Tehachapi, BC

[jereising]

I can't believe - with all the talent and entrepreneurship shown on this board - SOMEONE has not yet simply etched a bunch like Traincat did...I doubt he ever copyrighted it. 

Or maybe no one thinks it's worth doing. 

I was fortunate enough to build from his kits (had to roll my own hoods) and tried to order a couple more frets but the site was down by then.

It's easy enough to do - tubing is very available, and ladders and bases share a commonality with the searchlights.  So all that's really needed are the targets and hoods.

So why hasn't SOMEONE picked up the ball?

[C855B]

... So why hasn't SOMEONE picked up the ball?

Two things - first, the modeling challenge. Signals, especially color verticals, are more than just a target and hoods/visors on a stick. They have a back. I have yet to see anybody try to accurately portray the backs in a working 3-color signal. Showcase raised the bar big-time with their SAs. There are significant appearance differences between a US&S one-unit casting and a built-up Safetrans (1st or 2nd gen), f'rinstance. If you're going to do a color vertical, you'll also need at least two versions of targets (24" and 35"), three visor types, two or three hood types (old- and new-style "Vaders"). Then there's the challenge of attaching visors to the target.

Second is wiring. SAs work because there are a couple of entrepreneurs who have managed to contract (probably overseas) for a single-unit LED with attached wires. The tough job there is covered. Color verticals are going to need some manner of circuit board to place three separate LEDs and secure the wiring. Near the end of Bob's efforts, he had devised a cool scheme to make the target out of thin circuit board material, with drilled holes to set the LEDs into, making the target/lights a one-piece system to attach visors or hoods and backs of your choice. I've been thinking that approach through, the greater challenge being thin enough PCB material to not look clunky as a target.

Needless to say I've been dwelling on this a lot. The question is designing something that can be reasonably mass-produced and offer enough variety for the disparate prototypes.

[jereising]

I'd sure settle for less than perfect - anything is better than nothing.

I seem to recall a fellow over on the product board that was developing something like you describe but apparently didn't get enough orders to make his project go - then he sold his company, so that's likely dead.

And I was satisfied even though the backs of Bob's don't look great with all those wires, still, working signals were worth it...

And apologies to Gary for hijacking his thread

[GaryHinshaw]

No worries Jim, it's all topical.

RE era mixing: the longer I work on this layout, the more I find myself invoking Rule #1.     For example, see the windmill discussion above: I have no problem bending time (to a point...) for either practical or aesthetic reasons.  RE signals, I love these searchlight kits, and their prototypes -- especially the old signal bridges -- so I'm happy to use them.  That said, I am keeping the installation modular, so it would be relatively simple to swap in new signals when that brave soul does come out with a high-fidelity Vader kit.

And I do admire what Jim has accomplished with his signal build.  It adds so much to a mainline-style pike. 

[Sokramiketes]

Are you programming the signal logic into the CirKits boards, or running JMRI signal logic where JMRI has to be up and running? Logix?

I'd love to read about how someone else went about the programming.

[nscalbitz]

A little more progress to report.  Here was the state of the Loop shelf as of a few days ago:



You can see the tracks from the desert shelf coming in from the right...
...
Scenically, this is the most challenging section of the layout, but it is starting to take shape.  Thanks for looking - more soon.

Whoa there Gary- that's such a glaring background change in desert colour. Is that intentional? In your 'pièce de résistance' most dramatic scene?

I drove the area in 2002 visit so (late September) maybe variations, but that drastic?
I only ask because local layout I used to scenic used ancient (20 yrs+) disparate backgrounds, PVA-glued down and always irked the look of the 'theme' we were trying to achieve.
davew

[GaryHinshaw]

Whoa there Gary- that's such a glaring background change in desert colour. Is that intentional?

Not at all.  

If you're referring to the backdrop photos in the above shot, those are temporary, as is the tan paint.  If you're referring to the darker foreground section, that was an earlier piece that I painted with tempera paint - the only shade of cheap brown I had at the time without going to the store.  I have since painted it over with the basic tan colour:



(Tip: tempera makes a terrible primer for latex paint.)  Since this shot was taken, I have made more progress on the remainder of the shelf: the central hill is now mostly finished and painted tan, and I've started shaping the left side.  But I decided I better pause to finish up the signal wiring before I go much further than this, otherwise I'll be cursing myself for making my life more difficult. 

Eventually this will all be covered with a finish coat and, of course, real ground cover.  Colour matching will be very important at that point.

Are you programming the signal logic into the CirKits boards, or running JMRI signal logic where JMRI has to be up and running? Logix?

I'd love to read about how someone else went about the programming.

I'm actually using CATS to control the signal logic, so I need to have CATS and JMRI running to activate the signals, but I'm fine with that.  In CATS, I define the signal discipline for each block (mostly CTC) and I define masts that refer to specific signal heads in my JMRI panel.  In JMRI, the Signal Heads are defined using the DCC Signal Decoder definition.  In the Signalman, I tick the Mast radio button, then each mast (A,B,C,D) controls a different LED (Signal Head).  I'd be happy to start a thread with some screen grabs showing how the various elements are configured if there is interest.

I do employ Logix to control the approach lighting for these signals (IF blocks X,Y,Z are occupied then Heads A,B are lit, etc.).  There is an option for approach lighting in CATS, but it doesn't work with the configuration I am using.  Rodney (the CATS author) recommended Logix was a better way to go for my application.

-gfh

[jereising]

Just refreshing our memory - the company working on signals was Cutting Edge Scale Models and was acquired by Out Of The Box Models.

When asked how the signals were doing - back on September 6 - the answer was

"I am still working on trying to find an avenue to cost effectively manufacture the signals. The electronics are a different story, I need to hire an electrical engineer to figure why the board doesn't work, and then fix it, then find someone to make it."

[DKS]

I'm actually using CATS to control the signal logic, so I need to have CATS and JMRI running to activate the signals, but I'm fine with that.

Hey, I'm fine with it too.



Guess I have my future layout signaling covered... 

[Cajonpassfan]

Lol, here's one back at you🙀
Of course, this one isn't allowed in the trainroom any more; he's been known to chew on a water tower and rearrange my freight cars ...
Otto [ Guests cannot view attachments ]

[Mark W]

I did a little proof-of-concept on 3D printed signals over in the weekend update, for those interested. 

https://www.therailwire.net/forum/index.php?topic=46258.msg603331#msg603331

[GaryHinshaw]

LOL @ CATS.

I did a little proof-of-concept on 3D printed signals over in the weekend update, for those interested. 

https://www.therailwire.net/forum/index.php?topic=46258.msg603331#msg603331

This is a most welcome development Mark!!  I'll follow the 3d printing thread with interest and contribute where I'm able to.  Thanks for taking this on!

In the Picture is Worth a Thousand Words category, here is another mini-update:



On the single-head signal, I terminated the magnet wires on Wutter's board and added a pigtail to connect to the control circuitry.  I decided to go with phone cable because it's super simple and readily available, and the jack let's me unplug it for repair or maintenance.  Electrically, it works beautifully.  The only downside is that the cable is a little stiffer than I would like, so I'll need to add some strain relief so it doesn't pop the signal base out of its little holder.  I haven't quite sorted out how to mechanically mount these signals, but I'm picturing some kind of short tube that is rigidly mounted to the base which could secure the cable and insert into the scenery.  Suggestions welcome.

The other item in the photo is my first attempt at a dual-head signal.   

I spent more than an hour trying to thread the second set of magnet wires through the tube after threading the first lower set through a side opening.  I succeeded in getting 6 out of 8, and there is definitely room in the mast for all 8, but complete success has so far eluded me.  The task is made more challenging by the need to have all the hardware threaded on in proper order before wiring can commence, so the process is rather unwieldy.  If I can't succeed with a different order of operations, I may have to look into some ultra thin-walled SS tubing.

Here's a few quick shots of the completed signal in action.  LEDs never show up well in photos (even in HDR mode) but the colour and brightness look great in person.

[nscalbitz]

Not at all.  

If you're referring to the backdrop photos in the above shot, those are temporary, as is the tan paint.

'Twas a moment of terror. Sorry I missed any reference to mock-up stage earlier.

If you're referring to the darker foreground section, that was an earlier piece that I painted with tempera paint - the only shade of cheap brown I had at the time without going to the store.  I have since painted it over with the basic tan colour:

Anything to avoid the gawdawfull use/ vision of brown latex ever again!
Congrats on the effects-  cheers d

[ednadolski]

Man those sure look great.  There are times when I feel the pangs of regret from giving up on my Tehachapi project, but it sure helps to see that you're doing such fantastic work with this. 

Ed

[Wutter]

On the single-head signal, I terminated the magnet wires on Wutter's board and added a pigtail to connect to the control circuitry.  I decided to go with phone cable because it's super simple and readily available, and the jack let's me unplug it for repair or maintenance.  Electrically, it works beautifully.  The only downside is that the cable is a little stiffer than I would like, so I'll need to add some strain relief so it doesn't pop the signal base out of its little holder.  I haven't quite sorted out how to mechanically mount these signals, but I'm picturing some kind of short tube that is rigidly mounted to the base which could secure the cable and insert into the scenery.  Suggestions welcome.

The other item in the photo is my first attempt at a dual-head signal.   

I spent more than an hour trying to thread the second set of magnet wires through the tube after threading the first lower set through a side opening.  I succeeded in getting 6 out of 8, and there is definitely room in the mast for all 8, but complete success has so far eluded me.  The task is made more challenging by the need to have all the hardware threaded on in proper order before wiring can commence, so the process is rather unwieldy.  If I can't succeed with a different order of operations, I may have to look into some ultra thin-walled SS tubing.

Hey Gary, the 3D printed frame is intended to be glued to your benchwork or scenery and then covered with ballast, and to have the PCB be press fit or held by a small amount of tacky glue to the frame. It usually takes about a 7/16 or 1/2 hole under the frame to let the connector fit through.

For cabling, I like to use 4 POS 2.54 pitch cables that I just cut in half and solder the wire end into the board, it stays flexible but is easy to plug into standard headers on a breakout further down the line. I basically use one 4 POS cable per head on a signal, like these: https://www.banggood.com/5pcs-4-Pin-20cm-2_54mm-Jumper-Cable-DuPont-Wire-For-Arduino-Female-To-Female-p-914008.html

What size magnet wire are you using to thread down your tube and are you using 1/32 or 0.8mm? Using 38ga already takes a bit of work to get 8 wires in, so if you're using something bigger, it might be worth it to look at smaller wire. The question about the tubing is if you're currently using 1/32, I think the ID of the tube is smaller than metric tubing which comes in 0.8mm OD and 0.6mm ID, so that gives you a little bit more room on the inside as well.