Problem with Z axis Helical Bellows type Couplers on Lead screw to stepper motors especially with direct drive extruders. The weight of the x axis travelling from one side to the other causes these couplers to expand and contract thus changing Z axis height during bed leveling and printing. A solid coupler is not an acceptable solution has anyone tried the 3 piece couplers? My temp solution was to push the lead screw till it bottoms out and reconnect to stepper with one problem being too short for the top frame seems to work no wobble in prints ? any better ideas short of linear slides
This is why simple tube couplers are better than the fancy helical ones.
Oldham coupler, or a bearing block above it that preloads, or a flexible tube coupling like whosa said.
You can also find a 5mm bearing, stick it in there between the shaft and the screw, and bottom it out.
Try inserting a bearing ball (single ball) between your stepper shaft and your screw. It will still carry weight down while giving you more Z height
So the problem is real made it a real pain to level the bed thx all
@Whosa_whatsis is this happening because the screw is seeing side-loads? I don’t understand what is going wrong where the coupling does ANYTHING with X moves…?
@Ryan_Carlyle First, it’s a mendel-style bot of some flavor, with the X bridge (which carries the extruder) lifted by the two Z screws. What he’s saying is that as the X axis (which carries the weight of the extruder motor) moves from one side to the other, the difference in weight distribution is enough to change the compression force on the helical Z couplers (which are essentially springs). Personally, I suspect there’s also some bearing resistance coming into play, and the weight of the motor is assisting the tension on this spring tension in pulling one side down against this resistance. It will be a bigger issue while probing the platform, with the way the Z axis moves down and back up for each probing position.
If he’s saying that solid couplers aren’t an option, they’re also probably fairly bad (or at least poorly-aligned) screws.
Of course, with the helical couplers (and this is probably my biggest objection to them, personally) they will also act as torsion springs, with the rotational force of the motor fighting some non-zero and difficult-to-predict (and probably not even consistent, considering how much it relies on lubrication) amount of rotational friction as well as the friction on the linear guide (and with the discrete stepping of stepper motors, you never really know if this is static or dynamic friction).
When you model a load as purely inertial, they seem to make sense, but when you factor in friction and (in the case of a vertical axis) gravitational loading, they just seem like a horrible idea. A 3D printer’s Z axis is pretty much a worst-case scenario for these couplers. Platform compensation algorithms that keep the Z axis moving slowly most of the time complicate matters, but they don’t create enough Z axis motion to mitigate these problems, and may actually make matters worse.
Honestly, I’m becoming more and more convinced that any multi-screw system, even using timing belts to synchronize them, is a really bad idea, which is why I’ve been exploring alternatives for bots where a single-screw cantilever is not ideal.
@Whosa_whatsis the shifting weight thing makes sense. I have always heard that you need to make sure the shafts are touching inside the coupler to prevent that.
One problem with perfectly rigid screw couplings is that they are more prone to resonance. When you get into quad-stepping on long Z travel moves, the resulting effective 1/2 or 1/4 microsteps can allow mechanical resonance to mess stuff up. Any kind of flexure mode in the drivetrain slightly decouples the masses so the resonance gets either damped or broken up.