Signal chat: BeNscale vs. Showcase Miniatures

[peteski]

I am looking at what it will take to make it possible. Kingbright released a new form factor earlier this year in 0402 single-color that is ultra-thin, with overall dimensions of 1.0mm x 0.5mm x 0.2mm. It may... or may not... be possible to cluster three inside a Showcase head. We'll see.

Good luck with that!

All 0402 LEDs have a standard footprint of 0.040" x 0.020" (1.0mm x 0.5mm).  The thickness can vary.  the problem with ganging them together (side-by-side) is that the ends are all metalized and are electrically live.  Or are you thinking of stacking 3 on top of each other and  have the light emitted from the edge?

[jagged ben]

I believe the "flying nun" signal is more commonly referred to as a "darth vader" signal. ...

Or to get technical about it, a Type D signal.

[draskouasshat]

Unfortunately the only thing modern, if you want to call it that, about signal heads is the snow shield that are being used about everywhere on the bnsf for new installs. The signal itself on the otherhand isn't exactly new. I actually can't think of a "new" design of signal head off the top of my own head, just new terminal blocks and bulb sockets in essentially the same outer housing that's been around for half a century. About the only new thing to railroad signaling is the microprocessor based logic controllers and ptc equipment and that's all our of sight in our signal bungalows. Everything else was either designed in the 1920s and is still in use or its the same 1920s outer shell but freshened up with new fancy equipment inside of it. If it's not broke,  we don't fix it, well we try not to but engineering likes to try to redesign the wheel sometimes too. You'd be surprised at what we still have in service and how long is been out there.
As for leds in real signals, they're actually in a grey area with the fra as there's no rule on how much of the light is out percentage wise before it's considered a dark signal. I only know of a few spots we have put them in on the bnsf. We are no longer putting them in for n.

[C855B]

Good luck with that!

All 0402 LEDs have a standard footprint of 0.040" x 0.020" (1.0mm x 0.5mm).  The thickness can vary.  the problem with ganging them together (side-by-side) is that the ends are all metalized and are electrically live.  Or are you thinking of stacking 3 on top of each other and  have the light emitted from the edge?

Thought at the moment is the three arranged in sort of a pyramid, facing inward with the common holding them together. A drop of resin or clear paint in the middle becomes a lens/light pipe. The 0.2mm super-thin of these new LEDs is what might make this work.

[peteski]

Thought at the moment is the three arranged in sort of a pyramid, facing inward with the common holding them together. A drop of resin or clear paint in the middle becomes a lens/light pipe. The 0.2mm super-thin of these new LEDs is what might make this work.

I see.  That will require some micro-soldering skills.    The other problem is that the signal head is all metal (conductive). You will have to isolate the LEDs from it somehow.  It is doable (as a proof of concept) but I wouldn't want to do couple of dozen 2 and 3 head signals. 

[C855B]

Micro-soldering, yes. Neglected to mention I've done something very similar with three 0402 ambers to make a rotating beacon simulator (Ngineering's design). Used a bead of amber Tamiya transparent paint to form a dome. So, signal idea is the same thing, but the LEDs turned inward.

If making it fit in the metal head is too much an issue, it might be easier to use the Showcase casting as a master and just cast a copy in resin with the LED bundle embedded. But it'll be a while before I mess with this since we're in Run 8 trying to get the layout building prepped for winter.

[eric220]

Or to get technical about it, a Type D signal.

Well yes, or that.

[C855B]

All 0402 LEDs have a standard footprint of 0.040" x 0.020" (1.0mm x 0.5mm).  The thickness can vary.  the problem with ganging them together (side-by-side) is that the ends are all metalized and are electrically live.  Or are you thinking of stacking 3 on top of each other and  have the light emitted from the edge?

New information: 0404 RGB LEDs are now available. This will fit inside a Showcase head. Yes, there's still micro-soldering to get the leads on, but there's no assembly as with my idea to use three discretes. The work then comes in programming a PIC or something like that to convert incoming R-Y-G (or R-Y-Lunar) logic levels to the color mixing to drive the RGB. Red and green on these chips are correct and can be driven at 100%, amber and lunar will require management. But then there's the opportunity to get fancy with the PIC and simulate things like the warm-up/cool-down of a flashing indication, and the red flash when going from green to yellow and vice versa.

[railnerd]

Here's a "glossy" for the 404 RGB LEDs: http://www.lumex.com/images/Lumex_QuasarBrite_0404SMDRGB_ProductBrochure.pdf

I have been playing with using digitially programmed diffused RGB LEDs, with an eye toward using them to drive the showcase signals (http://www.adafruit.com/products/1938)  I've been getting good results mixing in some blue to get a more "signal green" look.

A small bit of diffusing material in front of the 404 LED might work well.

-Dave

[C855B]

I've been getting good results mixing in some blue to get a more "signal green" look.

Hmm. The NeoPixel green might be a little off, since the Lumex and another RGB chip I'm looking at spec green peak at 518nM, which is spot-on for signal green, maybe already on the blue side. Also, give mixing-in a little blue with the red a try. I'm recalling running across info that when Corning Glass standardized the signal colors way-back-when, signal red has a hint of purple in it. I do like that NeoPixel integrated-controller design. Very nice.

How is the amber? This is the main hesitation I have with RGB. Even with diffusers, I've not had the best luck getting a homogenous mix - the eye can still pick out the component colors.

[peteski]

That is a good news!  If a 3-die 0404 LED does exist now, the factory could be set up to manufacture a RYG led in the same package. But I'm sure there would be a a minimum quantity required (probably something like 5,000 or 10,000) pieces. I wonder if someone like Richmond Controls or Ngineering would be intrested in doing this...

[railnerd]

That is a good news!  If a 3-die 0404 LED does exist now, the factory could be set up to manufacture a RYG led in the same package. But I'm sure there would be a a minimum quantity required (probably something like 5,000 or 10,000) pieces. I wonder if someone like Richmond Controls or Ngineering would be intrested in doing this...

I'd rather not have to pay $11 per LED...

-Dave

[peteski]

I'd rather not have to pay $11 per LED...

-Dave

I agree, but using an RGB LED plus a microcontroller to get the right colors out of it will probably cost more than $11 per signal head.  With RYG LEDs the circuitry would be much simpler.

[robert3985]

...flying nun... Darth Vader...

[sigh]

Snow hoods. They've been used on the UP in snow country since at least the late '50s. One of the Canadian asshats will have to verify whether UP was the pioneer or if it was something we imported from them. The problem they solved was blowing snow accumulating on the visors of conventional three-light color signals, blocking the aspect above, and accumulated snow in general over the lenses of approach-lit signals. They are especially vogue these days because of LED signals, which do not generate enough heat to melt any snow or ice which might stick to the lenses.

Additional benefit was better aspect visibility in glare conditions.

Aesthetically they suck, although Lennart's Safetran mid-dwarf example above (used mostly as siding exits) has a smaller hood so doesn't have that cumbersome, top-heavy look the full signals have. I like these half-size signals and hope to incorporate a few on my layout where appropriate.

I personally think they're beautiful and am sad to see they're being replaced between Ogden and Wahsatch on the Weber and Echo Canyon helper district.  Here's a prototype photo for ya...



Here's one on one of my scratchbuilt U.P. Cantilever Signal Bridges.  I had to build it out of .003" brass.  Not too difficult.


Here's some on a modified Traincat etched U.P. Cantilever Signal Bridge kit that I did for my friend Nate (Nato) for his layout:


The .003" brass they're made from really contributes to a prototypical look.  I built them from prototype blueprints I found in one of my old U.P.H.S. "Streamliner" quarterlies.