i had an issue with making my first print about temperature so i wanted to check the fan an filter at the back of the H2D, to do that i had to lower the bed, doing so i realise when i lowered with biggest setting that it hit the floor of the printer, making a noise that made me ask myself.
does the H2D know the level of the bed ? or is is a dumb machine, like if i am above the bottom at 1 cm and make the bed go down 1.2cm, does the H2D know where to stop ?
i hope so on a +2000 euros machine, can bambu confirm this ?
assuming it was leveled initially and no power loss, I’ve never had the H2D lose track of itself. I can’t say the same for other printers I’ve owned though. X1C, for instance, seems to forget after a while after printing, and it keeps prompting you to rehome when that happens so that it can remember where it is. Not really sure why this happens, except maybe it idles the steppers after printing rather than keeping them at full engagement. Less amps going through them, so then they can cool down a bit rather than stay hot.
H2D lose track after a while as well. It’s the nature of stepper motors, you cant keep it always on, would burn them much faster. Unlike servos.
But the even dumber thing is that even if you just homed your printer, due to the way the AP and MC board communicates it somehow can still go beyond the limitation if your operation is “fast enough”. Like clicking multiple times on the movement buttons or clicking too fast. It was a problem on X1C and it still presents on H2D.
In general, no it’s not gonna hurt your machine too much if it happened just very occasionally. Usually your bed would come reasonably leveled, the noise you’re hearing is just the belt skipping when it has resistance. So assuming you don’t have anything under the bed, this process actually could help your bed to level a bit by belt skipping. at the cost of a little bit wear on the z belts though However, after this incident you must redo your bed calibration and probably tramming as the leveling of the bed has been changed.
If you’re really worried, give your z axis belt & tensioner a check a couple thousand hours later if it’s not causing a problem.
as long as the sepration between mcu and application processors (actual cpus) exists, this kind of issues are hard to prevent…you can mitigate and fix one by one but cannot fool proof it from low cost junior engineers
you could do a sensor under the bed with some low electric current, if yes then stop going down, if no keep going.
putting those on a spring so it give time to the system to stop without damage, could have been an easy fix before they ship
It’s not like some printers, where the bed literally falls all the way down when the power to the steppers is lost. It could easily make a conservative assumption about how far it might have fallen and then subtract that from available z-height, thereby allowing you to move it down or up most of the way but still not ram it into the floor or top it out. But, they didn’t, and judging from the X1C, probably never will. Instead, it’s chinese logic: re-home when you’re told to and stop complaining. Meh, irritating, but ultimately not a big deal. Just a different way of doing things than maybe what a more western mentality might be.
It’s not really Chinese logic, it’s just modern machine design. Why add a limit switch if it isn’t necessary? The z axis is unlikely to damage itself by crashing into the lower limit, even if the operator can’t drive properly.
Yea just extrapolating from “confirmation of approaching end of travel” which seems to be what the op thinks should be included, beyond the loosely implemented soft limit.
The “more western mentality” might call for a prox switch near end of travel, but the only difference between that and Chinese logic, is NPN vs PNP
Actually, you raise a really good point! However, I believe it does it slowly, rather than at full speed, for homing purposes. So, as a compromise solution, it could simply move slowly (homing speed) without actually homing when it loses track of where it is. That would be a sensible improvement.
I agree that homing would be done with slow speeds and low current (low torque).
Initial positioning could be done with a higher (not excessive speed) but remaining a low torque.
As soon as you hear rattling (what others are reporting), either the belt is skipping or the stepper motor itself is skipping a step. To me that feels like unnecessary stress on the motion system that I prefer to prevent at any time.
Soft limits are often used on inexpensive CNC machines. Just one home switch per axis, but then you configure in the software where the ends of travel are from home.
Sounds like that’s partially implemented here, but you can defeat it by just clicking too fast?
However, after this incident you must redo your bed calibration and probably tramming as the leveling of the bed has been changed.
Meh, irritating, but ultimately not a big deal. Just a different way of doing things than maybe what a more western mentality might be.
