (since there’s people with 2.0mm of warpage behind denied, but some with 0.5 mm being approved for replacement) and make this right.

[ThanksForAsking]

@anasazi37

I’m using a simple tape-based workaround for now and it’s giving me…acceptable results. …

I started doing this very thing, having done it on an earlier printer. I’m running .93 mm gap @ center. It was a big enough gap that I’ve ordered some thicker tape than what I have and I’ll re-do it since I tried something ‘clever’ which didn’t work … LOL.

I still have to submit my ticket.

@anasazi37

someone at BL made some technical assumptions about the bed and those assumptions were not adequately tested before and after manufacturing began.

That may be, or other factors we know nothing about occurred. You seem to have enough of a background to understand exactly how many things can go wrong in manufacturing.

They also may have received samples with small errors that were fine, nad with the ABL’s correction window, excellent. Then got screwed when the main shipment arrived. All is speculation at this point. {Sigh} But a big thanks for the pics and the work in detailing this for others (besides you saved me the having to do the post LOL).

@anasazi37

I should not have to do this with a $700 printer that has so much amazing engineering that went into its design.

Yep and I shouldn’t have an >$1500 investment in the same boat. If they send you a new bed and it doesn’t work out they have made it clear they will refund people. That’s all you can ask of any company, particularly a startup, eh?
And yes, them being a small startup gives them a hall pass. The rest of the printer is so thoughtfully built, and at a price point that shames the existing ‘big guns’, I’d also suggest that we all need a hall pass from time to time. I know I sure have.

@xi.2dvh

see a lot of people here defending Bambu by saying they’re a new company and they’re still learning. But the thing is, there SHOULD be control plans to prevent parts with such poor quality

And you nailed it on the head too, but in real life sometimes the most brilliant engineers are not the most brilliant business people.

For all our wishing they did this, or they did that, the fact is we have no idea of the actual issues, the order those took place, the players involved, on and on. Armchair quarterbacking. The solution to problems are always easy if you’re not in them.

Just for the record - I am just as absolutely PO’d at this whole thing as you or anyone else here. But pitchforks and threats to never buy again is just noise, and accomplishes nothing.

If you read the long posts from some of us suggesting a calmer discussion (which is not condoning it at all!), we have been having discussions of what could be the problem.

Posts like @anasazi37 and others have been temp solutions to help others. More than a few people discussing the possible causes may just jog something that results in a fix, and that’s good for all of us. The 3D industry has been built to where it is today in just this manner. Bambu put together lots of ideas from many places, cherry-picking the best to make these machines. Overall, they nailed it.

I know that I got out of the box amazing prints that blew my and my Wife’s mind despite the bad bed. $10 worth of alum tape and a few minutes virtually solves the problem for now until proper beds are ready. If that’s too much for you personally, they will refund you.

I’m no stranger to bad beds, I have been at this since late 2015 and have yet to buy a printer that did not require some serious work on the bed. I actually straightened a CR-10S (310x310) bed that was as wavy as a washboard. Manually. Deadblow hammers may have had some role, I’m not sayin. So trust me, I’ve been here before. Was ANY of the manufacturers willing to refund me? Nope. Not one. Replace the bed? Nope. Not one.

So I’m not absolving Bambu in any way, but b*tching about it here does nothing to help fix it, now does it?

[msinger]

@xi.2dvh

So it baffles me that they can produce a printer that is leaps and bounds beyond anything on the market, but they somehow don’t have what equals to manufacturing 101 quality control.

And it’s equally as baffling to see some attacking Bambu Lab and assuming they should have put plans in place ahead of time to resolve issues (warped beds in this case) that haven’t yet occurred. So crazy.

[xi.2dvh]

I think as paying customers we all earned the right to complain when a product we receive is defective. Why you feel the need to defend a multi million dollar company for free is a topic for another day.

It’s not crazy to think they should have plans in place to prevent these things, PFMEAs and DFMEAs are a thing that gets performed in pretty any product before launch, and they’re something that engineers learn about while in college.
And if not, then root cause analysis is a thing, and the simplest solution would be to put a stop to shipments, identify faulty beds in already assembled printers and replace them at the factory/warehouse. The customer shouldn’t be the ones left to identify the issue, go through troubleshooting and then be expected to take 2-3 hours out of their time to replace the bed with something that might not be acceptable again.

I obviously don’t expect the printers to be perfect, I don’t think anyone does. A loose belt or backed out screw that takes minutes to fix I can excuse, the bed being such an integral part of any 3D printer taking days of communication and hoop jumps with customer support to receive a replacement is inexcusable.

[msinger]

@xi.2dvh

Why you feel the need to defend a multi million dollar company for free is a topic for another day.

I’m not defending Bambu Lab at all. Just pointing out obvious things that some don’t think are so obvious (for whatever reason).

@xi.2dvh

It’s not crazy to think they should have plans in place to prevent these things

It sounds a tad crazy to me. I’d rather a company focus on development than spending all of their resources making plans to fix thousands of issues that will likely never occur.

@xi.2dvh

The customer shouldn’t be the ones left to identify the issue, go through troubleshooting and then be expected to take 2-3 hours out of their time to replace the bed with something that might not be acceptable again.

Welcome to the world of 3D printing.

@xi.2dvh

And if not, then root cause analysis is a thing, and the simplest solution would be to put a stop to shipments, identify faulty beds in already assembled printers and replace them at the factory/warehouse.

And I guess you’re assuming everyone (or nearly everyone) has a warped bed? If that was the case, then, yes, that would likely be the best route.

[xi.2dvh]

@msinger

And I guess you’re assuming everyone (or nearly everyone) has a warped bed? If that was the case, then, yes, that would likely be the best route.

My guess is that this a more widespread issue than we know. The thing is, the majority of the Bambu user base probably won’t notice the bed warpage because they use consider the printer as another cool gizmo they can play around with and make silly little things.

I think Bambu is suffering from Tesla syndrome. Demand for product increases, the suits push for manufacturing to keep up with sales and quality suffers because compromises have to be made to ramp up production in a short amount of time.

But again, these are just educated guesses.

[GuyH77]

Someone on a discord I use has an early X1C with flat bed, they recently received their second X1C which has this warped bed issue. Clear QC problem. I want to order 3 printers for work but won’t until they resolve the supplier quality

[msinger]

@xi.2dvh

the majority of the Bambu user base probably won’t notice the bed warpage because they use consider the printer as another cool gizmo they can play around with and make silly little things

Not that Bambu Lab printers are by any means expensive. But, are most users really spending this much money to buy a printer that cannot print objects with a flat bottom? There are so many other less expensive options out there for people who have low standards.

[xi.2dvh]

@msinger

But, are most users really spending this much money to buy a printer that cannot print objects with a flat bottom? There are so many other less expensive options out there for people who have low standards.

That’s pretty much my point, a printer at this price point should at least be able to print flat parts without them coming out like a banana.

[ThanksForAsking]

@xi.2dvh

But again, these are just educated guesses.

Well, they are guesses at least. I think I pointed that out earlier. You, and I and everyone else here do not have enough data to know what’s going on, how many are affected, etc.

I’m not belittling you at all, but I have to disagree that anyone here is “making an educated guess” no matter how much we know about printers.

I don’t think a reasonable person would think that Bambu is doing some sort of coverup, nor do I think they ordered thousands of beds that they knew would be defective. I don’t think you do either.

Having been involved in this issue right from when it started being reported I can tell you that the number I’ve seen is actually fairly small overall, but the noise level is high because people are outraged.

[Matt]

@anasazi37

Short version: I’m using a simple tape-based workaround for now and it’s giving me…acceptable results. I’ve also told BL, in engineering terms and with an accompanying diagram, that they can’t “software” their way out of this situation. Read on if you’d like to know more.

Thanks for this detailed post. I’m with you that their proposed solution of a software fix isn’t going to cut it. It might be a temporary solution, but I really don’t think it’s good enough, especially considering how warped some of these beds are.

@anasazi37

I should not have to do this with a $700 printer that has so much amazing engineering that went into its design. I think this bed warping issue represents significant flaws in both BL’s design and manufacturing processes. My best guess, and this happens on almost every complex engineering project that I’ve been involved in, someone at BL made some technical assumptions about the bed and those assumptions were not adequately tested before and after manufacturing began. The important thing is that they own the mistake and fix it, and it sounds like that is happening. If the people behind the printer are smart enough to make a machine this impressive, they will be able to figure out a long-term solution.

Agree with all of this.

@anasazi37

What I don’t want to see happen is that early adopters across the X1 and P1 lines are told that the solution is to get their next model with the new bed design. If that happens, and we don’t get an offer to replace our current printers (best case) or a massive discount (worst case), I won’t do business with BL again.

Agree with this too. Even though my bed is 100% flat now, I have concerns that it will remain flat long term.

@xi.2dvh

I see a lot of people here defending Bambu by saying they’re a new company and they’re still learning.

Just in case this is directed at me, I’m not so much defending Bambu Lab as much as I am defending the reality of the situation. People are upset that their beds aren’t flat, and that’s totally understandable. BL has a hot mess on their hands that they absolutely need to fix.

But being mad and demanding they fix things overnight just isn’t going to happen. It’s not going to get them to work harder or faster and magically come up with a solution. That’s just not how these kinda things work. This is a complex problem that is going to require complex solutions.

@xi.2dvh

It’s like you said, obviously they have some really good engineers behind these projects doing some really complicated stuff. So it baffles me that they can produce a printer that is leaps and bounds beyond anything on the market, but they somehow don’t have what equals to manufacturing 101 quality control.

Because it’s not really manufacturing 101. If it was, every company would have no problems with quality control. But the reality is, EVERY single company runs into these sorts of problems. It’s just a lot easier for companies with loads of experience and deep pockets to pivot and fix them faster.

This is very likely the first time they’ve had to do something of this nature, and they’re having to do it while trying to scale. They’ve got to hire hire and train new people to get them up to speed on their hardware and software. They probably only have a handful of people who have been working on this machine from the start, the rest will take some time to get up to speed.

@ThanksForAsking

Just for the record - I am just as absolutely PO’d at this whole thing as you or anyone else here. But pitchforks and threats to never buy again is just noise, and accomplishes nothing.

If you read the long posts from some of us suggesting a calmer discussion (which is not condoning it at all!), we have been having discussions of what could be the problem.

@ThanksForAsking

I’m no stranger to bad beds, I have been at this since late 2015 and have yet to buy a printer that did not require some serious work on the bed. I actually straightened a CR-10S (310x310) bed that was as wavy as a washboard. Manually. Deadblow hammers may have had some role, I’m not sayin. So trust me, I’ve been here before. Was ANY of the manufacturers willing to refund me? Nope. Not one. Replace the bed? Nope. Not one.

Same. My last printer had a bed that took work to keep flat. For sure did not get offered a refund. Was basically just told to deal with it. And I did. As @msinger said, tinkering with 3D printers kinda comes with the territory.

@xi.2dvh

I think as paying customers we all earned the right to complain when a product we receive is defective.

You absolutely do. I don’t see anyone saying you shouldn’t complain if your bed is giving you problems. Bambu Lab certainly isn’t, they’re giving you the right to complain on their forum. And they are giving you several options, including a full refund if you aren’t happy with the bed or the ambiguity on when a legit replacement bed will be available.

[anasazi37]

Dang, I didn’t want my “here’s a quick and dirty fix” post to spark this kind of discussion. Sorry, everyone. Just wanted to share that I got a functional, completely reversible hot fix in place and maybe give everyone impacted by this issue a little hope and a path forward until we find out what the long term fix is (and when we might get it). I am frustrated, though, and I think a lot of that is due to the lack of communication from BL on what the issue is, plus the inconsistent responses from their customer support team. Even if BL hasn’t definitively nailed down the source of the issue, it would be nice for them to engage with their user community and say something about it. Uncertainty and silence can drive people crazy, and it makes it more difficult to build your brand.

Looking back on the past week or so of posts on this thread, I think I now have another idea on how to do something even better than just tape. If you completely cover the plate with thermal paste, level it off with a metal straight edge similar to how you’d do the same with a 2x4 after pouring a concrete slab, then apply a single layer of aluminum tape (or a thin aluminum sheet cut to size) to cover that, you’d theoretically have a very level surface with less heat loss than multiple layers of tape. Plus the layer of aluminum would keep the paste away from the build plate, resulting in easier post-print cleanup. Not sure when I’ll have time to try this out, but wanted to throw the idea into the mix…

[Matt]

@anasazi37

Dang, I didn’t want my “here’s a quick and dirty fix” post to spark this kind of discussion.

Don’t be sorry! Temporary solutions to this issue are great and your post was excellent. They help big time because, as you say, a permanent solution is gonna take some time. The last thing we want is for them to rush something out and have it still not fix the problem. People who had a warped bed and were sent worse replacement beds were probably even more frustrated than they were at the start.

@anasazi37

I am frustrated, though, and I think a lot of that is due to the lack of communication from BL on what the issue is, plus the inconsistent responses from their customer support team. Even if BL hasn’t definitively nailed down the source of the issue, it would be nice for them to engage with their user community and say something about it. Uncertainty and silence can drive people crazy, and it makes it more difficult to build your brand.

Totally, it sucks. And it needs to get better. Brand communication when you’ve made mistakes is tricky, especially for new companies who haven’t dealt with an issue like this before. Social media and the drama that can come with it makes it even trickier. Plenty of large US firms have botched their responses to problems way worse than this one.

@anasazi37

I think I now have another idea on how to do something even better than just tape.

This is good and will probably work well.

[Sayenah]

Checking this lively discussion after nearly a week and LOVING your approach.

Help me a little with your last post. Are you recommending applying the thermal paste on the black magnetic plate itself?

Your idea then is to level it off using a metal edge ruler, and then placing aluminum tape on top of the thermal past. Am I getting this correct?

[DHK]

I am planning to do something similar with thermal glue or silicon so the tape is not needed and it can be removable. I do not think thermal glue (or paste+tape) will have significant advantage over silicon. We have thick rubber magnetic plate which is not a good heat conductor itself and 0.5mm of silicon probably will not make much of a difference over thermal paste.
Have to figure out the other issues first though.

[dana.nelson]

Something like silicon or glue that would cure solid sounds like a good approach. Thermal paste could easily squish out - the build plate exerts a significant force when sucked down from the magnetism and would probably be inclined to follow the curvature of the bed, and I would not think a little foil tape would keep the paste contained.

A quick search for thermally conductive glue turns up a lot of options like this:
https://www.amazon.com/dp/B072MSXHJD

[ThanksForAsking]

@DHK

I am planning to do something similar with thermal glue

Hmm … interesting thought. What sort of thermal glue are you thinking about? Removable? I’ve not messed with thermal glue. Gallons of paste over the years tho’

@DHK

which is not a good heat conductor itself and 0.5mm of silicon probably will not make much of a difference

Two days ago I had an idea and tried silicone tape (thicker) for the center area rather than multiple layers of aluminum since I had thin alum tape.

It worked well but was uneven. The silicone tape was tapered at the edges, gave me ripples. Overlapping the edges worked but was very hard to get even since it instantly adheres when touched to itself.

I saw the problem, but finished it just for research. I got a fairly smooth bed (certainly workable) but wasn’t happy with it.

I almost tried again with a silicon sheet I have from another project, but I decided silicone is too much an insulator and ordered some thicker alum tape to go with the thin stuff I have.

I ran some temp checks (it was fine up to 90°C) but it lagged behind a lot in heating and settled a full 10°C lower where the silicon was from the rest of the bed. Too much of a loss & too uneven. A few degrees, fine. 10°? Nah. Silicon sheet or tape is out. here’s why -

It causes the heater to work too hard, and there’s already a bug report of uneven surfaces when the heater cycles during some prints. I think BL didn’t allow much overhead in the power buss. Perhaps before they changed power supplies to Meanwell, which is an awesome unit (or it may be why they did). In any case I don’t what to push the supply, the bed being the main power hog.

This is something that bothers me with the “put a glass plate above the existing one” idea. Some worry it increases the Z mass and may cause artifacts, but the Z moves so slowly that’s not an issue.

Heating is an issue since it now has to heat the entire extra mass - longer warmup and works the supply hard. On the flip side the extra thermal mass probably reduces bed temp fluctuation which is a plus.

We just don’t have the power data to tell how much an extra load will affect the supply. Wire diameter and length, bed wattage, power supply overhead, etc. They engineered it for their heater and bed. It gets complex. We’re changing the design and hoping the change doesn’t cause a problem.

So I’m hoping my thicker tape comes today. But I like the thermal glue idea.

Tape on top of paste …? Rather suspect that would be damn messy. Paste and thin alum sheet, nice but $$ and I don’t know a source for thin accurate sheeting. Can’t tap AliExpress for that you’d get something bent in half in shipping LOL!

[ThanksForAsking]

@dana.nelson

A quick search for thermally conductive glue turns up a lot of options like this:
https://www.amazon.com/dp/B072MSXHJD

Yeah, nice spot. People also say they’ve removed it without much issue which is a plus.

This caught my attention tho’ - * Recommended coating thickness: 0.1-0.5mm, the thinner the better.

Since my bed is .92mm dip, that’s twice the recommended thickness. Not sure if that would be a problem.

Still, I like this whole idea, it’s an elegant and accurate solution.