That is an even more fails safe than anything electronic.
Well, maybe we just test the Motor power driven way and see if that interrupts the driver as Dan warned about.
And test the mains power only as the E-Stop.
If it needs an even smarter version, this can fall under advanced user/mods/20lb bed area.
1 Like
Yeah those smart ones I think they are trying to hold it with zero loss of position or something. In my mind If I lose power my parts start poppin off the bed in 5-10 minutes depending on the outside temp I am no longer looking into power recovery. I don’t see it being worth it. So I just want to prevent popping boards if a bed drops that fast. I assumed (wrongly) there was protection and a fall was not actually a big deal.
Very true. That is in my ending gcode. So yes you are covering a power outage which is good. Now what about a klipper SAVE_CONFIG? I do my best to remember to drive Z down before I do it but sometimes I still forget, and end up with a burnt SKR…
OK I’m late and just read the other comments below this one.
Firmware edits are stupid simple with klipper so that doesn’t bother me any. I just need to figure out a way to always remember to drive to Z max before I hit anything that will disrupt power. I already took M84 out of my end gcode and I have the idle_timeout turned off. Its just dummy proofing it from ME lol
Personally wasn’t able to prompt ChatGPT to suggest something different, but viable…
If your bed ends at Z max for it to be anywhere else you would need to do a save config after a power outage to catch your bed in the middle of travel right?
It should always end at Z max, How are you ending up in the middle of travel all the time? I am not sure what we are protecting against I guess.
There are likely cunning solid-state ways to do this, too, which would be smaller/cheaper and easy to integrate either onto the control board or as an add-on board that could live at the control board, at each stepper etc. The relay situation is a pretty easy one to handle, though.
I doubt it will ever stop the stepper dropping the plate, unless there’s a huge amount of friction in the system. The braking in the case of the stepper is created by the shaft rotation generating a current that opposes that motion. The resistance in the coil means the current won’t ever be enough to fully stop the stepper and it’ll slowly creep downwards. If the shaft isn’t rotating then the braking torque from the coil is zero.
1 Like
Any time you make a nozzle change and need to reset Z, that’s a save config. That you just finished with the bed all the way up. And my absent minded ass will forget that every time
That’s just one thing off the top of my head because that’s what killed the second SKR. First one was due to idle timeout and I’ve already fixed that one
Woo Hoo!
Good decision!
Good decision!
I’ll study more about your experiment. What really needs to happen is that when the TMC2209s are not enabled (and thus holding the motors), there needs to be a shunt across rthe Z stepper coils so that they have enough resistance not to crash thunderously into the bottom of the printer, nor generate enough back EMF to blow up regulators or drivers in the controller. A decent shunt should do that.
Isn’t that where the V1 store could offer standard parts. For example, pre-cut side panels, which more or less force the standard 300x300x300 to be square and properly aligned when assembled.
It’s awesome that Mike found that, I saw the Lulzbot part in one of the other threads. Some commentary: The lulzbot part is for thier mini, which uses dual Z axis motors with a single stepper driver. The Lulzbot part relay shorts the drivers when the machine isnt powered. It wouldn’t prevent a bed drop when disabling the TMC2209s as an example.
Relays are good for a shorting application, but sucky for responsiveness. Ideally you want something fast like a FET so when the TMCs are enabled/disabled the drivers never see the shorted coils. The problem with FETs is they have leakage currents, which might (MIGHT) mess with the TMC’s current sensing. WIth a firmware change, one could have the enable preceeded by a GPIO command to turn off the relays, and similarly a GPIO to assert the brake immediately after disabling the steppers.
I do believe whatever you do needs firmware support.
A geared down belt drive sounds interesting as well, mainly because it is simple- but I wonder if the trade between firmware complexity and mechanical complexity favors doing this with firmware and a shorting relay set. You could roll your own 3-motor relay shorting board fairly inexpensively. Maybe even set the standard for Z belt printers.
Also, need to have some kind of bumper at the bottom so even if the heavy bed does come down it doesn’t crash into the bottom of the machine. That’s just not good to have happen.
Any three points always makes a plane. Four-point leveling sucks. Just an observation from the dozen or so various 3D printers I’ve had or maintained. If you do this, it needs to be all handled in firmware whether it’s a flying bed or a flying gantry.
Probably not good enough to just do it with a separate power supply. You need firmware support as described above. Consider what happens when you enable the drivers while the relay is still shorted. Enables are fast relays are stupid slow.
Not with a modern hotend like the Revo.
2 Likes
What about moving the motors off of the bed? keeping them at the bottom and clamping the belt top and bottom to the z carriage? that will at least take some weight off of the bed. Sorry I know I’m skipping around but had to say it while it was on the brain
Also wouldn’t need a special looped belt for this, just cut the length you need
It is almost convincing. If we were not already diving into most those topics I would believe it.
Two leadscrew Z suggestion leaves out hardware leveling a plane.
That is what I am shooting for, that magnet through a copper pipe type drop…yeah that kind of smooth!
So like a nozzle offset save kills the stepper power? Equivalent to a marlin Baby stepping save? that sucks. I can;t see anything other than parking it helping in that scenario, or tieing into the enable pin of the stepper.
So a 3.3v relay instead of a 24v one I think that would be fine if they make them?
1 Like
I have a bunch of 5v ones here that function fine at 3.3v, or seem to at least. I can test more if you would like
1 Like
If we think of it in terms of your 10lb 300x300 bed, a stepper is .8 pounds (for the giant CNC steppers we use), think about the weight of your bed at 500x500. I don’t think it is worth the extra hardware and belt.
Absolutely! Mine will be here Wed. I would wire one to the enable pin and see how if functions before adding the stepper in.
I guess that is if enable is always high and not just a digital signal. You can check Vmot as well, Not sure if those are are always hot or not.
Enable is not always high. We know of several cases that turn off the enable:
Updating klipper or power cycling most firmware.
Ending Gcode in many slicers turns off the dirvers.
1 Like
Ok but looking at my bed at 10lbs, that brings it down to 7.6 lbs. That’s 24% lighter, nothing to shake a stick at. and if we are figuring my size to be the “standard” then I would fall into the majority then. Yes it wont make much difference to the big big ones but for the majority it would. And how much hardware are you thinking? most of that stuff is real cheap and your buying extra to get the amount you need anyways. and belt… I have enough extra belt here to probably wrap all the way around my house. again something we aren’t going to buy to the mm of what we would need so would probably already have enough anyways.
Unless there is something real expensive in there that I’m just not thinking about I don’t see the hold up
1 Like
That’s where using macros in klipper is nice, and you just set what you what in the slicer. You can do the same for marlin as well. I had it that way before i switched over to klipper
Sounds really cheap!?!
Bummer, but not the end of the world. That does make sense, it would need to be faster than gravity.
We are getting closer I think.
1 Like