One of the variables in the current design that needs work is the Z-axis…sure, it currently produces fine results, but a loss of power has a platform crash in the Z direction…which is sub-optimal. Springs can limit the damage but that seems, generally, like something that can be improved upon. (a 2am power failure should not result in a heart attack)
Case in point: Including a rotary damper (like, but not specifially this: http://www.mcmaster.com/#6597k8/=qynk6b) would make power failure a graceful failure, as would using Leadscrews in the Z axis (expensive, with cantilever issues), or a Worm Gear (expensive, untested)
Originally I wanted a topic to handle the Z axis as it seems that’s the component of the printer that needs the most development, but really, a category that listed all deviations from the originating design could use thier own category.
my answer to this is a worm drive. manufactured ones are expensive and hard to match a set for our needs. I have a set I printed and soon as my other life allows I will give some results on how well this approach works
I’m not to where I can produce gears yet…I could see having a bronze worm and a printed gear, I’m not sure having both printed would be a strong enough solution, and there’s a good chance the layers of the printed gears might cause artifacts.
The diy acetal nuts I posted about in the 3d printer group are a possibility, with the added cost of two acme leadscrews or a leadscrew and additional rail(s)
@Mark_Hindess what happens to the RAMP in an unintentional shorting situation? The motor could hold things in place (or provide enough drag to slowly lower the platform), but I wouldn’t want it to potentially harm the controller.
Another option would be to have a fixed bed and move the extruder in the Z direction, I’m sure that raises a host of other issues. (Including doubling the vertical envelope .)
@Mike_Miller I’ve not thought it through but it should be possible to have the relay switch from ramps to shorted so you wouldn’t end up with a short at the driver.
I don’t really trust that method of driving filament or cable though. This arrangement could be done with a toothed belt driven by a worm gear connected to one of the top pulleys. There is a method for making worm gear sets for use with threaded rods: http://www.astronomyasylum.com/gears.html
Another way to use Spectra or cable would be to connect it to nut running on a screw along one of the straight runs. The belt would isolate the bed from wobble, but you would need to find a way to keep the nut (and cable) from turning around the screw.
I like the idea of making a worm gear set, but I’m leaning toward the screw and belt idea as the easiest to implement.
The issue is stopping the bed from falling on power loss. Why not a brake solenoid energized through a relay or controlled output. Power fail > brake applied …
Even with “air breaks” as @Tim_Rastall rastall pointed out many moons ago the steppers have a timeout in firmware that would need to be changed. ergo if you leveled your bed and waited x amount of minutes for it to heat or any reason and it surpasses the set time the motors will go slack and drop the bed. though it would be cushioned it would still move. a spring on the rods is an easier and cheaper solution. If you want the bed to remain stationary a mechanical change needs to be made. I have thought about a ratcheting system but that wont work for variable z heights. I honestly believe the most elegant solution is a worm and wheel set up. The weight of the platform holds a constant back pressure and if the z isnt required to make fast movements backlash would be mitigated
The relay idea could potentially damage your stepper driver if it malfunctioned, even if it was wired so that the driver sees open-circuit when it is activated instead of a short, but you could to the same thing mechanically with a solenoid mounted so that its return spring brakes the Z axis when power is removed.
Anyway, you wouldn’t need a controlled output if you just want it brake on power loss. Just wire it directly to the power supply and the return spring will activate the breaking mechanism when power is lost. If you want it to brake any time the Z stepper is deactivated, even without power loss (like on reset), you could use a FET wired to the Z stepper enable pin.
@Whosa_whatsis You could do both I suppose … set brake on power fail, or fault detect of z-axis stepper. PSU to one side of coil, ground return thru a controlled FET. Either way when the relay/solenoid de-energizes > brakes applied.
I haven’t started to build an Ingentis yet, but when I do I plan on using lead screws or possibly a single lead screw solution. I like many things about the design in general and the basic frame system was more or less what I had stewing in my head, but I have never been a fan of of the belts on the Z.
If people insist on using belts then I think the cheapest and simplest solution would be a counter weight system. Yes, it will change the tension in the belts during a print but this probably only matters with very large parts.
counter weights will eat into your build volume to outside footprint ratio king of defeats the purpose of such a space conservative design. lead screws can be good but can cause artifacts and wobble ymmv. and with the reduction of a worm gear Ibelieve max resolution can be achieved. the better the worm setup the bettes your results.
Great to see this discussion happening, aside from a general unease about using spectra on X and Y axis (mostly irrational as it works fine) the Z axis has never really been completed to my satisfaction. I have a design for a large worm drive that I will print at some point soon, it’s scaled and has a pitch that should allow for good print quality. Aside from that, I also have lead screws on the way from misumi, I’m of a mind to try them too.
The belts do give a very smooth and consistent layer height, no Z ribbing or any of that malarkey and I’m glad I tried it. I’d also be keen to see someone try rack an pinion, the Z axis is the only viable candidate for this as it’s performance won’t be particularly effected by having the Nema mounted on the bed to drive the pinion, of course, this would suffer the same fate as belts if power was cut. Incidentally, tats, exactly what happens to alfinia UP! Printers.
@D_Rob I’m pretty sure a weight of the proper form factor could be worked in between the bed frame and main frame of the printer.
In regard to z wobble with screws there are many solutions to that such as flex couplers and not over constraining them. I have also been playing with the idea of using smaller diameter lead screws and 10 mm linear rod or one of the linear rail/carriage systems. If you keep the linear motion components stiffer and the screws less rigid then you have less to worry about when it comes to the lead screws affecting quality.
