Author Topic: Pennsy Position Signal Head -- a possible solution?  (Read 8843 times)

0 Members and 2 Guests are viewing this topic.

SkipGear

  • Crew
  • *
  • Posts: 2418
  • Gender: Male
  • Respect: +629
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #30 on: March 08, 2013, 09:38:17 PM »
0
I know how to make it work but it would take mass production with pick and place equipment and wave or hot air soldering.

The circuit board is the target back plate, hoods and cover plates, as well as the LED's are soldered to the front of it.

Painting would be the interesting part but a mask could be made to fit in the hoods to protect the LED's. It would also be the handle to facilitate painting.
Tony Hines

nscaleSPF2

  • Crew
  • *
  • Posts: 383
  • Gender: Male
  • knowwhatimean?
  • Respect: +103
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #31 on: March 08, 2013, 10:21:14 PM »
0
Tony,  I'm sorry, my description was lacking in detail.  There would be no circuit board per se.  The head (including the hoods) would be molded out of black plastic (no painting required) and the LED's would be hard wired first and then glued into the back of the head.  More hobbiest bench assembly, rather than mass production.  Sorry I didn't clearly communicate my intent.

Jim
Jim Hale

Trying to re-create a part of south-central Pennsylvania in 1956, one small bit at a time.

peteski

  • Crew
  • *
  • Posts: 32958
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5342
    • Coming (not so) soon...
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #32 on: March 09, 2013, 12:01:22 AM »
0
Alkem circuit board could be manufactured from a much thinner PC board material.  The LEDs look like they are 0805 size. Thue could be replaced by thinner 0603 size LED which. That would reduce the thickness.  The wires could also be made thinner.

Jim,  if your idea is to make the head out of metal and mount the LEDs directly in the head, you need to make some allowance for insulating the LEDs from the head. Otherwise, the LED terminals will short out though the metal head.
. . . 42 . . .

SkipGear

  • Crew
  • *
  • Posts: 2418
  • Gender: Male
  • Respect: +629
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #33 on: March 09, 2013, 12:52:41 AM »
0
Tony,  I'm sorry, my description was lacking in detail.  There would be no circuit board per se.  The head (including the hoods) would be molded out of black plastic (no painting required) and the LED's would be hard wired first and then glued into the back of the head.  More hobbiest bench assembly, rather than mass production.  Sorry I didn't clearly communicate my intent.

Jim

I've been doing a lot of soldering with 603's recently and it is very hard to control their orientation, and to control where the light goes when they are soldered to magnet wire. It would not be impossible, but very difficult to do that consistantly over a number of signals. Mounting the LED to a board is the best way in my opinion to get them consistant. If you do that, there is no need for a back plate, just solder the hoods over the LED's and done.

I have experimented with using my oven to solder surface mount stuff. It is a trick in the computer industry to take an intermittent problem circuit board and bake it to resolder a cold solder joints. I have also played with it using a heat gun with some luck. You could glue the LED in place on a new clean board with just a pre tin of solder on it, heat the board and everything would solder at once. It is concieveable that it would work.
Tony Hines

peteski

  • Crew
  • *
  • Posts: 32958
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5342
    • Coming (not so) soon...
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #34 on: March 09, 2013, 02:49:27 AM »
0
Tony, I often solder arrays of 0603 LEDs by placing them upside down on a piece of self-adhesive double-sided tape, then soldering the wires.  The wires are also placed on and held by the tape's adhesive (or sometimes by hand). It is vital to place a minute amount of flux (I use Radio Shack Electronic paste flux) on the LED terminals. That greatly improves the quality of the solder joint and speeds up the soldering time.  Then just place tiny bit of solder on a clean soldering iron tip and touch it to where the wire crosses the LED terminal.

If I have a series-connected string of LEDs (and when I can use bare and thinned wire) then I just place a single piece of wire across all the LEDs in the string and solder it to all the LED terminals. Then afterwards I cut into the wire in the middle of the LEDs, using a hobby knife with a new #11 blade.

I do not recommend using an oven to solder the LEDs.  They aren't designed to withstand long periods of time in temperatures which will melt solder. In tindustrial applications the soldering process is done very quickly.  Using solder paste (which is tiny spheres of solder suspended in a liquid flux) and a heat gun might work (as it will be much faster than using an oven).
. . . 42 . . .

nscaleSPF2

  • Crew
  • *
  • Posts: 383
  • Gender: Male
  • knowwhatimean?
  • Respect: +103
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #35 on: March 09, 2013, 08:13:43 AM »
0
Tony, I had to read your post several times before I got it.  (My Polish heritage is showing, again.)

I've been doing a lot of soldering with 603's recently and it is very hard to control their orientation, and to control where the light goes when they are soldered to magnet wire. It would not be impossible, but very difficult to do that consistantly over a number of signals. Mounting the LED to a board is the best way in my opinion to get them consistant. If you do that, there is no need for a back plate, just solder the hoods over the LED's and done.

I think you are talking about a circuit board that is clad with copper on both sides.  LED's soldered to one side and shining thru the board, and lens hoods soldered to the other side.  Right?  I like it.  Might need some kind of fixture to hold the hoods, but David K has apparently figured this out...
Jim Hale

Trying to re-create a part of south-central Pennsylvania in 1956, one small bit at a time.

SkipGear

  • Crew
  • *
  • Posts: 2418
  • Gender: Male
  • Respect: +629
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #36 on: March 09, 2013, 08:19:23 AM »
0
Tony, I often solder arrays of 0603 LEDs by placing them upside down on a piece of self-adhesive double-sided tape, then soldering the wires.  The wires are also placed on and held by the tape's adhesive (or sometimes by hand). It is vital to place a minute amount of flux (I use Radio Shack Electronic paste flux) on the LED terminals. That greatly improves the quality of the solder joint and speeds up the soldering time.  Then just place tiny bit of solder on a clean soldering iron tip and touch it to where the wire crosses the LED terminal.

If I have a series-connected string of LEDs (and when I can use bare and thinned wire) then I just place a single piece of wire across all the LEDs in the string and solder it to all the LED terminals. Then afterwards I cut into the wire in the middle of the LEDs, using a hobby knife with a new #11 blade.

I do not recommend using an oven to solder the LEDs.  They aren't designed to withstand long periods of time in temperatures which will melt solder. In tindustrial applications the soldering process is done very quickly.  Using solder paste (which is tiny spheres of solder suspended in a liquid flux) and a heat gun might work (as it will be much faster than using an oven).

Low temp solder melts at 345 deg. well below the point that the LED's would be damaged. How do you think all these LED's are used in production of cell phones, video games, etc. You can set an oven to 350 deg and the melt the solder without damaging the LED. A heat gun used for shrinking the covering on RC planes is hot enough to melt the solder without damaging the LED or board.
« Last Edit: March 09, 2013, 08:31:16 AM by SkipGear »
Tony Hines

DKS

  • The Pitt
  • Crew
  • *
  • Posts: 13424
  • Respect: +7026
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #37 on: March 09, 2013, 10:05:52 AM »
0
I can envision two approaches to making these signals. One is DIY, for the sake of economy; the other is RTR, although the price point may be on the high side. Both start with a one-piece casting like the one I made. It could be lost-wax brass or spin-cast pewter, but I see advantages to a resin casting. Being non-metallic, it avoids any possible electrical issues with the LEDs/PC boards. Also, it can be cast in black.

As for illumination, the DIY approach is not particularly easy, as it involves soldering wires to 0603 SMD LEDs. The difficulty of positioning/monting them on the back of the target, however, can be reduced by using SMDs with integral lenses, which would fit into the holes in the target (and provide a modest cosmetic improvement at the same time). 0603s with integral lenses are readily available.

The RTR approach will be costly, because it would involve having a die made to cut thin PC board parts. Ideally the openings for the LEDs would be rectangular, allowing them to sit flush in the PC board to keep the profile as low as possible (ultra-thin SMDs are available for this application). While a double-sided board isn't necessary, it would make wiring a little simpler. Copper-clad holes for the LEDs would be most desirable, as it would make soldering much easier, although this would be the most costly option.

Bob Knight was in the process of developing a PC board with LEDs mounted on the back that shone through holes. I have no idea how far along he was (although IIRC he posted a few images), or if the project is moribund.

Edit: Here's Bob's post. Sure looks like there was a hefty investment in this project...  https://www.therailwire.net/forum/index.php?topic=22151.msg253040#msg253040
« Last Edit: March 09, 2013, 11:22:04 AM by David K. Smith »

C855B

  • Crew
  • *
  • Posts: 10871
  • Respect: +2418
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #38 on: March 09, 2013, 11:43:53 AM »
0
Bob's approach is especially intriguing to me because with a slight variation you can use a thin PC stock (the .015 Pete mentioned), and instead of mounting 0603 reverse-lens to the back, I would put 0402 LEDs in the holes, allowing the possibility of better modeling of the conduits on the back of the signal. I'm looking at the possibility of this approach for standard color-light signals (US&S Type D and similar), and then 3D-print a variety of common castings to glue on the back (and also run the wires through).

I was hoping Bob would have followed-through on this project, although I fully understand his circumstances which may have tabled it. He had indicated in that thread the desire to extend the architecture to color-light signals. What is lost is that he created the artwork and cultivated the supplier relationships, and have now left the rest of us to walk down this road all over again. Bob, if you're reading this I do hope that if you aren't going to be able to do this in the near future that you can see your way to pass the baton to a capable producer such as David.
...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.

DKS

  • The Pitt
  • Crew
  • *
  • Posts: 13424
  • Respect: +7026
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #39 on: March 09, 2013, 11:56:37 AM »
0
...instead of mounting 0603 reverse-lens to the back, I would put 0402 LEDs in the holes, allowing the possibility of better modeling of the conduits on the back of the signal...

That was my point in my prior post. However, I don't see an advantage to using 0402s, as I believe the 0603s are adequately small.

Also, the board Bob was using is .010" thick, which is the same thickness as the ultra-thin SMDs I cited.
« Last Edit: March 09, 2013, 11:58:58 AM by David K. Smith »

C855B

  • Crew
  • *
  • Posts: 10871
  • Respect: +2418
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #40 on: March 09, 2013, 12:27:24 PM »
0
Actually, there is a point in using 0402s. The way I would design these is to have traces on the face to be used for soldering the visors. 0603s would make the holes just a shade too big to do this, with risk of the traces overlapping or the solder creeping into fillets which would then short the pads for the LEDs. The smaller LED form factor would also allow round holes, as rectangular in that size are going to be difficult to manufacture, creating a milling step not normally handled by the board houses.

There is also .007" board material available (the multi-layer board makers use it), but I've worked with it just enough to know that it's awfully brittle, pretty much dependent on living inside the laminate. I don't think it would be an issue after manufacture, but breakage loss during assembly of the finished piece might be enough of a problem to affect overall production costs.

FWIW. Just thinking out loud.
...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.

DKS

  • The Pitt
  • Crew
  • *
  • Posts: 13424
  • Respect: +7026
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #41 on: March 09, 2013, 12:35:03 PM »
0
Actually, there is a point in using 0402s. The way I would design these is to have traces on the face to be used for soldering the visors.

Yikes, soldering on the hoods? You're a better man than I am, Gunga Din.

C855B

  • Crew
  • *
  • Posts: 10871
  • Respect: +2418
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #42 on: March 09, 2013, 12:50:24 PM »
0
Solder paste covers a lot of sins with ultra-small work. I just wish I could find my jar of it. Expensive. :(
...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.

peteski

  • Crew
  • *
  • Posts: 32958
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5342
    • Coming (not so) soon...
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #43 on: March 09, 2013, 05:03:13 PM »
0
My thought was using a very thin PC board in the manner it was used on the Alkem signals.  While not ideal, it would look better than the rather thick Alkem's PC board.  A variation of this, to further reduce the overall thickness would be to make holes in the PC board for the LEDs with the solder pads on the back side. That way the LEDs could be soldered to the back of the board with their "lens" protruding through the hole.  That would result in overall reduced thickness.  Then  glue the circuit board on the back of the brass head.

Soldering individual hoods onto the circuit board, while it sounds doable,  is not something I would consider feasible.  Too time consuming and prone to misalignment.

Another alternative woudl be to have the LED dies mounter directly on a thuin circuit board. Then the dies would be sealed with a drop of clear epoxy to protect the dies and bond wire from damage.  That would be the thinnest possibe assembly.  This type of process is often used on multi-digit LED displays and on many of the small watch battery powered flashing LED trinkets which are sold as wearable items.  But going this route woudl require rather large sum of money and a fairly large quantity to keep the cost down.

Tony, as someone who used to rework SMD devices for living, I'm aware of many soldering techniques. You are correct that the industrially produced circuit boards with SMD components (including LEDs) are soldered using hot air, but that is done relatively quickly. This is especially important with the lead-free solders used nowadays (which have higher melting temperatures than the older solders with lead).

If you placed your board in an oven, it would have to sit in there for couple of minutes to reach the melting temperature of the solder. Look up some technical specs for any LEDs. They are guaranteed to withstand the elevated soldering temperatures for seconds (not minutes).
. . . 42 . . .

Mark5

  • Crew
  • *
  • Posts: 11031
  • Always with the negative waves Moriarty ...
  • Respect: +608
Re: Pennsy Position Signal Head -- a possible solution?
« Reply #44 on: March 09, 2013, 09:36:53 PM »
0
but I see advantages to a resin casting. Being non-metallic, it avoids any possible electrical issues with the LEDs/PC boards. Also, it can be cast in black.

Excellent point, especially for someone scared about soldering LEDs (me!)

Quote
As for illumination, the DIY approach is not particularly easy, as it involves soldering wires to 0603 SMD LEDs. The difficulty of positioning/monting them on the back of the target, however, can be reduced by using SMDs with integral lenses, which would fit into the holes in the target (and provide a modest cosmetic improvement at the same time). 0603s with integral lenses are readily available.

Hmm, thats yellow. Are these the equivalents for red and green?

Red: http://www.kingbrightusa.com/product.asp?catalog_name=LED&product_id=APTD1608SURCK

Green: http://www.kingbrightusa.com/product.asp?catalog_name=LED&product_id=APTD1608ZGC

Signalling is core for me in terms of recreating a railroad in miniature, so I am watching this thread with great interest.

If the CPL option does not make the cut as RTR, then hopefully the "kit" option will be available too.

Mark