Page 1 of 1

Hub screws for Wally and Simpson

PostPosted: Tue Jan 28, 2014 3:27 am
by owens-bill
GUS 1 has been puttering along with the original center hub, the one with the hub screw built in. And we'd already printed Wally's original hub, though he hasn't yet been assembled. But I like the idea of the Kossel accessories, so I printed the new versions of both.

There's a definite problem with the screws, however. Simpson's hub and screw have some wobble; I figured that once I tightened the screw against the hot end it would probably be okay, but I was a little bit concerned. Wally's hub and screw are way too wobbly. Mixing and matching, it appears that Simpson's hub has a slightly smaller hole, but the two screws are about the same.

I tried reprinting Wally's screw at 105% and that's perfect in the Simpson hub, just a little loose in Wally's (probably 106-107 would be right). There is a layer height difference; the Simpson parts were printed at 0.2 mm to match the existing prints, Wally at 0.3 mm because I was impatient ;) I don't think that accounts for the looseness though.

I know Brandon pointed out something similar in another thread (which I didn't want to hijack), but Lochemage said the fit was good. Any other observations? For comparison, my printer does most of the holes just a bit undersize, which I think is by design; for example M8 bolts fit just right, 608 bearings need a little heat.

Re: Hub screws for Wally and Simpson

PostPosted: Tue Jan 28, 2014 5:24 am
by brandonh
I had to use a ton of teflon tape to get a tight fit, and even then, I'm not convinced that the two-part hub bottom will be rigid enough for the accuracy I want. I think Slic3r is erring on the conservative side with angled volumes and not laying down filament in areas that are not 100% within the part volume - an entirely reasonable discretization choice. This would explain why threads tend to print a bit loose; both sides will be offset by nearly the layer height, in my case yielding ~0.5mm of slop. It also would explain holes for the 2mm screws for the endstops that are loose when printed horizontal but tight when printed vertical. Some kinda screw offset seems necessary.

I really don't think my printer calibration is the issue; the big Kossel is tuned to <0.1mm over 175mm, and all the 608s were heat-free press fits.

For the next hub bottom I make, I'm thinking the bottom will be a simple light press fit, with a tiny bit of intentional clearance for epoxy.

Re: Hub screws for Wally and Simpson

PostPosted: Tue Jan 28, 2014 12:22 pm
by PeteD
It's not just Slic3r. I've been using KISSlicer, and I get about the same 0.5mm slop on the hub. I found that a single layer of electrical tape wrapped around the hub screw will give me a good, snug fit.

I also get the same looseness on the 2mm horizontal holes, but I wonder if that is more due to the inaccuracies introduced by the layer heights than the slicing software.

Re: Hub screws for Wally and Simpson

PostPosted: Tue May 13, 2014 7:01 am
by brandonh
After 4 months without touching a printer, I'd like to get my 90%-there Simpson done, maybe even in time for the Maker Faire this weekend.

Has anyone had extra experiences with the GUS hub? My first (test) print at 15% infill with the integrated hub rod lasted for awhile but then broke off, and the printed one was too loose even with tons of tape, and eventually broke off anyway. I'm tempted to use the old piece and just go heavy on it. I recall a RepRap discussion with parts links to threaded studs but none affordable and long enough and with a center hole.

Thoughts appreciated.... printing the spools now; these are the only parts other than the hub that I have left. Will be switching to the 3x heavier springs when re-winding with the new spools.

Re: Hub screws for Wally and Simpson

PostPosted: Tue May 13, 2014 3:03 pm
by Nicholas Seward
I would just scale up the XY directions of the screw just a bit.

Re: Hub screws for Wally and Simpson

PostPosted: Tue May 13, 2014 5:15 pm
by brandonh
The OD of the non-threaded portion is a tight slip fit with the bearing IDs, so I don't see how a simple scaling would work here.

I have my theory about slicer discretization on angles being the root cause, so I think what I'll do is test that theory first to confirm the amount of extra gap from a given angle and layer height, then expand the screw troughs out w/o expanding the non-threaded ID.


Re: Hub screws for Wally and Simpson

PostPosted: Thu May 15, 2014 5:54 pm
by brandonh
I moved the thread profile out a bit (0.3mm) from the centerline on the hub_screw piece and modified it to only do threads on the top and bottom for exactly as long as needed (so that the bearings have a nice round surface to fit against). Came out perfect! Required a pretty heavy grip to tighten it in (a good thing), and there's zero play now. I really didn't expect it to feel so solid, being printed as two pieces, and also being printed in the weak direction (stacked layers vs a continuous plastic bead going vertical).

Got 3 new spools & the tighter springs, so it looks like GUS Red may enter calibration mode shortly.