CAD: Move flap thickness slop to spool slop #114
Conversation
This increase was a mistake. The difference in thickness (0.2 on either side) should have been added to the *spool* slop instead of the flap width slop, as the inner width of the spool (flap_width_slop) was not affected by the thickness parameter, while the outer width of the spool (spool_width_slop) *was*. This commit corrects that oversight. This reverts commit a68fc98 (squashed as part of 7be69db).
|
On my test display the spool will sometimes stall while rapiding. It appears that this is caused by imperfect assembly of the spool and the flaps sliding sideways, causing them to get caught on the upper corner of the front panel. I've traced the issue back to this change and I'm hopeful this PR will fix the issue. |
Nudges this from 4 mil to 6 mil, just for that little extra clearance to avoid binding. (And for peace of mind)
|
Thanks for following up on this @dmadison and good catch! Do you have any other planned tweaks or changes? If not, I'll go ahead and order a test of the latest 3d designs soon to validate for a new release. |
|
I should have at least one more tweak to make. My own enclosures just arrived so I'll try to unpack and test them this weekend. |
|
Updating for posterity: the original change was not a mistake. The Nevertheless! Although the comment and reasoning behind the change are wrong the change itself is still correct, as the spool is now tighter on the flaps and should prevent them from sliding out of position and causing the motor to stall. This matches the original design intent of having 0.1 mm (now 0.15 mm) clearance. |
On my latest order from Ponoko (3 mm acrylic) the spool struts are still loose and will need to be glued. This increases the interference a bit more (+ 0.04 mm) to hopefully reach a point where the struts will fit into the spool and hold there by friction alone. Also increased is the flap width slop, in a reversion to the change from #114. On my order based around f4e7714 the flaps do *fit* and swing but it's tight - a little bit of misalignment with the spool end pieces or burrs on the parts will cause them to bind. And after gluing the end pieces in place not *perfectly* parallel (+/- 0.15 mm across ~48 mm diameter) that's exactly what has happened: some flaps on the spool do not 'flip' freely which breaks the whole display. Originally I thought a slight increase to the flap slop would do the trick, but after giving it some more thought I think the slop decrease should be reverted in its entirety. It's one of the only critical dimensions in the design that can break the entire display if it's too tight, and there are too many factors that can influence the total clearance. Better to play things safe and keep this loose. If there is continued binding due to misalignment of flaps within the spool, the `spool_width_slop` value can be increased further. I also parameterized and slightly reduced the magnet hole interference. On my displays the magnet press-fit is tight and I ended up cracking one of the spools. The magnets seemed much happier and still held tight after some light sanding of the hole, so this value is being reduced ever-so-slightly.
Increases the clearance between the motor chassis and the right outside wall of the previous module. Also refactors this variable from 'slop' to 'clearance' for a better description of what it represents. This is still playing a bit of catch-up from the thickness change (#93). In the previous release where the thickness parameter was set to 3.2 mm but the manufactured panels were actually 3.0 mm thick, there was extra space between the motor chassis and the previous module. This change adds that additional 0.4 mm of clearance back as part of the motor clearance variable. The additional clearance also makes the widths between the two designs approximately equal - `enclosure_width` is 82.2 mm on v0.6, and 82.25 mm here. (The discrepancy is the additional 0.05 of flap width slop that was added in #114.) Note that #122 bumps this further to 82.35.
This increase was a mistake. The difference in thickness (0.2 on either side) should have been added to the spool slop instead of the flap width slop, as the inner width of the spool (flap_width_slop) was not affected by the thickness parameter, while the outer width of the spool (spool_width_slop) was. This commit corrects that oversight.
This reverts commit a68fc98 (squashed as part of 7be69db).