Last week I picked up a “Squealer” from Iowa Scaled Engineering at the Colorado RPM meet. I’ve been eyeing one of these for a while, and I’m so glad I finally took the plunge! If you’re not familiar with it, the Squealer is a static system installed at a curve on the track that detects the motion of a passing train and emanates recordings of high-pitched flange squeal sounds. While this may strike some like the torture of nails-on-a-chalkboard, I find the sound to be very prototypical and not annoying at all. In fact, now that I can hear my train going over a Squealer, when it’s not going over it, things sound a little empty which tells me this is a GREAT product!
Installation was very simple and took me about an hour from start to finish, to include testing. The basic Squealer consists of a sensor (about the diameter of a pencil and around 2″ long), a speaker with a ~1″ cubed souknd enclosure, and a circuit board with terminals for connecting power and sensors, a spot for a micro SD card (in case you want to load your own sounds… I find the built-in sounds to be great), and two buttons for volume control. It supports multiple sensors, so I also picked-up an extra “TrainSpotter – Motion Detector” for an adjacent track. The trickiest thing about my installation is it was into finished track with no way to drill a hole from underneath as the instructions recommend. Instead, I had to drill downward through the ties. The instructions recommend a 3/8″ drill bit–I used this on my first sensor install and found it to be overly large. For the second sensor install, I used an 11/32″ bit and found it to be just right. I wrapped my sensors in a couple layers of electrical tape to help them fit snugly and to mitigate moisture from ballasting getting to the circuit board. Per the instructions, I lined the sensor up in the correct orientation and set it about even with the ties.
I hooked up the sensors per the instructions. Since the white and blue wires each go to a common terminal on the board for all sensors, I found soldering the two white wires together and blues together made it much easier to install them (and keep them) in their respective slots. ISE recommends cutting a 3/8″ hole into the scenery for the speaker sound to escape. It was easy to cut a hole into my red-rosin-paper scenery and install the speaker to the adjacent sub-roadbed using a piece of 3M double-sided foam tape. I covered the speaker hole with a little piece of green clump foliage. For power, the Squealer can take a wide range of DC, so I hooked it up to my 14V DC accessory bus which works just fine. That was it for installation, and it worked as advertised right off the bat. The only adjustment I made were dropping the volume to a level that sounded right in comparison to my locomotives. I was able to carefully add and glue some ballast over the top plate of the sensor to hide all but the silver portion on top, and I noticed no decrease in performance.
My only complaints (a very strong word for it) are very minor regarding the sensors: 1) they’re pretty big, so if you can’t drill the hole from underneath, you’re tearing up some ties, and 2) the sensors are very reflective and create a bright glint when viewed from certain angles. It’s a pretty complicated sensor for its size, so I don’t anticipate this can be easily fixed at a reasonable price point, and the benefit is far greater than these drawbacks.
If you haven’t already, you should check out the video to hear what it sounds like with a train. To me, it now makes it sound like all my cars have sound decoders… at least in that one spot. That means I’m now plotting to put a series of Squealers all around the layout on all major curves (probably 3-5 per deck) so you can enjoy the extra ambiance no matter where you’re working. It’s not cheap, but now that I’ve experienced, I am a huge fan, and it would sound flat to go back, just like I could never go back to locomotives without sound. Install these at your own risk… if you install one, you’re likely to want several!
After years of hauling around black foam core inserts pretending to be coal loads, I finally got around to finishing some of them… 58 to be precise! I won’t go over the whole process here (article coming soon on Appalachian Railroad Modeling), but I’ll give you the basics here. Most people are used to seeing gently sloping coal loads that can barely be seen over the top of the car–this is a load that’s been on the road for a while and settled into the car. I model coal at the source, and this looks very different. In the ’60s and ’70s, much of the coal was loaded by feeding a car under a chute a few feet at a time using gravity or a winch to pull the car along in stages. This resulted in a series of high, distinct, and often uneven coal lumps, perhaps a dozen or more.
I haven’t seen too many modelers attempt this look, so years ago (like 25 years ago), I came up with a way to model this look using foam core and real coal. 25 years later, the biggest improvement has been the introduction of black foam core which does a MUCH better job of hiding any imperfections. I use 1/2″ black foam core, though some of the older forms were made from two 1/4″ pieces laminated with white glue. I basically cut them about 1/16″ smaller than the dimensions of the hopper, press them into the car to know where to cut notches for any bracing, and carve the load. I start by cutting a rough 45-degree angle around all sides, then I cut notches where I want the lumps to be. I try not to be too precise, and all the lumps are slightly different sizes. I then start rounding the lumps and eventually cut the top poster board layer off the foam core. The final shaping is done by compressing some of the foam with my fingers to smooth it out. If it needs it, I’ll add little pieces on the ends underneath and cut them to fit the car so the load sits a little higher.
Before I lose track, I label each load for the type of car it fits (easy with a silver Sharpie), and I add a couple of stick-on weights under each load (pinewood derby weights). I want my coal cars to “feel” heavier to a locomotive when they’re loaded, and I also want them to be a little top heavy like the prototype so crews will handle them a little differently.
I grind my coal the old fashioned way: with the butt of a butter knife in an old Cool Whip container. Then I sift it with a kitchen strainer to get a container of small coal. I paint the top of the foam core with straight white glue, then sprinkle the coal over the top and shake off any excess. I do this over the coal container so I can recycle any coal that falls off.
I set the loads on parchment paper to dry overnight. The next day, I spray the tops of the loads with “wet glue” (about 8:1 water:glue + a couple drops of dishwashing soap) until they’re saturated. This really sets the loads and keeps the coal from leaving dust on hands and models. Once the glue has completely dried, I go along each load with my fingers, knocking off any protruding pieces of coal, then I clean up any coal on the edges with an X-Acto blade.
The white glue tends to dull the coal’s sheen, so my last step is to hit the loads with a coat of clear satin-finish lacquer spray–this looks about right to my eye to bring the coal back to its original luster. All told, when you do them in bulk, it only takes a few minutes and a few cents worth of materials per load, and I love the way they look! I also like that each load is absolutely unique–something tougher to achieve with commercial loads.
As I’ve discussed previously, I like to build mock-ups out of foam core and paper of the larger structures on the layout that I will eventually scratchbuild. This serves three purposes. First, it gives me an opportunity to create a line-drawing / blueprint and make sure the drawing works before cutting more expensive materials. Second, it allows me to visualize a scene and make adjustments before I build the permanent structure. And third, it gives me a good stand-in on the layout until I can build the real one–something that makes operations a lot more fun than imagining there’s a big structure where you’re switching! For this third reason, I put a little extra effort into the drawings to give them some color and texture. I’ve covered the techniques before, so I won’t repeat them here.
This project involved the two “truck dump” tipples that were built in the late ’70s (as far as I can tell) near St Charles, VA. One is known as JAD Turner, and it’s probably still standing today. The other sat on the wye at St Charles for just a few years–I don’t know it’s name, so I’m just calling it the St Charles loader (super original, I know…). What made this project challenging is I didn’t have any good photos of the loader configuration I needed to model. JAD Turner was modified over the years with a second conveyor and second empty track, but the earliest photo I have of it (a grainy aerial from 1981) clearly shows only a single conveyor. I’m modeling it as if it’s the same tipple but with fewer added parts, so I took the dump shed / crusher section of the current loader along with a single conveyor and came up with this design that looks reasonably close to the aerial.
St Charles was a little more challenging as the ONLY photo I have is a grainy aerial from 1981 showing what looks like a pile of coal, a conveyor (maybe two), and what looks like a dump ramp but no shed. I didn’t have to look far to find something close. Just up the road between St Charles and Mayflower was a loader known as “Southwest” which had a similar dump and conveyor arrangement. Southwest was built after my era, so I won’t have to have two similar looking loaders on the layout. Who knows, perhaps they moved the loader from St Charles up to Southwest? That’s my story until someone proves otherwise…
Anyway, here are the results, and I’m really liking the scene now. I can’t wait to build the real things! But first the upper deck…
While I finished the basic scenery forms on the lower deck a few weeks ago, I decided it would be better to ballast the tracks before adding the upper deck benchwork while I still had good access to them. I’m so glad I did because ballasting gives the scenes a much more “finished” look even though there’s still a ton of scenery work to be done. I’ll be honest, I was dreading ballasting the track–I had little experience with ballast, but from that experience I saw it as a frustrating, tedious, and time-consuming job. I have now changed my tune! While it’s still time-consuming, I was able to learn and mature my techniques quickly to avoid the frustration and tedium, so I’ll pass along my method here.
First, I had to determine what kind of ballast I needed. This wasn’t as straightforward as I’d hoped. As best I can tell, most of the tracks in my area were at one time ballasted in cinders harvested from steam locomotives. The steam locos went away in the 1950s, and with them the ability to get cheap and ready cinders for ballast. Photos from the ’80s and ’90s clearly indicate most everything got covered in rock ballast–would the cinders still be around in the 1960s and ’70s? After some digging online, I found that cinders in many places lasted for decades after steam, and in the coal fields, it’s tough to tell cinders from spilled coal anyway, so an added incentive for cinders, at least on sidings and secondary tracks. For the main tracks, photos show the Southern’s ballast in this area was a medium gray. I toyed around with trying to find some actual rock to use as ballast, but in the end I decided on good old Woodland Scenics products made from crushed walnut shells because I can find it readily, it doesn’t cost an arm and leg to ship, and it’s pretty easy to work with. I used fine cinders and medium sized gray ballast in the big shaker containers for this project, and I was able to do the entire lower deck (12′ x 16′) with just under two shakers of each color (4 shakers total).
Second, I did a bunch of research on how to apply ballast, and I am so glad I did! In the end, I went mostly with the method Cody Grivno of Model Railroader lays out in the article here. The only other materials I needed were white glue (I bought a gallon), isopropyl alcohol (I used about XX oz), and dish soap. For tools, I used a spoon, a large flat brush, a white glue dispenser (like the ones kids use in school), two small jars with eyedroppers, a work glove, and my fingers. In the glue bottle, I mixed up some “just a bit wet glue” which is about 2 parts white glue, 1 part water, and a drop or two of dish soap–when you squirt it out, it should dissipate from its bead but not run. In one of the jars with an eyedropper, I made a mix of “very wet glue” of about 1 part glue, 6 parts water, and a drop or two of dish soap–it should look about the consistency of milk and absorb into the wet ballast (I’ll explain that in a minute) after a few seconds. You’ll need a LOT of the very wet glue, so you can either make a big batch or mix it on-demand when you run out (what I did… it was a lot of trips). The remaining jar and eyedropper are for the isopropyl alcohol.
I worked on the track in about 2 foot sections, usually one track at a time. If you’re doing two ballast colors, determine which ballast should be “lower” and work on that one first–for me, the cinders were replaced by ballast rock, so where they overlap, I did the cinders first. With about half-a-spoonful of ballast, I first apply it to the “gauge” (inside the rails) of the track. It takes a few tries to get a feel for how thick to lay it, but it becomes routine pretty quick. You want just enough that when you spread it the ballast fills the space in between the ties and rails with little on top of the ties and nothing on the rails. I found my finger to be an effective spreading tool, and I just rub it back and forth down the tracks, rubbing any excess ballast to open areas. Cody glues his ballast at this point, but I found it easier to lay the edge ballast first. I applied ballast to the edges by first running a bead of glue from the bottle down the side of the subroadbed and on top of the scenery–this helps the “slope” to hold better. Next I used the spoon to apply ballast inward toward the rail from about the edge of the ties until I couldn’t see the edge of the subroadbed any more. I used my finger again, first to poke the ballast under the rail a bit, then to wipe off the tops of the tie edges, and then to pat down the sloped edges until they looked smooth. I used a brush to clear off any unwanted ballast to outside the range of the glue (I vacuum it up later) and to remove any stubborn ballast from areas my finger couldn’t get to.
Once I’m happy with the ballast shape, I glue it down. The critical part of this process is to USE THE ALCOHOL AS A WETTING AGENT FIRST! If you just add the glue, the ballast will float (and float away) which makes a frustrating mess. You can avoid this by first saturating the ballast with an eyedropper of isopropyl alcohol–just drop until everything looks wet. I follow the alcohol wetting with the very wet glue, making sure I apply drops to every section of ballast until things were saturated and it took a couple seconds for the glue to soak in. If you drop a big glob of white glue that somehow didn’t get diluted, no worries–just dilute it with some alcohol, and it will likely settle in just fine. I wet and glued each section by starting with the gauge between the rails, then moving to the edges. I found for the edges it’s better to start the alcohol low and work up to keep things in place, and its better to start the glue high and let it work down.
Turnouts require a bit more care, and I probably didn’t take all the care I should have… it worked out ok, but I spent a couple hours massaging my turnouts to get them working smoothly again. I spread the ballast a little less thickly between the ties to make sure the points wouldn’t catch, and I took care to avoid putting ballast in the area of the throw. No matter how careful I was, there was always some piece determined to get stuck in the throw, so I used the brush (and the occasional X-Acto blade) to fish out any offenders. I used the very wet glue sparingly in these areas, but there was still some glue that stuck to the top of the ties causing the points to stick a bit. I believe Cody’s method is to drop the glue in first, then add the ballast under the points, and I think I’ll try this next time.
After letting the glue dry overnight, I clean up any excess ballast. First, I use a work glove and rub it over the top of the ties and edges to knock off any obtrusive pieces. Next, I clean out the area in the web (sides) of the rail using a matchstick rubbed back-and-forth followed by a fingernail. I used a flathead screwdriver to clean out flangeways if necessary. I cleaned up any excess with a vacuum. You’ll inevitably find spots you missed with the glue, but it’s easy to just add more ballast, drop some alcohol, then drop some glue to repair.
Ballast is scenery, so I also wanted it to tell a story. Because track repairs would have been made with gravel instead of cinders in my era, I picked a few spots along the cinder-ballasted yard tracks to fill with gravel (in this case, Woodland Scenics fine gray ballast) to simulate a replaced tie. I like the look! I also picked a few spots in prominent areas to lay some cinders on the scenery to go underneath the ballast rocks to show that some tracks were once cinders but had now been ballasted with rock. I also laid a thin layer of cinders in areas where I know tracks used to be, even though I don’t model them in my era. Finally, I added extra cinders to areas under tipple chutes and where locomotives sit to represent spilled coal and grime. I’m pretty happy with how these “extras” turned out, but they won’t fully tell the story until more scenery is complete.
Ok, the ballasting was the last step before adding the upper deck, but you’ve heard that before… We shall see.
Many of the grade crossings on the St Charles Branch didn’t have flashing signals and were protected by flagmen or fusees (see how I simulate fusees here), but a few of the more prominent crossings were protected by automatic flashing signals and bells. One of those crossings is Bonny Blue Road which crossed two legs of the wye in St Charles. I was looking for a way to make these signals work automatically with nothing required of the crews (beyond sounding the horn for the crossing) and no hardware needed on any rolling stock. I quickly settled on using IR sensors mounted near the tracks to trigger the circuits required for the crossing. While many of the major manufacturers of railroad electronics offer circuits for flashers and for triggering based on sensors, there seemed to be a lack of good, simple options for the sensors. So I did what many of us do when we’re looking for something–I turned to eBay.
I found a lot of products for flashing crossing signals, but one it particular caught my eye. A company called “WeHonest” was offering what looked to be decent looking LED signals that came with a flashing circuit for a very reasonable price. Being a little suspect of a foreign company calling itself “WeHonest,” I needed four signals, so I ordered a couple sets and hoped for the best. I ordered the signals with two heads instead of four (front and back) because my signals would only been seen from one direction, and the ones with four heads looked too thick front-to-back (I plan to add dummy heads on the back later). When they arrived a couple weeks later, I was impressed with the quality for the price. The lettering is easily readable, the construction is mostly metal, and the size and shape are good for HO scale. I had to clean up some areas of the metal crossbucks, and some of the silver paint flaked off, but these were easy fixes. I initially hooked up the flashing circuit to a pushbutton on the fascia, and the flashing circuit worked flawlessly and controlled all four signals in a synchronized manner.
The signal piece was solved, so now I needed a way to automatically control them. My confidence in “WeHonest” was bolstered, so I explored their options. They offer a “model train detector automatic signal controller crossing system trigger etc” (also called a “master board”) which shows a diagram of how it can be configured to trigger a grade crossing flashing circuit using simple, single-unit IR sensors that don’t require a broken path. I also needed a circuit that could support four sensors due to the tracks that would trigger this grade crossing, and while the board only supports two sensors, their diagrams show that you can connect more sensors via separately available splitter cables. They also offer a sound effect circuit with multiple grade crossing signal bells (and a rooster). I ordered a master board, sound effect board, two splitters, and some additional IR sensors.
The documentation you see on eBay is all you get, so it took some studying and tinkering to set things up, but it wasn’t difficult. The basic idea is the master board is connected to 12V DC and the IR sensors. The sound effect board and flasher circuit are daisy chained off the 12V DC “output” side of the master board which is only live when the IR sensors are triggered. The only surprise on wiring was there are no normal contact screws for the 12V DC input, only a plug for an adapter and a specific connector type (both of which are sold separately). I found a plug off an old RC helicopter I disassembled years ago that did the trick. I mounted all three circuits and the speaker on a piece of masonite to keep the wiring tight and organized. Rather than use the supplied speaker, I attached a pair of baffled cube speakers I had pulled out of a locomotive when I replaced it with a Scale Sound System speaker.
I installed the IR sensors between the rails and ties as the company indicates in the pictures. When anything passes over it within a couple inches, the IR sensor is triggered. There is no documentation on how the sensor works, but it has two elements, a blue dome and a black dome. I can only speculate that it transmits IR from one dome and receives reflected IR in the other dome. When I hooked everything up, it worked great… with two IR sensors plugged into the two separate sensor inputs on the master board. When I tried to use all four IR sensors, it would trigger the circuit no matter what I did even if nothing was present. I noticed some sensors were more sensitive than others, so I experimented with different placements and combos and even the positioning of the elements within the sensor. Unfortunately, I destroyed one of my sensors in the process, but thankfully they’re inexpensive, and I found the WeHonest customer service to be very responsive and helpful!
When my replacement sensors arrived, they did the exact same thing as before. Two sensors worked fine, four sensors triggered the circuit even with nothing present. I really liked the overall operation of these circuits, so I kept experimenting to see what might work. I speculated that the circuit detects based on a threshold of received IR energy–with one sensor, the ambient IR was low enough to stay below the threshold, but with two sensors, the ambient IR increased above the threshold to make it appear a train was present. I found that if I covered a portion of the black domed element on some of the IR sensors, it would keep the circuit from triggering but would still trigger if a train passed. After playing around, I found covering about 60% of the black element of all IR sensors with a small piece of electrical tape made everything work as intended.
Now that I’ve worked out the kinks, I’m very happy with the crossing! I’m able to control the sensor sensitivity via the electrical tape, I can control the flash rate of the LEDs via a dial on the flasher circuit, I can select the bell sound from one of several good options on the sound effect circuit, and all of this works automatically with no actions needed from the crew. I have two more flashing grade crossings to go on the upper level, and I’m satisfied enough that I’ve already ordered the parts to replicate this installation on those crossings.