N scale searchlight signal build thread

[GaryHinshaw]

Step 0: the ingredients
Now that I'm deep into the project of assembling searchlight signals for my layout, and having encountered several snags along the way, I thought I would share some of the lessons learned in the course of getting to this stage.  The story will be spread out over a few posts, starting with this intro to the components, and a few shots of the finished products.  Here are the hardware components:



There are 5 separate sources for these components, all of them basically mom & pop shops (or pop-up shops), 2 of whom are Railwire members.  Here is a brief description of each, with links.  More detailed info will follow, as needed.

1. RR-CirKits SignalMan-S driver board and pre-wired 3-colour LEDs (p/n SS-RGY-24).  The board can drive up to 4 LEDs (one typical control point) via JMRI, and no additional resistors are required to drive their LEDs.

2. Alvin Wutter's modular signal base.  These can be purchased directly from @Wutter (at least I think they still can...).  This consists of a custom board and a printed frame and the combination provides an excellent mechanical and electrical foundation for most any signal.  As noted below, and references therein, this product was developed in collaboration with the Modutrak group.

3. Showcase Miniature/Century Foundry kits, #514/515/516.  With 2 exceptions, all of the parts circled in 3 are from the Showcase kits.  The only difference between the various model numbers is the style of base: 514 has a pedestal, 515 has a single cabinet, 516 has a dual cabinet.

4. Ngineering.com thin walled stainless steel tubing (p/n N2032U-2 or -4).  The Showcase kit ships with stock .032" brass tubing which has a fairly thick wall.  This is fine for single-head signals with 4 LED leads, but for multi-head masts, do yourself a favour and use the thin-walled tubing in its place.  (You can extract the brass tubing from the cabinet bases with some pliers and a little bit of effort.)  The -2 p/n is a 2-pack, -4 is a 4-pack.

5. 3-d printed H2 signal heads from @C855B.  I have raved about these elsewhere - they are much finer than the Showcase heads, so do what you must to bribe Mike into printing you some, or download the .stl file (link posted below) and print them yourself.

Here are a few examples of finished signals.  One thing I love about the prototype signals on Tehachapi is that every one of them is a little bit different, and the kits allow you to reproduce that variety easily.  In subsequent posts, I'll walk through a sample build and give some tips (none of which are going to be profound).  It might take a few days though, so bear with me.

[ednadolski]

Fantastic thread, thanks @GaryHinshaw !   It's a lot of work for sure to put this sort of thing together, I know can speak for all of TRW to express how much we all appreciate this!

Ed

[GaryHinshaw]

Thx Ed.

Step 1: preparing the heads
The first step is an aliveness test (the first of many along the way).  I unspool some LED leads and screw them directly into the appropriate Signalman ports to make sure they work properly.   I keep a spare board around for this purpose.

Next I prep the signal head by cleaning it up.  If you're using the Showcase housing you'll have to drill out the lens hole, and you might have to ream out a bit of material from inside to allow the LED to fit squarely in the head.  If you're using Mike's, you'll want to give it at least one coat of paint inside to help with light leaks.  (More on this later.)  Before installing the LED, I trim the leads to be just a bit longer than I'll eventually need (depends on your mast height) because you don't want to thread any more wire than necessary through the mast.  I then install the LED in the head and use a bit of canopy glue to help secure it (the fit is pretty snug, so not much is needed.)  I then add the back plate with canopy glue and let it dry.

Once dry, test the head again to make sure nothing happened to the leads during this step.  I usually assemble 4 heads at a time as long as I have the test board handy.

Here are a few random shots for illustration, showing both a Showcase head and a C855B one:





Note in the last shot that the Showcase head casting has small breakthrough in it that will leak light if not sealed or replaced.  Order extra castings and try to cull these ones out....   Here is a C855B head ready for installation:

[C855B]

Before we get too deep into extolling the virtues of my fooling around with the Photon    , how was the breakage loss with that last shipment?

You see, I am pressing the limits of the Photon and its resins, with many areas of the finished heads 0.008" or even less, similar to that breakthrough Gary highlighted on the Showcase casting. It doesn't help that the cured resin is quite brittle. The first batch I sent Gary had about half of the heads broken on arrival IIRC. Bottom line - while I now have decent repeatability with the printing, the big challenge in making them available is the packing for survival. I'll post when I think we have a technique that's protective enough.

[GaryHinshaw]

Good question Mike.  I haven't done a careful inspection yet, but the survival rate is definitely higher than 50% from a quick pass through them; probably closer to 80% or more.  I'm still working through the first batch you sent, and my experience working with the ones that survived shipping is that they're extremely durable.  Of the dozen heads that I've finished, I've had zero issues with them during the builds.  I was worried that they might be fragile, but that hasn't been the case. 

On the subject of heads, I went through my stash of antique (read: discontinued) heads and took a couple of quick 'head-to-head' comparison photos (sorry for the messy stage).  From left to right: Showcase/C855B, BeNscale, Traincat, Sunrise.




(Please also forgive the fact that I installed the Traincat target backwards.  )  The Sunrise head holds up pretty well, especially given its age, but it's a moot point.  I've never lit any of the other models up, so I don't know how that 'aspect' compares.

[GaryHinshaw]

Step 2: preparing the base & mast
With the heads prepped, the next step is to prepare the base and mast.  For this example I am building a 2-head signal with a double cabinet base for the south entrance to Walong siding.  The ingredients for this step are shown below:



(Looks like I need to file the ladder sprues a bit more.)  You'll need to refer to prototype photos to determine how tall to make your mast and where to place the bracket for the lower head.

Since this will be a 2-head signal, I'm using the thin-walled tubing, so I extracted the brass tubing from the relay case by twisting and pulling on the tube with pliers.  Here's a shot that shows pretty clearly the gain available with this tubing:



By my estimate, there is about 80% more clear area, which makes threading the leads for the 2nd head dramatically easier. 

The next step is to file a feedthrough hole in the mast for the leads of the lower head.  Before you do though, I recommend determining where your ladder support brackets will be located, so you can avoid those locations for the feedthrough.  I go ahead and form the ladder and trim it to length now and figure out which support holes in Wutter's base I'll use to receive the bottom of the ladder.  I can then determine which rungs of the ladder are the correct distance from the mast to accommodate the upper and lower bracket.  In this case, I'll place the upper bracket on the 5th rung down from the top, so I've marked that elevation on the mast so I know to avoid putting my feedthrough there:



In the above shot, nothing is glued in place yet.  I have the base sitting on .060" styrene spacers so the mast pole will protrude that much below the base, but there is nothing critical about that dimension.  The main point is to have the mast pole supported at it correct position relative to the ladder stock. 

You can now cut the feedthrough a bit below the lower head bracket and clear of the ladder brackets.  I use a triangular file for this, per the Showcase instructions.   Be sure not to go too deep with the cut, lest you weaken the mast.  For the upper head, you can either cut another feedthrough below the upper head bracket (and clear of the top of the ladder), or just feed the leads directly down the top of the tube.  The latter is easier, the former is more prototypical.  I chose the easy way.

Once the mast is prepared, de-burr all the cuts.  I mainly twirl the tip of a #11 blade for this, but there are a variety of techniques that work.  Next, clean the parts and glue them together, making sure the mast is square to the base.  It should look something like this now:



The mark above the feedthrough is where the lower head bracket will go. 

The next post will cover threading the LED leads and wiring the base.

[babbo_enzo]

Whau! Gary, awe tutorial! is complete training on what I plan to do in near future!
I'll print everything for my reference and wait your next installement. Good suggestions on tubing too.
I wonder "if" Mike will share his file for the head for us overseas? ... I, "probably" have to plan a Photon printer in my future ... also have some files kindly send me Ken Harstine of some SP details (phone boot, telltales and switch stand !)
Can we create a sort of RailWire library of 3D files (need to ask to Ken too)...
Keep your good work going!
Take care all and ... mask on your face
Cheers
Enzo Fortuna
modeling the Friendly ... in Italy

[GaryHinshaw]

I wonder "if" Mike will share his file for the head for us overseas? ... I, "probably" have to plan a Photon printer in my future ...

Can we create a sort of RailWire library of 3D files (need to ask to Ken too)...

Thanks Enzo.  Paging @C855B...

In general, I do like the idea of hosting a file sharing site for 2-d and 3-d artwork that people are happy to share.  But I haven't thought through any of the logistics or costs.  That's a discussion for another thread, and a decision for Tom to make.

[C855B]

When I get over to the studio later today, I'm going to put the STL files on my web server and see if they remain "whole" when downloaded. Some browsers may have trouble downloading non-HTTP, non-image, non-text files. Maybe a .zip file will solve that, but I'm hesitant given that a lot of anti-virus software and firewalls barf on .zips due to "the bad guys" commonly hiding malware in them. I know this doesn't help with folks that don't have a 3D resin printer, but I think the whole shipping thing is more than I'm interested in tackling on a wholesale basis.

Obviously I'm not too worried about the IP aspects, esp. given my sister is a high-powered IP litigator. If somebody wants to make coin off my artwork, have at it, but I kinda, sorta, wouldn't mind a smidge of credit. I'll post the links once they're up and tested.

[glakedylan]

Gary (and all others involved: C855B and Wutter) this is fantastic!
to be able to model these in N scale with such accurate
and sharp parts is remarkable, to say the least.
wow. just, simply, wow!
thanks for the write up and the list of parts needed.


sincerely
Gary

[C855B]

OK - 3D print files (STL) are up on the server:

US&S H-2 with Phankill and long visor. These are what @GaryHinshaw is using: http://www.everywherewest.com/Signal_Heads/H-2_1.2_SC.stl
US&S H-5 with Phankill and long visor. Later model, Santa Fe, among others: http://www.everywherewest.com/Signal_Heads/H-5_1.3_SC.stl
US&S H-2 with small target and short visor: http://www.everywherewest.com/Signal_Heads/H-2_1.2_SC_Mini-Dwarf.stl

Screen shot of H-2 CAD artwork, showing the alignment pins on the back which are press-fit to the body: http://www.everywherewest.com/Signal_Heads/H-2_Head.png



I recommend using gray resin for best opacity, although as Gary said above, black paint on the insides to help avoid light leaks. The rings around the visor are for seating the Showcase Miniatures target and target brace.

[Sokramiketes]

If somebody wants to make coin off my artwork, have at it, but I kinda, sorta, wouldn't mind a smidge of credit.

In this vein, let's also not forget in this thread that the PCB signal bases, specifically designing them to transition from magnet wire to durable wire, is a Modutrak innovation.  It stemmed from the BeNscale innovation to solder the signal together on a base, and originally soldering to extra lettering traces on the bottom of Ben's design.  This group has really honed in on a near perfect solution, through many iterations over 5+ years in this space.

Now, where do we get to the bouncing roundel part of what makes searchlights have souls?

[Wutter]

In this vein, let's also not forget in this thread that the PCB signal bases, specifically designing them to transition from magnet wire to durable wire, is a Modutrak innovation.  It stemmed from the BeNscale innovation to solder the signal together on a base, and originally soldering to extra lettering traces on the bottom of Ben's design.  This group has really honed in on a near perfect solution, through many iterations over 5+ years in this space.

+1, the inspiration for the signal base-resistor breakout is from Mike's original build of the BeNscale signals, and the subsequent etchings and base breakouts way back in 2013: https://www.therailwire.net/forum/index.php?topic=31256.15

[GaryHinshaw]

Thanks guys.  I confess that I forgot all about Mike's earlier thread...     I definitely want credit to go where it is due and I'll amend my OP to reflect this.  I'm just happy to stand on the Shoulders of Giants.

[GaryHinshaw]

Step 3: threading the LED leads and wiring the base
With the mast prepared and the heads wired, we're ready to thread the LED leads down the mast and wire them up.  Before you thread the lower head, be sure to thread any do-dads onto the mast that will live below the head.  In my case it was the two ladder support brackets:



When running the leads through a side hole, I push through one lead at a time and work carefully.  The threading goes pretty easily, especially with the thin-walled tubing, but be careful not to nick any insulation on the wires as you do.   Before threading the upper head leads, thread the remaining do-dads onto the mast, then thread the leads.  Since I am threading the upper leads straight through the top of the mast (un-prototypical, but I'm ok with that) I just push all 4 leads through at once.  This is very easy with the thin-walled tubing.  Here is what it looks like at this stage:



Be sure to keep track of which leads go with which head.  I trimmed one set to be shorter than the other so it was unambiguous.  Before gluing anything to the mast, I terminate the LED leads to the base and attach the pigtail cables that will plug into the signal electronics.  Here is a shot of the terminated LED leads before attaching the pigtails:



A word about the baseboard layout:  the feedthrough vias that are numbered 1-4 in the photo are connected to the solder pads labelled A-D, respectively, using internal traces that are not readily visible in the photo.  If you need to use step-down resistors for your LEDs, you can terminate the LED leads on the inner pads and solder a resistor across to the corresponding outer pads (where the white lines are).

For the pigtail cables, I'm using CAT5 wire attached to a standard phone jack.  (I tried using phone cable itself, but all of the brands I tried were too stiff.)   Here is an old shot of how I made the CAT5-RJ11 connection:



Just cut a short length of phone cord, poke the wires out of the insulation, and feed the CAT5 wires through in the proper colour code order, then crimp the connector as usual.  This has proven to be very simple and reliable.  Here is the pigtail soldered into the vias on the baseboard:



and a completed signal with a pair of pigtails and jacks, one per head:



Of course, be careful and consistent with your colour coding.  (Duh.)  Now plug the signal in an test all the connections before going any further with assembly.  Easier to repair a connection or replace a head now, if there are any issues



The jacks are labelled with their Loconet IDs.  Next up - completing assembly and preparing for painting.