H2D rapid Z movement Knocking on 01.04.00.00

Bambu Lab H2 Series
21

I am a mechanical engineer. I’ve designed and built CNC machinery in the past and have a very good understanding of how these machines work.

A new sound is cause for investigation, not for concern. If you want to discuss this in engineering terms then put away the hyperbole and let’s have a logical discussion.

This isn’t a case of someone in Bambu turning up speeds for faster print times. That’s loose talk and is provably false. You can see that for yourself in the printer profile in Bambu Studio - under Motion ability you can see what all the maximum speeds and accelerations are - you compare them to previous versions of Studio and see they haven’t changed. You can even turn down the jerk if you really think there’s an issue.

Bambu has issued immediate fixes for issues in the past; they are very aware of this sound and if it were detrimental to the machine they would have done so here.

You can roll your firmware back to 1.03 if you truly can’t live with the sound. I will just say my printer is making this sound, and I’m not taking any action.

If you’re certain this is a problem then raise a ticket with Support. I recommend you stick to the facts and be nice to them.

We’re just other Bambu owners and users, I don’t understand why you are so keen to argue with us about this in the first place.

22

Who is ‘us’? I don’t even know you. You are the one that got hurt because I was mad at Bambulab. You don’t think that’s a problem good for you.

23

Can people post their total hours on printers that knock?

Is this a z axis lead screw play issue?

Any play in your lead screws?

When Bambu releases 1.04 as a download, I can try on my H2D running POM lead screw nuts.

24

I did a full cleaning, belt tensioning, and recalibration and the issue persists. I’ll try the z-hop setting changes in the slicer to see if I experience the same dampening of the noise and report back.

25

Having the same problem. Just noticed it today for the first time

26

Incase you missed it, Bambu has CONFIRMED to me via support ticket that this is abnormal noise and they are working to resolve it with the next firmware release.

28

My H2D has been running non-stop for like 4 days now printing some stuff for a friend to sell and it’s got the knocking on z-hop as well.

29

you should all do what I suggested. Raise a ticket with Bambu. Describe the situation, specifically that reverting to older firmware resolves it. Mention my ticker number US260601389001 for htem to link it all together and start building a large case.

And complain. Complain and complain.

This has become an absolute farse in my opinon. It has been months since I reported it, Several more months since others did and STILL NOTHING !

Bambu needs to feel some pressure from angry customers before they move a finger to fix this it seems.

Put some pressure on then .

30

Mine started at around 200 hrs. Its not nut wear. I followed the Nut Wear thread where you and a few others did a ton of excellent work and immeditelly I thought it was that. First resposne was Bambu sending me screw and nut set but before I put them it, I did a lot of investiagtion into it and found that my nuts are perfect and its actually something else (firmware). I replaced Z motor, did several belt remounts and retensions. tried jerking the Z belt by hand trying to replicate rapid direction changes during operation when the knock happens. Cant do it. I even removed the Z motor, put the belt on and tried the same that way just to check if there is any mechanical issue in the nuts or screws anywere. Nothing.

The knock only happens when the printers is runing on its own, during spiral Z Hop. Normal Z Hop is a lot quieter, (sounds almost as if knock is gone, but its still there, just very faint). I know that Z hops suggest its not JUST the Z axis, but it is. There is a video made by another user where he maxed out ZHop distances to exaggerate it to see when knock occurs. Its exactly at the change of direction of the bed. That video is in somwhere my thread linked below.

Aaaand - old firmware does not do this on the affected printers. There is no play anywhere. No mechanical issue to speak of. Even replacing the Z motor makes no difference. My guess is perhaps the driver board but Bambu is adamant that NO. Saying that best course of action is to wait for firmware update.

Oh, in the meantime they were sending me a few variants of gcode that runs Z noise compensation calibration with extra undocumented parameters. Those made no difference even though Bambu says this fixes things for others. They also confirmed that the fix coming in the firmware release is different to the gcode attempts they were sending me so in guessing even if gcodes did not work, firmware should ?

Having said that, they did offer to replace the printer for me … so I am not all that convinved they beleive they can really fix this. Time will tell. I am giving them the benefit of a doubdt and will wait for the firmware. The moment it out and does not fix it, I am replacing the machine.

There is more on that in my origial thread here

When I noticed it, seems I was alone. Now, especially after 1.04 for H2D, there seems to be a ton new reports of this. I find it baffling as H2D had the new motor noise compensation (and new stepper curves) since 01.03.00.00 (20260303) and from what I understand, the first reports of this stem from that firmware. Then it spread onto H2C with 01.02.00.00 (20260601) which is when I started noticing it. Reverting to 01.01.05.00 (20260211) fixes it completley just like for everyone else affected that I spoken to.

Oh one final thing. Even though I spoken to users who said they had this since April on H2D’s and did raise tickets with bambu, when I raised mine, the front line support had no clue about this being a known thing. I imagine that might still be the case, which is why I am suggesting to everyone to mention my ticket number. It took a while of back and forth and futile tests, tweaks and parts (screws, nuts, z motor), but eventually this has been passed over to their RnD team who were issuing me those gcodes to run and said the firmware will fix it. Hance, when you do raise your ticket, expect complete blank face and totally missing-the-point reply.

But do raise your ticket. Pressure, pressure. pressure. One angry customer is easy to ignore. A whole bunch screaming at them may be a bit harder to.

31

@Suchy_101 Something to check, The z axis linear rods for play up and down. Certainly sounds like play somewhere that is slapping during Z hop…

32

This has not worked in the past on much bigger issues so chances are it won’t work here.

33

all solid. no play anywhere.

the knock sounds like its coming from the motor.

My dumbass theory is that they messed up control during ZHop of the z motor. and at the direction change, perhaps due to deaccel and accel. curves - there is choppy response from the mottor which sounds like a knock, as it last miliseconds ?

Its coming from someone who knows nothing about drives, motors, electronics, physics or anything at all for that matter. SO …

34

thats just wonderful. perhaps a waterfall of printer replacement requests would make them take notice ?

35

You want a printer replacement because of a noise? While it might be slightly annoying to some, replacing a printer because of it is pretty drastic measure that I don’t think many will be willing to do.

36

well, abnormal noise is abnormal. I payed normal money, I want normal printer …

noise is never anything good - an indicator of stress. Something will fail. Prematurely, eventually. but it will.

I should not be made to gamble with my money to find out if it will or will not,

I want normal pronter that operates as indented.

37

Possible mechanism: the bed lifting off its leveling screws

Here’s an observation that might be relevant, and a hypothesis that would explain why this is firmware-dependent but still only affects some machines.

On my H2D I can lift the heatbed by hand. It moves roughly 1–2 mm upward, evenly at all three lead screws, and the screws do not rotate while doing so. It cannot be pushed down at all.

That by itself is normal and by design: the four leveling screws define the bed height as a hard stop from below, and spring preload pulls the bed down onto them. Nothing during printing pushes the bed up — gravity and nozzle pressure both act downward — so the bed normally stays seated and Z position is repeatable.

But that only holds as long as the bed is never accelerated downward faster than gravity can follow it.

If the Z axis decelerates or accelerates downward at more than ~1 g (9810 mm/s²), the bed becomes momentarily weightless relative to its carriage, separates from the leveling screws, and then lands back on them. That impact is exactly the kind of sharp, metallic knock people are describing here — and it would happen on rapid Z moves only, which matches the reports.

This would explain three things at once:

  • Why it’s firmware-dependent. The motion profile changed with the noise-cancellation feature (01.03.00.00 on H2D, 01.02.00.00 on H2C). Different acceleration profile, same hardware.
  • Why only some machines are affected. Spring preload and leveling-screw seating vary between units. A unit with slightly lower preload separates at a lower acceleration.
  • Why changing the Z-hop mode helps. A different lift type means a different motion profile, and may simply stay below the threshold.

One open point I can’t resolve myself: the slicer’s machine profile declares machine_max_acceleration_z = 500 mm/s², which is about 20× below 1 g. So this could not come from moves the slicer requests — it would have to come from firmware-internal motion (homing, leveling, Z-hop, or whatever the noise-cancellation shaping does), which isn’t bound by that value. If someone has a way to log actual Z acceleration, or can film the bed at high frame rate during a knock, that would confirm or kill this theory quickly.

Testable prediction: if this is right, the knocking should correlate with downward Z acceleration, not with Z speed. Slow moves with harsh deceleration should knock; fast moves with gentle ramps should not.

Worth checking whether others can also lift their bed, and how far.

38

I am not sure I can agree.

First, I do not believe there is any spring loading there.

Second, I jerked the z belt with my hand far more aggressively than it happens on its own during z hop, trying to replicate this at least in the slightest. Does not happen unless its the motor driving it.

Third, the noise does not come from the screws. It’s coming from the motor.

And fourth, I just did TPU print and to my surprise, it knocked twice during single z hop, once on the way down, second on the way up.

All your seemingly valid arguments are sadly not relevant nor true… Thanks for your input though

39

Oh, and PS: Bambu confirmed it to be an issue which they will resolve with future firmware update, so …

A lot of investigation has already been done by multiple users. No concrete culprit found. Bambu is refusing to confirm which part is causing the issue and are keeping it extremely vague, despite confirming the existence of the issue.

Despite direct request for a replacement of the offending part, which they normally are quick and happy to do, they kind of refused to do here, insisting that the best course of action is to wait for firmware fix.

And just to clarify, official Bambu description of this issue is „a combination of environmental and specific printer conditions”, which both means absolutely nothing and is evading answering which part is at fault or why…

Is it because they don’t know, or because truly the only fix is firmware, who ‚ knows..

40

Sigh. Let’s go through this, and you tell me which of these paragraphs are not true.

The venn diagram for sounds that you percieve to be bad and sounds that are actually bad is not a single circle.

We know that this issue appeared with a firmware update, meaning that whatever mechanical condition makes the sound was already present.

Nobody has identified any physical print issue, or any mechanical failures corresponding to the sound.

As time passes, if there was a print issue or a mechanical failure related to the sound, the chances of that appearing on machines would go up - meaning if this actually led to printer or print issues, there would be a time incentive to issuing a fix sooner.

In the past, Bambu has issued patch updates to correct issues outside of their normally planned releases, meaning they can and will correct issues when needed.

When you evaluate all of these data points together, it’s pretty clear that although unpleasant, this isn’t an issue that’s ruining prints, or ruining machines, and isn’t due to a mechanical failure. Firmware made it appear, and firmware will make it go away.

My advice to you - and everyone else concerned about this sound - is to stop speculating. If you want to make a case that this is a mechanical issue, you need to bring facts and data. Find the exact component that’s making the sound, and the folks here with a lot of years of engineering experience, myself included, will be happy to make an assessment as to if this is truly an issue or not.

I’ve never pulled punches when Bambu is in the wrong and I don’t intend to do so here, either.

41

Hey CarbonForge,

Always great to hear from you, so thanks for taking the time.

Having said that, I am not sure if your response was directed at myself or AlexE1978 ?

Either way, to address what you said a little - respectfully, kind of seems you did not fully follow all the threads, including my own, on the matter. Apologies if I’m talking rubbish, I mean no offence.

As I said, we already know the noise is coming from the motor, even though it’s not super easy to pinpoint as it resonates through the chassis and screws a lot.

Swapping said motor for a new one does not fix it. Swapping screws and nuts does not fix it. Retightening and resitting the belt does not fix it. I personally have no abnormal play in my screws. This does not seem mechanical, nor specifically motor related. I asked bambu for a new MC board, thinking perhaps there driver side is at fault, having ruled out mechanical components including the motor itself. They said no. As far as my limited understanding goes, this is the last component in the z chain that has not yet been swapped to see if there is any difference …

I have been in touch with bambu RnD team about this (well, they were in touch with me), and even they, despite acknowledging the issue and calling the noise abnormal, are not confirming what component is causing this nor providing a replacement for it. Quite the contrary, they specifically said replacing MC board will not help and said to wait for firmware… or replace the printer which they also offered…

So respectfully, I kind of feel like the big leagues, including yourself, have not gotten properly involved in investigating this, likely because your printers are not affected and there is little you can actually do remotely.

The advise we received, if I remember correctly, related to checking backlash and wear on the nuts, then same for the stabilising sliders and not much else…

The noise comes from the motor though, which seems to be collectively ignored by everyone as if it was impossible.Bambu has kind of confirmed it is the new motor noise cancellation that is causing it, having issued Gcode that runs the routine for z axis only, with extra undocumented parameters. Did not help me despite 3 different Gcode versions I received from them. Thy said this helps others.

They also said firmware fix will include a different solution to those Gcode, so even if those don’t work, firmware will.

Long story short - it’s formally confirmed to be an issue. Z motor noise compensation related. Bambu swears no part swaps will fix it, which is hard to believe, as replacing the printer will, so swapping ALL the parts fixes it, surely we don’t need ALL of them to be swapped to fix this, just one … but Bambu is refusing to say which or provide it …

So there we have it. I would love one of the smart guys to lay their hands on an affected printer and share thoughts with us, as I am sure you could tell us more than bambu does…

And naturally, you are 100% correct that prints are fine and the printers are not getting destroyed. So it seems it’s one of „ high annoyance, low impact on functionality” issue which undoubtedly has low priority for Bambu.

It is annoying though, terribly so…and can’t shake the feeling I would rather have a printer from the „not affected pool”… for the piece of mind having a unit that matches the standard tolerances and QA.