Belt driven Z axis as an improvement?

Just reading the list of new and improved features of the new Lulzbot Mini 2 and noticed they say that a belt driven Z axis is better??

I thought that the Acme rods are the most reliable and give the most accurate movement especially over belts. I know they are slower than belt but I really don’t think speed is an issue on the Z axis.

Am I wrong on this, thinking that going to belts is actually a downgrade not the upgrade they claim? Seriously want to know, maybe I have mis informed on belts vs acme??


:slight_smile: I’m glad you asked, 'cause I’ve been wondering about this myself.

Seems to me that belts would add some common problems we see on the X and Y axis that we never had before on the Z – things like belt tension needing regular attention, and then there’s the occasional skipping of a tooth (just consider what that would do on the Z if it were to happen). Repeatability would seem to be an issue to me as well – as the belt warms won’t the tension drop a bit, and the result would be dropping of the Z axis as the print progressed, ending up with slightly thinner layers as the printer warmed up. And then, what happens when someone smacks the power button in an emergency situation – won’t the weight of the print head and X rods result in the entire gantry plummeting down onto the print (or worse, onto the bed surface itself)?

They were looking at geared steppers in the past. That would allow proper resolution for a belted z axis. They have also added a custom relay board that locks the motor position when power is lost. So, the Z isn’t going to fall when you power off the machine. The z axis isn’t in constant motion so, I don’t think heating will be a problem. Lulzbot uses quality belts so frequent tensioning isn’t an issue. I have Taz 5 machines with thousands of hours on them that have never required belt tightening.

The main advantages of belt drive z are that the z can move through its range much faster during homing and non-print moves and there is no leadscrew to apply pressure to the X axis during the print. A leadscrew that isn’t perfectly straight can cause z banding on the sidewalls of the printed part as it pushes the x axis around during its rotation.

I suspect this will be a good change but, time will tell. If it works out, I hope they go ahead with the current Taz 7 design that includes belted Z.

A belted Z completely eliminates Z wobble. You only have the machine constrained to the bearings then. It also eliminates an expensive part of the machine (twice in the case of the taz) The downsides will be belt tension, fraying belts potentially, Risk of a belt anchor failure causing a catestrophic print failure like can currently occur on Y or X but very rarely occured on Z, etc. With the Einsy style rambo controllers, they may be able to add a self layer tramming feature as well, which would compensate for one of the belts being off a step. Thats not currently possible on the 6 leadscrew setup with the current rambo board.

I’ve toyed with the idea of belted z before, and its a solid concept. Head crashes were always my concern, but a geared stepper pretty much eliminates that. You could also add an Z/X axis counterweight later to balance the system if they find that layer transitions are too springy with the belts. I think it will be a good setup, and a less complex option to running an anti-wobble on a leadscrew.