TechPaladin Acme Nut Block

[Enraged]

I don't get how it wouldn't be printable? If anything, it's more printable, because the holes will be properly round, instead of being printed vertically.

I just made a quick model to get my idea across:

If it's split vertically, it makes assembly easier (you can have the nut off of the threaded rod, or have it assembled) and it makes a nice flat face for mounting.

If you mean non-printable because of overhangs and such, you can simply rotate your part in netfabb so it's oriented properly.

[orcinus]

Because of overhangs, yes.
But yeah, it can be rotated, you're right.

The second reason i went the horizontal route is - integrity.
If it's horizontal, the weight of the gantry rests on a *single* flat surface, with only one "joint" (the one between the plate and the nut block) bearing the load. If it's vertical, the weight of the gantry rests on two surfaces and there are two joints (between the plate and first half of the block and between the first half and the second half).

With enough constant shear applied to it, the two halves can eventually work themselves out of alignment.

Re: holes - it doesn't matter if they're perfectly round or not, you'll be cutting the threads inside them anyways.

[nismobg]

I got a quick design too for you

[orcinus]

Why put the M5 bolts and most of the mass on the bottom?
That makes no sense, all the force is on the top of the nut block, not the bottom.

(unless the render is flipped)

[orcinus]

In other news, tried the blocks in action and they work perfectly.
No opening at the seams, no backlash, no wiggling.



Unfortunately, i'm not that enthused by the Acme lead screws.
Turns out they're extremely easy to bend and were a slight bit less straight than the "stock" threaded rods out of the box too.

We'll see how they behave in print, but right now, at the extremes of the Z range of motion, they whip enough to cause binding occasionally. I'll see if i can straighten them out or at least minimize the whipping to just one end.

[Enraged]

are you not using bearing with them? if not, I can't see a major benefit to the change. The entire point is the Acme screws are cheap and accurate, and provided you buy from a good seller, they should be machined for bearings top and bottom.

[orcinus]

Well, i was hoping for:
a) less backlash during lifts
b) straighter rods
c) tighter fit to the blocks (nut in a block vs. tapped delrin)

So far, c is the only real improvement i'm seeing (the delrin blocks have developed quite a bit of slop over time).
I might try constraining the top with a bearing at some point, although i don't see what good that will do.

Constraining the top will only move the deflection caused by the screws not being straight down towards the useful portion of their length.

[Enraged]

Constraining the leadscrew will prevent it from whipping, which will keep it from bending. You could design a simple single bearing contraint and have it printed in no time, I think there is even one on Thingiverse already.

You can reduce backlash by using dual nuts with some form of spring/pressure between them.

edit: found it: http://www.thingiverse.com/thing:22217

[orcinus]

Re: bearing - thanks, will give it a go.

Backlash is fine too now.
Forgot to tighten the coupler on one side

[orcinus]

One for the naysayers:



Three months later.
Still no wiggle, still no separation nor backlash.