Upon my completion of the benchwork from Castlegar to the east switch of the summit siding of Farron, I began looking toward the next section – the 11 foot pair of sections encompassing Farron siding and a delightful little pond on McRae Creek. then it hit me – If I built that, how would I get lumber from the saw to the rest of the layout? For you see, though it’s fairly high off the ground, I also realized that I could add some house-like framing made of 2x4s to give a huge amount of stability to it. Can’t fit lumber through THOSE openings! So even though I already had the rest of the lumber cut and was maybe 2 months from having a fully operational grade to the summit and a place to turn my trains…
…I decided to build the rest of the benchwork first. ‘Cause then I could use my saw easily and stuff.
This meant a big ol’ pile of lumber. What you can’t see are the 40 or so members already cut and hiding under the layout.
The lower deck will be standard L girder, the same as under Castlegar. The upper deck will continue the design that I’ll outline in the next post. I already managed to get all the lower deck done save for legs and braces. Quick, huh?
This weekend’s project turned out to be more of a doozy than expected.
Step 1: Clean the wheels. Spread regular metal polish thinly on the rails, ran a loco in place at full speed for a minute or so before thoroughly cleaning the wheels and track off, spreading a thin layer of no-ox-id on the track, running them in circles for a bit, and cleaning everything thoroughly.
Step 2: Program the decoders for lighting effects.
What I wanted was something more prototypical than the vanilla “Headlights on with F0, front on going forwards, back on reversed” approach. I decided that I wanted front lights to come on with F0, dim with F4, and rear lights (when applicable) to come on with F1. They would stay on regardless of direction, because what engineer says “Back half a car, better turn this light off and -get up- this one on”. Really! Fortunately for me, I really only have 2 kinds of decoders. Digitrax 163 series (FX3), and TCS. This meant that I only had to spend hours wading through manuals instead of days…until the addendum at the end for NCE.
Starting with Digitrax, I decided on F4 for dimming because you can’t remap that.
The digitrax decoder manual is horribly outdated. Don’t even look at it! It talks about CV61, which hasn’t been relevant since the FX3 163s came out a decade ago. Somehow, digitrax hasn’t updated it to include what you need for 163 and 165 decoders!
Now, it took me a while to figure this out….the first thing to tackle is function mapping. CV33, 34, and 35 control the applicable functions. CV33 is F0 on your cab (controller) in the forward direction, CV34 is F0 in reverse, and CV35 is F1 in reverse. Function 1 on the decoder is the white wire to the front headlight and function 2 is the yellow wire to the rear headlight. (I will not be abbreviating these to F1 or F2.) These functions are the same whether there are visible wires or the LED is mounted on the locomotive’s circuit board or decoder. Thanks, NMRA!
In order to control these function wires seperately, you must change CV33-35. By default, CV33=1 (white wire/front headlight on in forward only) CV34=2 (yellow wire/rear headlight on in reverse only) and CV35=4. (green wire in either direction, not applicable in this circumstance.) We want function 1, the front headlight, to come on regardless of direction, so we keep CV33 at 1 and also set CV34 to 1. This means F0 does the same regardless of direction. The rear headlight (yellow/function 2) is function key-less, so we map it to F1 by setting CV35 to 2. If you have more lights, you can map F3 on your cab to be the rear light by changing CV37 to 2 instead of CV35
You can do more calculations including more CVs/functions using this digitrax link
Stopping it auto-dimming on reverse was, as it turned out, impossible. Therefore, if you want a headlight that dims on command with F4 but also dims on reverse, program CV49 to 104. I ended up setting it to 0, since I’d rather have constant headlights all the time than dimmable ones that mess up when you reverse. This means all my future decoders will probably be TCS…
Moving on to TCS (Train Control Systems): TCS decoders generally have more lighting options. They have become my favorite decoders, being only slightly more expensive with more functionality.
In TCS decoders, the front and rear headlights are controlled by 49 and 50. To set the front headlight to manual, set CV49 to 32, and CV50 to 32 for manual reverse. To have the front headlight dim on command from F4, set CV49 to 40. Next, I remapped the reverse light from F0 to F1 by setting CV34 (reverse light function wire) to 4. Easy peasy! I didn’t bother with any rear dimming.
The only complication I had (besides the units with TCS CN decoders refusing to dim) was my RS-10 – It’s set up for long hood forward operation. This means that front is rear and rear is front…the decoder is programmed for the correct direction of travel, but that doesn’t change what is physically wired to what. Since it’s TCS, not a problem. Set CV33 (what would be the front light) to 4, and CV34 (rear light come front) to 1. Next up, this means that F4 dims the rear headlight, so switch the CVs for 49 and 50. (CV49=32, CV50=40). Voila!
I hope this answered a few questions for you. I do have more DCC projects coming up, so keep tuned!
Update: NCE (North Coast Engineering):
It took me a while to figure out NCE decoders. I didn’t have any when I originally wrote this article but recently installed an N12A0 into my Atlas SD7 – A Great Northern unit that runs long hood forward.
Important: NCE does not use CVs 49 and 50 to control anything. They configure functions with CVs 120 to 128 instead.
As with other decoders, CV 33, 34, and 35 map functions (lights) to the function buttons. CV33 is F0 forward, CV34 is F0 reverse, and CV35 is F1 in either direction. By default, as with Digitrax and TCS, these are CV33=1, CV34=2, and CV35=4. To control the front light with F0 in both forward and reverse, set CV33 and CV34 to 1. To control the rear light with F1, set 35 to 2.
What’s that? You came here for directional lighting? Good, me too. This is where I had the CV49/50 problem. CV 120 and 121 act the same way as 49 and 50 on Digitrax decoders. To stop lights from being directional, just set 120 and 121 to 32. If you try to set one to 40, though, it just dims on reverse instead of with a button. Dimming isn’t something I plan to figure out in the near future though.
Special case: As I mentioned, this was a long hood forward locomotive I was programming. This means I had to do a number of things. After setting CV29 to 39 while consultingthis calculator (and being very sure to tick the long address box, otherwise the locomotive wouldn’t respond) I tackled the problem of the light outputs not being reversed with the direction. It was easy, though, just had to remap the functions. CV33 and 34 control the short hood, which on a LHF locomotive is the rear end, so I had to set them to be controlled by function 1, setting CV33 and 34 to 2. The long hood light, now the forward, is controlled with F0 by setting CV35 to 1.
Next up may be MRC, as I have that one locomotive….time will tell.
I have barely touched the model railway since I left.
I have excuses. Want to hear them? Well, first and foremost, this is my current project:
…Yep. That’s a lot of work.
Second, my computer’s power supply bit it and I decided that it would be more economical to buy a new most-of-it than to replace the PSU, and I don’t regret this. I can actually photoshop now! Speaking of computers, one related thing I’ve been up to is that I got myself a 9-pin serial cable and USB adapter so I can program through JMRI. I’m leaving speedmatching until I have a speedometer, and we’re going to see how that fares compared to the HORRENDOUSLY PAINFUL method I have now.
Now, let’s make a real post: I had 4 major excursions from Creston to other parts of the Kootenays, and I’m doing them in order. First, this post will contain photos of the hotshot from Nelson to Trail. Next, we biked the old Nakusp and Slocan from Nakusp to Rosebery (the barge landing) and back. We were going to do to New Denver/Denver Canyon or Three Forks, but it was incredibly difficult going between summit lake and Rosebery with bushes and washouts and ponds on the trail! The next post will contain a very exciting bit, a cabride on the hotshot from Nelson to Trail and back! This was done through proper channels with paperwork and yada yada so nobody will get in trouble 😉 Post #3 will be the Columbia and Western post, including photos from our bike ride from Castlegar to Grand Forks, a really wicked two day excursion that went very well.
Now, I planned this trip referencing google maps and old photos, trying to guess what would be good, and what wouldn’t…and it mostly wouldn’t. Tree and bush growth has made 3/4 of what were formerly good shots unusable..
Anyway, to the photos!
The crew is called at Nelson at 0630. I arrive at this time also, as a test run for the cab ride. Don’t want to be late for that! I look along Government road for an overview of the yard which once existed, encountering only trees. Moving on, I decide to go to Taghum. Taghum is the set of bridges that I will be tackling next in N scale, and the word itself means ‘six’ in the Ktunaxa language. The wait for the train gives me an opportunity to take reference photos. While waiting, I was passed by a couple on a tandem bike. When the train showed, I was surprised to see the power it had! This means that they’re a little short of SD40-2s out there, probably had several in the shops at the time…
There’s no ore for Trail on this train, but carloads of coal for the smelter are visible.
This photo is looking towards the aisle on my layout. I will not be modeling the road bridge.
Next, I tried to find a shot at a fill I’d like to model when I have more room between Nelson and Castlegar, but I encountered my old enemy: Trees.
Anyway, between there and the junction at South Slocan, there were a number of shots that I tried to get which were overgrown as well, most notably at Cora Linn dam, the site of a derailment this spring.
South Slocan has a remnant of the Slocan sub (the one to the barge) coming off at the east end of the wye. The old station foundation is just behind me in this photo. This was meant to be a storage track, but like many of the remaining sidings on the line it’s had its switches pulled because there wasn’t anything to store…and it shows. Not far down it becomes the Slocan Valley rail trail, which is quite well maintained.
A little bit to the west of there is the bridge over the Slocan river, at the halfway point between Nelson and Castlegar. I don’t have any intention to model this unless I end up with an enormous layout.
Onward from South Slocan are the upper and lower Bonnington dams.
Then I went to Thrums, which was a 40ish car siding. It became a storage track, and then was torn out after a small creek washed out the culvert a quarter of the way down it. Not an interesting photo, but it gives you some feel of the area. A dog was barking at me for the entire 15 minutes or so I was waiting…and flagmen on the road had me worrying about the next shot.
As it transpired, I didn’t need to worry about making it to the next shot, because the next shot didn’t wait for me. I leaned out as far as I could (After hopping the fence with the danger: cliffs sign) but alas, 3 years can do a lot to a shot. Compare this to the header image!
This is the Brilliant dam, the last dam on the Kootenay river before it joins the Columbia. Following the railway west, you can find the community of Brilliant, populated mostly by Doukhobors – you might remember them. Refresher: They’re the ones who bombed a train at Farron in 1924, blew up the power line across Kootenay lake in the 50s, and gave CP an excuse to end passenger service!
Past Brilliant, the line descends to the crossing of the Columbia river. This bridge was completed in 1900, but before that the barge from Robson to Robson West was the railway link across the Columbia. While waiting on the highway bridge, the couple on the tandem bike passed me again!
This shot features the original swing span (on the circular masonry pier) which is still in place, gears and all!
It might seem to be in a strange spot, but that’s the deepest part of the channel. Steamships regularly plied the Columbia around the turn of the century, and the last sternwheeler operating on the Arrow Lakes was retired in the 1950s.
Here’s a shot showing the whole bridge:
The line loops around behind that ridge, where the Kootenay river lies. Castlegar station is very close to the right, so I drove to the south end of the yard. This is where the hotshot sets off freight for the pulp mill, taking only cars for the Trail smelter and the Warfield fertilizer plant with it.
Here’s a shot of the south end of the yard, which is in the hallway on my railway.
Those are Trail cars on the main track. To remind you, here is this spot on my railway The 3 tracks on the right are staging, while the other 3 are prototypical.
Anyway, the train eventually dropped off the celgar freight on two tracks.
South of Castlegar, there are currently many 10MPH slow orders in place. This meant I had to wait an excruciatingly long time for the train to catch up. Incidentally, on this date the line’s speed limit had been reduced to 15MPH from 20MPH due to the derailment at Cora Linn in the spring.
Leaving the community of Blueberry Creek, the line parallels the highway over a large fill and through a cut. This alignment was created in the 1960s during highway improvements, and the fill was constructed to serve both modes of transportation – the line originally went up and around the valley of Blueberry Creek. I find this nifty. The couple on the tandem passed me again.
The next stop was the big hill at Poupore. If I model the line to Trail at some point, this has to be a feature! From Castlegar to Poupore, the line’s on a glacial steppe, but this ends at China Creek. Between Poupore and Genelle, the line cuts across this fill sharply and then enters a curve, going down an immensely steep 3.6% grade. The top of this hill has been undercut dramatically by China Creek, and railway engineers are worried about the entire hillside simply slipping away.
For photos, predictably, trees ruined the best one. Here’s all I could manage.
The train is now descending the steepest part of the grade and the dynamics are howling. The screeching wheels note that the sharp curvature is helping hold the train, but as it crawls past at 5MPH you can’t help but wonder how far it is from a runaway.
Past Genelle and through Birchbank, the grade evens out as the railway moves away from the highway, so we rejoin it at a somewhat boring spot where the tandem passes me yet again!
Further down the line, the valley becomes drier and the line climbs onto another bench, affording better views.
Rivervale is a community just upstream of Trail, and the train will shortly be entering the yard of Tadanac.
A hundred feet from the switch of Tadanac yard in Trail, the line crosses a gulch called Stoney Creek. Since this is on an uphill, in the old days the brakeman would run across the top of the train and run along the bridge ahead of it to throw the switch so the train wouldn’t have to start again! After this, the train enters the smelter yard (very much so private property) and I went on my merry way to conduct my business. Perhaps another day I shall follow it back.
This was not the end for the day. The smelter may be where the hotshot turns, but another interesting operation comes out of the other end of the yard! The Trail switcher makes a daily run up the hill to Warfield, where a fertilizer plant still generates daily carloads. Unfortunately, spring is the right time to take a look at this line, since that’s when most of the fertilizer is produced. On this day, the 2 GP38-2s didn’t have to push past run 4 to get their cars up the hill. The stack in the background is the smelter.
“4 cars and notch 4? What are you smoking?”
You may very well be asking this, and, well…
This is one hell of a hill. It’s still the CP Rossland sub, and the original line built in the 1890s as the narrow gauge Columbia and Western. It originally went past Warfield through a series of switchbacks to Rossland, but everything past Warfield was abandoned around 1960.
What remains today is the most torturous class 1 track in Canada. With curves with up to 20 degrees of curvature and a grade of 4.2%, you need a locomotive for every 2-3 loads. During high points in the 1980s it wasn’t unusual to find 5 GP38-2s on the head end and a pair of GP9s pushing on the rear of a 20 car train. Of course, I’ll return in spring to get this for myself! For now, the 4 Procor cars are the only reliable traffic into the plant.
Above here, the train loops around 2 horsehoe curves to bring it back 100 feet above me and just behind, in less than a mile! Unfortunately, this has also grown over, leaving me at the grade crossing where the Rossland sub used to pop off to the right.
A neat little thing which I would like to explore more, the train goes through a tunnel of tall trees. Next time I’ll pull ahead more so I have a chance to use my telephoto…standing in the gauge is bad mmkay?
A videographer, Matthew Robson, was also following the line that day and the crew seemed to enjoy it. I wasn’t hassled at all by security, but he was – perhaps because hiding in the trees is more my modus operandi. We started chatting in the parking lot, and over the scanner drifted “They’re plotting something…”
No, I wasn’t arrested, they were obviously joking!
Anyway, the engines cut off from the cars and bobbed around the yard a bit.
The plant, built in 1931, was the largest shipper in the entire division until the smelter was upgraded. In the 1950s, it was putting out an incredible 13,000 carloads a year! (The smelter was only putting out about 7,000, although taking in 18,000 loads!)
I somehow expected something…bigger.
Here’s a photo of the plant without that bothersome train in the way. It was surprisingly quiet and odor-free!
Anyway, you can expect some minor updates on the Taghum area after the deck is done. I’ve decided to do the water before I even put the bridge in, because it’ll be much easier!
For the record, my keyboard is full of PC board and wooden tie ends.
I’ve discovered that N scale stock rails at the point where they’re filed do not hold ties very well, and adding additional pliobond glue just mucks up the points. For the ties between the ‘S’ or ‘H’ tie (depending on your point type) and the last PCB tie I recommend putting CA or other superglue sparingly on the stock rails to avoid gumming up the turnouts or losing ties.
Another few weeks of turnouts has come and gone, and I’ve finished 26 of them. TWENTY SIX. That’s a lot. It’s almost to the point where my thousand dollar investment in fast tracks tools and templates has paid itself off.
Here are two shots of all the Castlegar turnouts arranged in close to their final positions:
When I took this photo, I’d yet to put ties on any of them.
Method for ties, since paying an extra $7 per turnout for quicksticks (lovely as they are) is a real turnoff. Print off the template from fast tracks, put double-sided tape on it, position crossties so the end you need to cut is pointing away from the diverging route, and cut them all in a straight line with rail nippers. Finished product:
Then apply the glue to the rails, put them down on the ties, weight them down, wait for glue to dry, and hopefully remove the template for reuse without incident.
So, the lessons? Here we go…
Filing points using fast tracks Pointform tools: Make sure not to put pressure on the part of the file that’s contacting the block, but rather the rail. Don’t use too much pressure. Try to file on as much of the block as parallel to it as possible. I’ve nearly worn out my #6 pointform just from bad filing practice! It’s got a rut in it like a male elk in fall. I cranked out 32 #5 points in the past 2 days with a little bit of wear on the #5 pointform, as compared to the 44 #6s I did, and it looks a lot better.
I’ve been doing my multiple turnout spacing using the point-to-point spacing from xtrackCAD as well as the frog-to-frog spacing. THIS IS BAD. I kept having to redo stuff because in at least the #6 N xtrackCAD template, there is a full half inch more between the frog and the points in real life! So that screwed me up, measuring frog to frog was WAY more accurate.
Also, as you can see in the second picture there, I’ve actually put all of my procab stuff in place and am very much enjoying running trains on the 1900′ of track I have. You know, running a loco or two back and forth and dreaming about what I’ll put there…
Goals for the month: Finish ALL the turnouts for the pulp mill (all #5s) by Monday the 13th (All the ties by the end of the month) and make major inroads into building my bridges, most of the parts of which have arrived. Unfortunately, Walthers sent a few wrong things aand the saw at the local hardware store is FUBAR, so I can’t finish my fascia or lay track for a while. Still, even in this small bit of the layout, there’s a lot to work on!