Gibbon, Cozad & Western - "The 100th Meridian Line"

[GaryHinshaw]

Mike, I confirm that the width of the Dick's LED is 1.3 mm, but I find the height to be more like 1.7-1.8 mm.  (The leads are soldered along the long axis, so it's a bit tricky to get a consistent measurement.)



It would be amazing if you were able to get somewhere with an SA head, as this part is the most problematic piece in the Showcase kit (which is otherwise outstanding).  For reference, the i.w. of the Showcase head is just a shade over 1.2 mm, so the LED does not really press fit into the head without a bit of reaming, and I don't have a good process for that yet.  In the heads I have assembled to date, I have been trying a combination of filing the walls (hard to do on these small parts) and mounting the LEDs at a slight angle inside the head.  The o.w. of the head is 2.0 mm, so the wall thickness is about 0.4 mm.  (FYI - I don't trust my numbers much below 0.1 mm.)

The LED thickness, including the leads, is about 0.8 mm, so ideally, if the back ~1 mm of head had a slightly thinner wall thickness (0.3 mm instead of 0.4 mm) the LED could fit perfectly in the back of the head and be stopped by the thicker wall in the front in a very controllable way, and the leads would comfortably clear the cover plate on the back of the head.

I'd be perfectly happy with a visor thickness of 0.4 mm or so.  The Showcase part is about that thick and it honestly looks fine to me.  They also have an etched brass extension visor, but I have not been bothering with that myself.

I'll be watching with keen interest! 

[C855B]

Excuse the thinking out loud...

Making some headway. I dug the signal out of the garden shed and discovered it to be a US&S H-5, acquired from an ATSF signal bridge. It is a different case style than the Showcase, which represents the US&S H-2. Tehachapi, etc., is going to be mostly H-2. The signal at the museum - ex-Frisco - appears to be an H-2, thankfully. The H-5 O.W. is 11.5" and by what I can tell from the ground and internal photos found online, the older case is a little narrower. As soon as we get the ladder back up I can confirm; it's off for safety reasons until I fasten the Plexiglas anti-climber to it.

That raises two issues. First, if the H-2 case is in fact narrower - let's say 10" - then the Showcase version @ 2 mm is 20% over-width. Yeah, that usual N scale compromise thing. I think my tack at this point is to make the replacement case fit the electronics. Let's say I can hold 0.010" (0.25 mm) on the shell, so if the 1.3 mm of the RR-cirkits LED holds across the batch, that means the O.W. has to be 1.8 mm, corresponding to ~11.5". Oops, that's the width of the H-5. [...sigh...]

Anyway... I'm going to do both styles since the H-5 is on Santa Fe signal bridges, my main need for searchlight cases, but I do need H-2s for Afton Canyon and Meadow Valley Wash. I'll fudge if I have to on the H-2 width. Internal height is no problem. Also, let's see what I can do with a "perch" for the LED boardlet, as I did something similar with the 3-color vertical for mounting the individual LEDs.

I'm going to spend more time on redoveryellow.com to get more rear views. One difference I'm aware of between the H-2 and -3/-4 is a bulge on the lower part of the rear cover to clear a quick-change connector. (Example here on the Santa Fe approach to Mojave.)

OK... so, with or without Phankills? (Without. For the most part. I didn't notice any on the Tehachapi shots at redoveryellow.com, but I'll look again.)

[If you think the diesel-spotter rivet-counters have a monopoly on detail minutia, think again. Signals are brutal.  ]

Which reminds me, I need to get off my butt and take some detail shots of the GRS SAs remaining on the BNSF before Adam @draskouasshat and his buddies pillage what is left during their PTC upgrade project on the Beardstown Sub. It's going to be a pill drawing the art-deco visor profile on those things.

[GaryHinshaw]

I feel like this project is in excellent hands now.   

[C855B]

Well... let's wait for results first.   

For clarity, H-2:



And H-5 with Phankill:



That H-5 is heavy. My guess is somewhere between 150-200 pounds. OOF!

[draskouasshat]

You mean one of these?
And yes, ill be in centralia at the start of the year it sounds like working south.
  [ Guests cannot view attachments ]

[C855B]

Yeah, Rodney @rodsup9000 mentioned we were on your project calendar. Stop on by anytime.

[C855B]

Well... let's wait for results first. ...

Preliminary report: H-5 proof-of-concept just out of the Photon. Everything printed nicely on the first pass... I'm sort of astounded, actually. The LEDs arrived today as well. I tried test fitting with a mushy uncured head and everything so far appears as planned. I'll know more in the morning after curing. It's good enough to start adding details to the artwork like the alignment sights, hinges and locks, the (separate) back, and a mount socket compatible with the Showcase bracket.

I'll work on the H-2 case once the ladder is up for square-on pix and tape measure access. How many dozen will you need? 

[GaryHinshaw]

I'll work on the H-2 case once the ladder is up for square-on pix and tape measure access. How many dozen will you need? 

Fantastic!  Can't wait to see some test shots.  I have 68 signal heads on TBC, so I'd be down for ~6 dozen plus some spares.   

[C855B]

In the interim, to further whet your appetite here is the CAD preview for the detailed H-5:



I'll run it through the slicer in a couple of seconds here so it'll be ready to go after I have my coffee in the a.m.

[C855B]

Progress comes with a "but...". Good results from the proof-of-concept batch, and Dick's chip fits with no carving, albeit at a small angle since it's ever-so-slightly taller than the case innards. The #36 wires create a bit of a problem with their noticeable thickness and is going to make routing out of the case a chore in micro-management. Moving the conduit port to the center of the back will alleviate some of that; the test was off-center per the H-2 location. Couldn't do a photo of the chip in the case without gluing; all those springy wires were working against holding the case, the chip and operating the camera.

The "but"... as you will see in the picture, my observation that the Showcase head is 20% oversize is more true than I thought. I double-checked the finished, cured dimensions of the H-5 tests against 1:1 and it is accurate; surprising that the Photon holds fine dimension so well. The problem now is my artwork does not work with the Showcase etched bits such as the target and brace. Since our objective here is to minimize our work, I need to swallow any exactness and prioritize towards "aftermarket" to work with other products. No interest here in getting into etching my own suite of trim-out accessories, at least not yet.



Back to the drawing board, so to speak, but a lot learned. The redraw should be quicker.

[peteski]

I think that slightly oversize head is a good compromise between realism and being fairly easy to produce.
Just look how far we have come from the old NJ International signals with those Grain of Rice 12V incandescent bulbs sticking out of the front and back of the head (which was just a brass tube).

[GaryHinshaw]

That looks really great Mike.   I know what you mean about fiddling with the leads - wait until you have to string 8 of them through a mast.  But the task will be a bit easier if the LED at least fits well and securely in the head.

I'm not nearly the signal connoisseur that you are (I mean that as a compliment), so I would be perfectly happy with a clone of the Showcase head that held the LED with less effort.  But by all means, you should do you.

BTW, are the Richmond Controls LEDs pre-wired?  If so, what kind of leads do they have?  Just curious.

[C855B]

Thanks! And I agree about the challenge of threading two bundles of those wires through the tubing... it just ain't going to work with three heads, nosireebob. The Richmond chips come with #38 wire, and I do the same. It is funny/interesting that one size up makes that much of a difference. A couple of years ago I tested pushing three bundles of four #38 through a Showcase mast and it sort of worked, tho' it wasn't easy, and due to other complexities with a three-head mast I never finished the build.

I'll work on the compatible version. The modeler wants to be accurate, the layout constructor wants it done, the easier the better. So I'll put on my "layout constructor" hat.

[GaryHinshaw]

I'll work on the compatible version. The modeler wants to be accurate, the layout constructor wants it done, the easier the better. So I'll put on my "layout constructor" hat.

I know what you mean about the layout constructor's hat.  It's a hard hat.

If you make your signal installation modular (Wutter's bases, plugs, etc.), you can always go back and redo them when you have more modeller's time available to tinker with etchings and such.  In fact, if there were a way to fashion micro plugs for 36 ga wire, you could make just the heads modular.  (Not sure how one would practically do that though.)

I can't stress how grateful I am that you're willing to take this on, but I need you to wrap it up soon, so you can start on the next 3d project I have in mind.   

[C855B]

OK, Showcase-compatible version of the H-5 is in the soup. "Ah, crap!" moment when pulling into the driveway back home - I had scaled it up 20%, thinned the visor back to where it was... and then it dawned on me that the hole for the Showcase bracket scaled-up, too, so the hole is now too big. No matter, this batch will be done in about 45 minutes and I'll head back to the studio to put it in the curing oven for test fittings in the morning. Fix the mount and anything else that pops up, and I think we'll have something. Then the H-2 will mostly be cosmetic adjustments to the artwork.

"...next 3d project..."? Ruh roh.