Author Topic: 3D-Printed Signal Heads  (Read 6464 times)

0 Members and 1 Guest are viewing this topic.

narrowminded

  • Crew
  • *
  • Posts: 2305
  • Respect: +743
Re: 3D-Printed Signal Heads
« Reply #30 on: January 28, 2019, 07:06:57 PM »
+1
A perfectly-filled 3/64", 0.005"-walled tube of 38AWG wire:



Practically speaking this can never happen, but it conveys the amount of relative space we have to work with. I don't see any issues.

My real concern isn't the theoretical but the act of starting multiple delicate wires down the tube by bending them 90 degrees as the first official act.  Pushing a wet noodle comes to mind. ;)  Maybe this won't prove to be an issue but I'd try it before to commit.  If it's not a problem, GREAT! 8)  The idea is to allow for the smallest practical tube size, be it 3/64" or something larger or smaller.  I don't have any direct experience with this exact thing so maybe I'm being overly cautious but... :|

Another thing I might be able to do, especially if a saddle is to be used, is to put the hole in by plunging with an end mill at an angle to make a more gentle approach, more like a soft angle that the wire is asked to negotiate.  That might work and would have the added advantage of minimizing the sharp edge that might try to scrape through the varnish insulation.  I would be happy to try some of these things before to have folks commit if it would help. 8)
Mark G.

Lemosteam

  • Crew
  • *
  • Posts: 5919
  • Gender: Male
  • PRR, The Standard Railroad of my World
  • Respect: +3666
    • Designer at Keystone Details
Re: 3D-Printed Signal Heads
« Reply #31 on: January 28, 2019, 07:54:46 PM »
0
Thread some 0.005 pb wire up the tube and solder all of the leads onto that, then pull the wires through.

Mark W

  • Crew
  • *
  • Posts: 1988
  • Respect: +2125
    • Free-moNebraska
Re: 3D-Printed Signal Heads
« Reply #32 on: January 28, 2019, 08:10:48 PM »
0
..is to put the hole in by plunging with an end mill at an angle to make a more gentle approach..

This would probably be the preferred method. 

Alternatively, what do you guys think about either of these ideas?  Yellow = Brass/Tube, Gray = print.  The first one would be the cheapest option and easiest to assemble, but would lose a lot of strength.
Then again, perhaps that's yet another benefit!  Cleaning track is risky business.  If something were to snag, I'd rather it break in a way I could *hopefully* just glue it back on.  That first option would have some natural break points for sure.   8)


https://i.imgur.com/E1ik7Ht.jpg
Contact me about custom model building.
Learn more about Free-moNebraska.
Learn more about HOn3-mo.

peteski

  • Crew
  • *
  • Posts: 32924
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5323
    • Coming (not so) soon...
Re: 3D-Printed Signal Heads
« Reply #33 on: January 28, 2019, 09:06:47 PM »
0
I think this approach would *SERIOUSLY* compromise the tube's strength. When snagged, not only the head assy. would detach - the half-tube would bend easily.
. . . 42 . . .

Mark W

  • Crew
  • *
  • Posts: 1988
  • Respect: +2125
    • Free-moNebraska
Re: 3D-Printed Signal Heads
« Reply #34 on: January 28, 2019, 10:52:13 PM »
0
I thought I'd catch the group up on some of my T-scale rolling stock projects. 


Worlds smallest CPL, GM50, and the cliche giant penny.


Alright @Jesse6669, spill it!  Are these T GAUGE signals 3D printed and operational?!  :D :D
Contact me about custom model building.
Learn more about Free-moNebraska.
Learn more about HOn3-mo.

narrowminded

  • Crew
  • *
  • Posts: 2305
  • Respect: +743
Re: 3D-Printed Signal Heads
« Reply #35 on: January 28, 2019, 11:30:48 PM »
0
This would probably be the preferred method. 

Alternatively, what do you guys think about either of these ideas?  Yellow = Brass/Tube, Gray = print.  The first one would be the cheapest option and easiest to assemble, but would lose a lot of strength.
Then again, perhaps that's yet another benefit!  Cleaning track is risky business.  If something were to snag, I'd rather it break in a way I could *hopefully* just glue it back on.  That first option would have some natural break points for sure.   8)


https://i.imgur.com/E1ik7Ht.jpg

I think the second one with the milled slot would be practical to make and easy enough to work with.  Another look at those pictures from Robert's post might be useful for adding a little detail that might also help with strength.  Add the "flange" on the spigot/ saddle piece on the signal back, per the prototype, that could receive wire wrapped as the U-bolts do, all glued up with the wire reinforcing the whole rig. 

Before to panic, that wire could be something like .008" PB wire that can easily be hand bent and at a length that can be handled, and instead of trying to fit through holes in the flange there could be narrow slots in the edge that you put that wire into, then bend each side over towards center, overlapping them, pulling them taught and laying on the face behind the signal back, on the flange face, projecting out each opposite side, and pulled up taught.  CA on the grooves and overlap on the flange face (also sitting in a shallow groove?) encapsulating the wire should hold it, add the detail, and add a little more strength to the assembly.  And of course, once the CA cures snip the wire pigtails flush.  I don't think that will be too fragile, any more than the whole rig already is.

And to that end, maybe some real thought into a base design that makes the thing serviceable would be in order.   Maybe a receptacle with a socket to receive the tube that's somewhat large and could pass wire and maybe even a plug that could be permanently set into the board and any oversized flange able to be scenicked in.  A robust attachment that affords an easy way to remove the whole assembly should the need arise.  The foundation pad visible in the photos might be part of the permanent base.  It could have a socket to receive the tubing with all of the detail per the prototype.  Even if it wasn't easy to do but could be done should the need arise, somebody, some day, will be thankful for it. 8)

Again, intended only to help, not intrude on the project.  You guys are doing pretty darn good all by yourselves. ;) :D  Really! 8)
« Last Edit: January 29, 2019, 12:09:13 AM by narrowminded »
Mark G.

C855B

  • Crew
  • *
  • Posts: 10860
  • Respect: +2415
Re: 3D-Printed Signal Heads
« Reply #36 on: January 29, 2019, 12:04:33 AM »
0
...



I'm warming up to the milled slot/insert idea on the right. Has the advantage of a near-straight shot at feeding the wires down the tube. Like Mark pondered, add flange details to the fitting with appropriate application of CA as adhesive and filler and I think it's something we can work with.
...mike

http://www.gibboncozadandwestern.com

Note: Images linked in my postings are on an HTTP server, not HTTPS. Enable "mixed content" in your browser to view.

There are over 1000 images on this server. Not changing anytime soon.

narrowminded

  • Crew
  • *
  • Posts: 2305
  • Respect: +743
Re: 3D-Printed Signal Heads
« Reply #37 on: January 29, 2019, 12:06:10 AM »
+1
I think this approach would *SERIOUSLY* compromise the tube's strength. When snagged, not only the head assy. would detach - the half-tube would bend easily.

The second one, the one on the right, I don't think would be that bad. :|  It's up high, the cut out void would be glued snugly to the signal housing, filling and supporting the void to some degree, and then while consumed with this potential issue, there's the whole signal housing that will be somewhat delicate.  Also, the cut section is high on the mast so by leverage, the point at the base might be just as bad or worse, depending on the joint execution.  In the pursuit of prototype fidelity I don't think you'll be able to entirely eliminate the delicate nature of the whole thing.  And this helps make the argument for a serviceable mounting method. 

Maybe the signal could just be set into a socket so if it is caught it could turn, relieving the strain in at least one direction.  And if it was just set in a socket, held with some light friction, maybe it could be removed and laid down during maintenance that invites the trouble you're worried about.  And in general, NO LONG SLEEVES working over any layout. :)
Mark G.

narrowminded

  • Crew
  • *
  • Posts: 2305
  • Respect: +743
Re: 3D-Printed Signal Heads
« Reply #38 on: January 29, 2019, 12:55:50 AM »
0
@Mark W  Another thought just struck me.  Maybe while designing the head back assembly you could add another removable saddle up high that would register to the tube to aid in aligning, squaring, and lightly clamping the thing during assembly.  Don't glue the support piece.  Just let it help with positioning.  Once assembled and all of the glue cured, snip it off.  Not unlike a typical printing support.   8)  And I'll try to quit now. ;)
« Last Edit: January 29, 2019, 12:57:28 AM by narrowminded »
Mark G.

narrowminded

  • Crew
  • *
  • Posts: 2305
  • Respect: +743
Re: 3D-Printed Signal Heads
« Reply #39 on: January 29, 2019, 01:09:57 AM »
0
Oh yeah.  And cleaning up CA excesses, you can use acetone.  It's the solvent for the CA and won't attack the resin.  Sidenote, MEK doesn't get the resin either.
Mark G.

jj3

  • Crew
  • *
  • Posts: 32
  • Respect: +9
Re: 3D-Printed Signal Heads
« Reply #40 on: January 30, 2019, 01:40:30 PM »
0
Just tossing this out there for the right hand option that folks seem to like best, what about more of a C shape that would snap around the mast.

So it's slightly longer vertically than the slot, and would go past the brass cut half tube. Exaggerating for clarity, the round portion of the flange attaching to the mast would cover 3/4 of the actual diameter of the tube

Think of a C clip, where it's snapped onto the other part and requires a bit of flex (or tool) to remove it

RAILCAT

  • Crew
  • *
  • Posts: 142
  • Respect: +7
Re: 3D-Printed Signal Heads
« Reply #41 on: February 07, 2019, 10:30:45 AM »
0
Any further developments.

Specter3

  • Crew
  • *
  • Posts: 866
  • Respect: +156
Re: 3D-Printed Signal Heads
« Reply #42 on: August 01, 2020, 06:48:29 AM »
0
@Mark W @C855B

Did anything come of this? As I start construction of my layout, I am fully intending on lighted signals. You guys seemed to have come within spitting distance of a decent D type signal. For my Southern layout I need a bunch of these. Any further results?

DKS

  • The Pitt
  • Crew
  • *
  • Posts: 13424
  • Respect: +7026
Re: 3D-Printed Signal Heads
« Reply #43 on: August 01, 2020, 07:15:31 AM »
0
Might this be a viable option to reduce assembly complexity and increase strength?


C855B

  • Crew
  • *
  • Posts: 10860
  • Respect: +2415
Re: 3D-Printed Signal Heads
« Reply #44 on: August 01, 2020, 10:06:57 AM »
0
@Mark W @C855B

Did anything come of this? As I start construction of my layout, I am fully intending on lighted signals. You guys seemed to have come within spitting distance of a decent D type signal. For my Southern layout I need a bunch of these. Any further results?
Might this be a viable option to reduce assembly complexity and increase strength?



I can't speak for Mark although we did talk about a mutual frustration with the various designs we were fooling with: soldering leads to the LEDs in a consistent, reproduceable manner. It is a tedious PITA, and I can't imagine for a second doing the 50 or 100 or 10,000 (it seems) needed for my MRR.

The mount, classic or modern, is not an issue. We can work that out. It's the LEDs that are the stumbling block.

As for my own (lack of) progress, I am standing by for Showcase Miniatures. They announced a three-color vertical several months ago, tho' with no details about what era or prototype. Whatever they do is going to require a pre-wired light board of some kind; they contracted with Richmond Controls for their SA-type LED, so I would surmise something similar with a three-color. Anyway, no news from Walter or Debbie on that front since.
...mike

http://www.gibboncozadandwestern.com

Note: Images linked in my postings are on an HTTP server, not HTTPS. Enable "mixed content" in your browser to view.

There are over 1000 images on this server. Not changing anytime soon.