Monthly Archives: August 2013


I’ve been pondering and worrying how to actually build the open grid that will form the foundation of my benchwork since I started tearing down the old stuff over a month ago.  My old theory was build the stuff to survive WW3, and then oversize by a bit.  Okay, maybe that’s an exaggeration, but I think mountains of 2x4s were some significant overkill for the problem.

My new approach, as suggested by many other modelers, is to use good 3/4″ AC plywood ripped into dimensional strips.  My working theory is to build an open grid where the back will be 4.25″ tall, while the front and cross-members will be 3.25″ tall.

The cross-members will be drilled with several (2-4) 1-1/4″ holes for wire routing towards the front and bottom, and will also have four pocket screw holes drilled.  (I was introduced to the Kreg pocket hole jig earlier this week by Michael Petersen.  It was one of those things that as soon as I saw it in action, I knew I had to have it.  It’s the blue thing in the pictures below.)  Then, using wood glue and a 90 degree clamp, we’ll build up the grid.  Once each piece of grid is built (probably in the garage), it’ll be hauled downstairs and installed on the layout.  The back will be held to the 2x4s with one pocket hole screw and one direct (horizontal) screw.  The lowest level will also likely get a 2″x2″ leg every four feet or so.

The grid will be sized such that the outside dimension, front to back, is 1-1/2″ shorter than the planned finished dimension.  This provides room to attach a 2″x2″ chunk on the front every 16″ or so, and the fascia will eventually mount to these.  That leaves plenty of space for installing switches, indicator lights, cab bus plugs, etc. between the front dimensional member and the hardboard fascia.

I tried building a short piece of this tonight just as a test, and it went spectacularly well.  I’ve included a few pictures from the test.  I’ve got some Iowa Scaled work to do this weekend, and a photo train on the San Luis Central on Sunday, but hopefully I can get some benchwork construction done somewhere in there.

Proto-Freelancing: The Third Copper Crossing

The Copper River & North Western crossed the Copper River three times on its way from Cordova to the mines at Kennicott – once across the delta on miles and miles of trestles, fill, and the occasional truss bridge, once at the famed “Million Dollar Bridge” near Miles Glacier at milepost 49, and then again northeast of Chitina at milepost 132.

Both of the lower crossings were permanent and relatively durable.  The miles of temporary trestle were slowly filled with rock where possible, making for a very sturdy embankment between more permanent trestles and spans.  The Miles Glacier bridge stands to this very day as a testament to its engineering and construction, although it did need a few million in repair work in the past decade to undo 1964 earthquake damage that dropped the last span into the water.

The last crossing, however, was never permanent.  The CRNW built the Cordova-Chitina line to a slightly better standard than the Chitina-Kennicott line, because it initially considered the latter segment to be a branch line off of a mainline that had ambitions of reaching further into the heart of Alaska.  (That, of course, never happened.)  Due to the expense of yet a third large steel bridge on what, at the time, was considered a branch, the MP132 crossing was constructed as a timber trestle with steep approach grades on both sides.  Each year the spring break-up of ice (and sometimes due to glacial lake outflow flooding – more than once a year) would take out significant portions of the bridge, and each year, CRNW B&B forces would drag out the pile driver and put it back.

What actually happened was that the 1938 trestle (the one that carried the last train on November 11, 1938) washed out in the spring ice break-up of 1939, just as expected.  It was replaced by a cable tramway high over the Copper River, which lasted into the 1960s.  In 1971, the current deck girder road bridge was completed roughly where the trestle was located, and the grade was turned into a slightly less primitive road.  The road was improved again in the 1990s, giving us the state of the McCarthy Road today.

Alternate History

Now stepping into the world of the CR&NW that only exists in my mind at the moment, and is turning into reality in my basement…   The CRNW has survived into the present day, and is still hauling ore from expanded workings around McCarthy.  Clearly some rinky-dink trestle that gets destroyed a couple times a year is not going to have survived this long.  The third bridge would have been built at some point in this alternate history – but when?

It’s likely that no bridge would have been built during the WWII era due to material shortages.  Either it would have been built in the 1930s – not likely because of the low demand for copper due to the Depression and cheaper copper available from other Kennecott properties – or in the late 1940s/early 1950s following the war.   Given that Kennecott dieselized (or electrified) many of its steam properties immediately following WWII, it’s reasonable to assume that a surviving CRNW would have followed suit.  So, for our alternate CRNW history, we’ll just say that diesels, mine improvements, and track improvements were all part of a large capital project starting around 1951, as the light rail, untreated trestles, and other such would have been in need of renewal and improvements by this point to keep the railroad safe and efficient.

Okay, so now we’ve worked out an approximate date that would be likely for construction of the Chitina bridge.  Thanks to an engineering diagram that Ron Simpson posted on the forum (full-size image here) a couple years ago, I know what the original proposed (1909) Chitina steel bridge would have looked like.  The planned structure would have a 360′ Pennsylvania truss crossing the main channel on the Chitina end, followed by three 275′ deck truss structures similar to those used on the Kuskulana bridge.  The railhead would have been approximately 94ft. off the river bottom over the main channel, and 70ft. off ground level on the east end of the bridge.  We’ll take this as the baseline.

I’m torn on how to model this.  On one hand, I know what the planned structure was supposed to be.  On the other hand, the railroad lived with the temporary trestle until 1938, and per the reasoning earlier, that temporary structure may well have lived until the very early 1950s.   If a 1950 engineer was looking at the same river, would he arrive at the same conclusion as the 1909 engineer who came up with the original structure?  Or, instead, would the 1950 engineer look to newer, less time consuming and costly construction styles, such as deck girder?  Would that bridge have lasted to present day?  Just to throw another complication in there, would the State of Alaska still have built a road out to McCarthy?  Okay, okay, let’s start working through those one by one..

Various bits of data I can find from the USGS and other qualified organizations suggest that the immediate Chitina area has not received any earthquakes violent enough in the last 60 years to have significantly damaged a large, well-constructed bridge like the one proposed.  It’s simply been too far from various epicenters of large quakes (1964, 2002, etc.)  Given that railroad bridges are typically designed for 100+ year lifespans, there’s no reason that a 1950 bridge should not still be in service today.

Given other large-span railroad bridges built in the 1950s and 1960s, I have to assume that some form of truss would still be the preferred methodology.  (The Wisconsin DOT bridge manual for railroad structures still lists “truss” as the preferred type for spans over 150 feet, even in 2003.)  As for which exact truss style would be used, that’s still up in the air.  I could match the rest of the railroad and use a Pennsylvania for the through truss design, but I

I have to assume that either Kennecott or the State of Alaska would have eventually connected McCarthy into the road system rather than being completely dependent upon rail access.  If Kennecott had built the access road, then the only logical choice is to share the road/rail bridge at McCarthy.  If the state built it, then it’s still an option.  Public data shows that even the 1971 highway bridge that exists today only saw 166 cars/day on average in 2000, and 400+ cars/day today.  That’s pretty low to justify a multi-million dollar bridge.  (I admit, I’m biased on this point.  I’ve wanted to model a joint road/rail bridge ever since seeing the combined BC Rail bridge at Fort Nelson decades ago, so I’m totally justifying myself here.)

My current thought is to go forward with the original bridge design, selectively compressed to fit within the 5 feet of allotted space that I have (the real one would have been around 8 feet), but make it a shared road/rail deck .  It’ll obviously be one lane, so we’ll add some traffic lights at both ends, and interlock them with the signal system.

It does raise the question, however – if the road and railway shared the MP132 bridge, what would they do when crossing the Kuskulana?

Weekend Update – Aug 24-25

Unfortunately I haven’t accomplished as much this weekend as I’d hoped.  I’ve spent most of my time getting projects ready for Iowa Scaled Engineering’s next printed circuit board run.  Some bits were model railroad-related, but most of it wasn’t.  So I don’t even have any cool new toys coming up that I can tell you about.

But hey, our first ad in Model Railroader appears in the October issue, which should be to subscribers shortly.  I’m excited and hopeful that this will help get us a bit more attention.  Seriously, if you’re considering servo switch motors or networked fast clocks, at least give us a look.  The more time I can justify working on model railroad-related projects, the faster the CRNW gets done…

I finally found that one pesky screw still holding the panel that will become the new Cordova yard to the 2×4.  So, upon removing it and cutting one metric crap-ton of wires, I pulled the two sections of plywood out and stored them on the other side of the room.   Then I removed what was left of the old industrial-strength benchwork (heavy 2x4s), the rest of the staging yard, and more wiring.  So now I’m down to bare wall framing again.

The next step – hopefully tonight or tomorrow – is to start noodling how open grid benchwork will attach to the wall frame.   I’ll probably build a couple prototypes, but I want to very quickly get to building benchwork on the wall frames that are complete.  You see, I have too much stuff “in storage” in the basement, and the shelves it sits on are in the way of Chitina.  So the faster I get the base benchwork built, the faster I can slide some storage shelves under it and get stuff in the way, well, not in the way.

CRNW LED Lighting, Round 2

While I was in Indianapolis, I received a call from DHL…  They had a box from China, and I hadn’t been home for the first two delivery attempts.  Woohoo!  The LEDs for lighting the layout arrived (all 500 feet of warm white and 250 feet of cool white) early!

When I arrived back in Colorado (which turned out to be a long and annoying tale on its own – thanks a lot, United…), the other piece of the puzzle was in the mail – an RGB LED strip.  As you’ll recall, my plan was to combine two warm whites, a cool white, and an RGB strip, so that I can vary both the intensity and color of layout light.  Eventually, I plan to integrate that with the fast clock system so that over the course of an operating session, the light will change to model the real change in daylight.

While I want to get going on benchwork again this weekend, I couldn’t go get more lumber because it’s (quite unusually) pouring rain outside and I loaned my truck to a friend.  So, given that I had some power MOSFETs and a MRB-GIO sitting around, I decided to try building a prototype MRBus light controller.  Turns out it was pretty darn easy, aside from some thinking about correct snubbing on the FETs to prevent inductive kick from killing them.  I’ll post a schematic soon, but for now, I’ll show some pictures of the prototype.

In the pictures, you’ll see the MRB-GIO, the breadboard with the power drivers and snubbers, my trusty Lambda power suppy, and then the test LED strip, showing all the different ways it can be turned on and dimmed.  Then there’s also a couple photos of some early tests – I could just send intensities over MRBus to the controller, and lights would change.

This is going to work beautifully…

Brief Update

I haven’t had much time to work on the layout in the past couple days, but I did manage to get the wall built that will separate the Cordova Dock (lower deck) and the Kennecott mine facilities (upper deck) from the workshop.  I also demolished the old helix, as I don’t plan to reuse it in its current form for the Chitina helix.

I also removed most of the screws holding down the existing part of the new Cordova yard.  My plan is to lift the plywood and track structure out whole and replace the benchwork underneath to match new standards.  Then I’ll install risers and cleats, cut down the plywood around the yard, and mount it back in place.  That’ll allow me to remove all of the heavy 2×4 benchwork under it – it’s all being replaced with 1×4 open grid – as well as what’s left of the old staging yard and the old electrical cabinet.

Unfortunately, I have to head for Indianapolis next week, so there won’t be much progress for another 10 days or so.  Then, when I get back, we’re firmly into fall photo charter season here in Colorado, so many of my weekends will be consumed out having fun on the narrow gauge.  As a result, progress may be a bit slow for the next two months.  I’ll try to fill the gaps with historical posts, or posts about layout standards and such things I’ve already decided upon.

LED Lighting Rethought

So last night, as I was trying out various LED strip configurations, I was trying to match my old fluorescent tubes in terms of illumination without regard to overall power consumption, cost, or anything else.

Today, sitting at work, reality smacked me upside the head.  Each deck has approximately 110 feet of linear run, so 220 feet of lighting for the whole layout.  Assuming four strips of LEDs going strong for main lighting, that works out to 4.8A/4ft, or ~1.2A per foot.  That’s 264 AMPS of 12 volt power, or 3.2kW.  Yikes!  Given at a 20A house circuit can only source 2.4kW under ideal conditions, that’s probably 2-3 house circuits just to light the layout. (Remember, any power supply isn’t going to be 100% efficient.  In fact, 75% would probably be a good day.  Then, unless it’s power factor corrected, that’s going to get even worse.)  Given that my breaker box is buried in a finished wall, running new circuits would be painful.

After some tinkering when I got home, I came to the conclusion that two LED strips is probably “good enough”.  That’s only 1.6kW, which even assuming crappy power supplies should easily fit on existing circuits I can tap from the basement.  So, I ordered enough LED strip to complete the layout this afternoon…

Now off to continuing to demolish the old layout and clean out the basement for new construction.

CRNW LED Lighting

My old Canadian Arctic Railway layout was planned to use dual 4′ T8 fluorescent fixtures mounted above each level.  I’d never been particularly happy with that solution, since even the decent-quality electronic ballasts would buzz slightly, and the fixtures were completely un-dimmable.  Plus, I always had some residual concerns about UV light from the bulbs damaging scenery.

I’ve been convinced for some time that LED strip lights were the way of the future.  Small, silent, efficient, easily controlled (they’re only 12 volts!), and available in a wide variety of colors, strip lighting seems to promise massive improvements over under-layout fluorescent fixtures.  The downside is the need for very large 12V power supplies and – until recently – the pure cost of it.  Fortunately, with recent advances in LEDs and direct access to Chinese vendors via eBay and Alibaba, those two concerns have largely gone away.

In addition, the infinite dimmability of LEDs and the variety of available colors opened up the possibility of something I’ve wanted to try for a long time – modeling the day/night cycle!  The basic idea would be to have the layout lighting adjust the intensity and color blend based on the fast clocks, so that as the day progresses, dawn grows into mid-day, which fades to afternoon and the “golden hour”, which gives way to the “blue hour” and then eventually night.  With the right mix of colors and some simple power controllers attached to the MRBus layout network, such a thing should be relatively simple.

I ordered three 5 meter spools of 5050 (5.0mm by 5.0mm component) LED strips from China via eBay (my secret source…): one warm white (3200K), one cool white (6500K), and one blue.  After two weeks of waiting on the proverbial slow boat from China, they arrived this afternoon.  I wasted no time getting them downstairs, finding a suitable 4 ft. board (same as the fluorescent fixture), and cutting off strips to test.

I finally settled on 3 warm white and 1 cool white strip, when lit together at 12V, gave me similar lighting to the old tubes.  One cool white and one blue strip, lit at about 10% power, gave me a reasonable “night”.  No, it’s not as dark as a real night, but it’s a decent approximation for my operators.

No mix I could find, however, gave me a suitable sunrise or sunset color, so I’ve ordered another strip of 5050 RGB LEDs.  I’m hoping that by using the warm and cool white strips as the “bulk” light, and then “flavoring” it slightly with color from the RGB strip, I can get satisfactory colors for all times of day.  More to come once those LEDs get here…

I’ve included pictures of this evening’s test below.

Also, as a teaser, the first part of my CR&NW’s roster showed up today…  Meet the 600s, a set of 6 SD60M three-window units that will eventually be numbered 600-605.  (Yes, there’s a spare shell there, just in case I’m not as good with the ol’ airbrush as I used to be.)  Obviously with the real CRNW calling it quits in 1938, they never had diesels.  However, as I’ve mentioned, mine will be set in the present day.  In my world, the SD60Ms showed up on the property in 1990 to replace the aging fleet of SD9s that had originally replaced steam in the 1950s.  But more on my fictional modern day roster another day…


And Construction / Destruction Starts…

Here’s a last look at the start of the Canadian Arctic Railway, my fictional bridge line from Fort Nelson, BC, to Anchorage, AK.  I started planning this line almost a decade ago, and began construction shortly after buying this house with my wife.  Then she became my ex in 2008, and the railroad went into an seven year holding pattern.

As of tonight, CR&NW plans in hand, I started tearing down pieces that aren’t staying in the new design.   The helix came out, and I’m in the process of removing the under-level staging yard.  (The staging yard is getting a new life as a testbed for some Iowa Scaled Engineering projects, so it’s still around, just shorter.)

Tomorrow or Sunday, I’ll remove the upper deck and extend the wall framing out through the area the helix once occupied.  Then comes the task of replacing the old heavy upper framework with a new, lighter open grid system.  The bottom through this section will likely remain the same – Fort Nelson is being rebuilt as the main Cordova, AK, yard.  All new construction on the bottom, however, will follow a similar lightweight construction method.  I’ll also be installing a valence above the upper deck, and shelving below the lower deck.

And so the journey begins…