Choosing an Era

I'm often asked why I model the 1930's. After all, I'm only 40, I'm in high-tech, and I'm a bit lazy about reliability issues. Why wouldn't I be modeling the 1970's, or modern day? I could run nice, reliable diesel locomotives instead of balky steam locomotives. I could spend hours computerizing the layout so I had automatic signals and dispatcher control instead of avoiding the computers and electronics problems I normally like.

It's a complex question to answer.

First, modern-day railroading just wasn't that interesting for me. One obvious reason is that modern trains are big. The locomotives are big, the freight cars are big, and the industries that have rail service are big. Fitting those kinds of scenes into the space I have - either a shelf layout or a single garage layout - just isn't easy. Older trains are smaller, so I can model more.

Even if I had the space for long trains and big engines, I'm not a railfan. Spending a day going out to the tracks and watching trains run by isn't that interesting. Even worse, I grew up on the San Francisco peninsula in the 1970's. I grew up watching dirty, unmaintained Southern Pacific diesels going by, and watching all the industries disappear. The steel mills and warehouses in South San Francisco either got torn down or stopped getting rail service. The canneries and food processing plants of the South Bay closed. San Francisco's Bayshore yard turned into a ghost town as all the warehouses in San Francisco turned into office space, dance clubs, and outlet stores. There's just no romance for me in that era, so trying to model the SP in the 1970's just doesn't sound fun.

My strongest reason, though, is to learn more about and recreate my grandparent's world. I've seen photos from the 1920's and 1930's, and I like the reminders of California's agricultural past, and of a time when railroads were the new and modern thing. I can see the pictures of my grandmother and her brothers and sisters in Santa Cruz for the day; they've just driven down from Hayward, they're dressed up, and they look like they're having a wonderful time. What was the drive on Highway 17 like for them? What sightseeing did they do?

My mom's family was from Hayward. My great-grandfather came over from the Azore Islands in the 1870's, earned money as a shepherd for the Miller and Lux, and bought ten acres above Hayward in the 1890s which he planted in apricots. He lost two wives, then met my great-grandmother and raised four children with her. The family dried the apricots on the property; every year for the harvest, my great-grandmother would prepare food for the family and the workers they hired to pick, cut, and dry the apricot. This photo of my great-grandfather shows him proudly looking over a drying tray of apricots sometime in the 1930's. Those apricots would get sold to one of the dried fruit packers down at the railroad tracks; the packing houses in Hayward probably looked a lot like the ones I model in San Jose and Campbell. My grandmother even worked as a bookkeeper for one of the packers for a while. My great-grandparents' fruit ranch probably looked like every farm in the Santa Clara Valley. When I'm researching and building my models, I'm also learning more about the world my family came from.


I also see the pictures of what the Bay Area was like. In this case, my grandmother took a photo of one of her friends sitting on top of a "gravity car" on the Mount Tamalpais Railway. My father took me and my sister hiking on Mt. Tam pretty often when we were growing up; I hadn't realized then that my grandparents had gone hiking on the mountain too, or that they'd stopped and taken photos at places like the Double Bowknot where I'd walked by 50 years later. My grandmother would have ridden the Northwestern Pacific to visit relatives in Sonoma County, or taken commute trains on the Peninsula when they lived in Redwood City. A couple decades later, my dad remembers working in the Hunts and Del Monte canneries in the summer, and knows that he got paid well for unloading cans from boxcars and checking the fruit cutting machines.

It's all memories. It's learning about family and the world they lived in. It's learning about how the Bay Area's changed from agricultural to high-tech, and from empty spaces to wall-to-wall suburbia. It's knowing what kind of orchards occupied the space where our house is, and knowing the importance of Sunnyvale's historic apricot orchard. It's about understanding the San Francisco Bay Area during one of the other booms the area's gone through, from the wheat fields to orchards to industry to electronics to the Internet.

How much difference could a few years make?

As I've been building my layout, I've also been looking at historical documents from the 1930's to figure out what to build. What stations were busy during the 1930's? Where were signals installed along the line? How many passenger trains ran through Los Gatos? I've looked at old employee timetables for some of the facts, picked details out of books for others, and made wild guesses for other details. I'm not being too picky about sources; when I found a 1939 timetable excerpt hanging on the wall of the Santa Clara depot, I used its schedule to plan what trains I'd run during operating sessions.

A 1939 timetable? That's only seven years later than the 1932-or-so era for my model railroad. How inaccurate could it be?



Well, I got a chance to find out. Bruce Morden visited my layout a few weeks ago, and asked the simple question: "Is the Vasona Branch really what that piece of track was called?" I told him I didn't know, and I'd been hoping to see a 1932 timetable so I could learn for sure. A couple weeks later, Bruce sent me copies of pages from a 1928 and 1932 Southern Pacific Coast Division employee timetable. I've been poring over them ever since in hopes of learning something about the small bit of the world I model - the branch line starting in San Jose, continuing through Campbell, Los Gatos, Wrights, and Glenwood.

I've found that even in the short time between 1928 and 1939, a bunch of little details changed on the layout. The way the track was controlled changed, signals were added, branches and spurs removed, and trains added and removed.

First, start with the question of the name of the layout. I chose the modern name for the branch of track - the "Vasona branch". The line's last stop - Vasona Junction - gave the branch its most recent name. But in 1932, it was the "College Park - Santa Cruz branch" because the line broke off the main line at the College Park station (near Bellarmine High School in San Jose's Rose Garden neighborhood). The original main line turned east there, crossed over the top of downtown San Jose, then headed to Los Angeles by heading south on 4th Street. The Santa Cruz branch instead headed south past the modern-day Diridon Station, then turned southwest towards Los Gatos and Santa Cruz. In 1939, the main line followed the old branch from College Park to San Jose's Diridon station, and the former "West San Jose" station became the "San Jose" station. The 1939 timetable shows the branch now being called the "San Jose - Santa Cruz" branch.

Second, there's yard limits. Railroads are very strict about when a train is considered to have "authority" to run on a stretch of track. Generally, a train running on a main line must have approval from the central dispatcher to run on that track, and must be able to get out of the way of any trains with higher priority. In some places on the railroad, either the amount of switching to be done is too great, or permission from a central authority too troublesome. For example, in a railroad yard, the crew would go nuts if they had to send a telegraph message each time they blocked the main track. For this reason, railroads declare some tracks to be in "yard limits." Permission to run on these tracks only requires permission from a local yard master, and all trains (except passenger trains) following those tracks must assume that another train might be on the main track, and go appropriately slowly. (All trains must still get out of the way of "first class" passenger trains, and passenger trains can travel at a higher speed speed on main tracks in yard limits.)

Parts of the San Jose - Santa Cruz branch were considered in yard limits, and were owned by the yardmaster in San Jose. This may have been because the tracks were infrequently used, or because the switching was heavy. Strangely, the range of yard limits varied over time. In 1928, it only stretched from San Jose past the canneries to the Western Pacific crossing a mile down the tracks. In 1932, yard limits stretched from San Jose all the way to Los Gatos. In 1939, yard limits was pulled back to only go from San Jose to Campbell.

The choice of yard limits matters because it affects game play on the layout. With yard limits, trains can run without explicit warnings, and can be told to just "stay out of each other's way." If the trains had to rely on a dispatcher, freight train crews would need to ask permission each time they wanted to leave a town, and couldn't just "avoid" other trains they encountered. I've typically been running everything from San Jose to Vasona Junction as "yard limits", and require trains only ask for permission if they go further on. Now that I know how much yard limits vary, I'll need to think whether the Vasona Junction approach still seems reasonable, and whether extending or shrinking yard limits might be fun.

Third, there's the question of signals. Signals add animation to the model railroad and give me a chance to play a bit more with electronics than I might otherwise do. I'd learned from pictures that the whole railroad line from Los Gatos to Santa Cruz had signals, so I added signals on the upper deck a few years ago. What about on the lower deck? Did signals ever go beyond Vasona Junction?

According to the timetables, signals only existed from Vasona Junction to Santa Cruz in 1928. There were no signals along the Los Altos branch between Mayfield (California Street in Palo Alto) and Vasona Junction. Some time between 1928 and 1932, signals got added to the West San Jose to Vasona Junction branch. This work would have occurred as the Southern Pacific was rebuilding the San Jose main line to run around the west end of San Jose. I'm surprised that the signals were installed in Campbell but not on the Los Altos branch; the Los Altos branch got more of the fast trains and was the connection to San Francisco, while the San Jose to Los Gatos branch was a slower track with lots of industries and fewer passenger trains.

So do I add signals all the way to San Jose? I'm leaning against it. Signals take a lot of work to install and build; they're fragile, and there's usually so many freights working on the lower level that the signals would constantly be red. I'm willing to reconsider this at some point, but for now I think I'm going to keep the signals off the San Jose - Los Gatos line.

There were a few other interesting facts I grabbed from the timetable. In the 1928 timetable, the Campbell station entry has the notation "TO-R" next to the name. This indicates Campbell was a train order station, and a train register book existed. A train order station has an operator who can relay messages between the dispatcher and passing trains. Trains on the timetable usually don't need train orders; the timetable gives those trains the authority to run along a particular piece of track. Train order stations are much more important for extra trains not on the timetable. Extra trains need to get telegraphed orders from the dispatcher explaining where they can run, what trains have priority over them, and where meets will occur. The fact that the station agent or telegrapher in Campbell had to prepare train orders suggests there may have been a lot of trains going by. There wasn't as much traffic elsewhere; on the Los Altos branch, the only train order stations were at Mayfield and at Los Gatos. Once a train got onto that line, orders couldn't be changed til it got to the other end. Because yard limits only stretched from West San Jose to the WP junction, trains likely would have needed orders at Campbell.

Not all freight trains are extras; the 1928 timetable also showed one freight train, #202, on the schedule. It ran on Tuesdays only from San Jose to Campbell and then disappeared from the timetable. Because of its disappearance, I'd guess it turned aside at Campbell and went on the Almaden branch towards New Almaden, then probably returned to San Jose by following the other part of the Almaden branch towards the main line at Lick (now near Capitol Expressway and Monterey Highway).

The train register book is our reminder that the Almaden branch existed then. A train register book is kept at a junction between two lines, or at a place where trains can start or end. When a train passes, it writes down the time it passed. If another train is supposed to wait until a specific train has passed, the crew can check the book to see if the track is clear. In our case, if our train coming from Los Gatos to San Jose is supposed to wait until train #202 passes Campbell and goes on the Almaden branch, our train can check the book and see if #202 has signed in. If not, our train waits til #202 passes. If #202 did sign the book, we can continue to San Jose knowing the track is clear.

The Almaden branch disappeared in the early 1930's. The 1932 timetable shows that Campbell is no longer a train order station, and no longer has a train register book.

A final tidbit from the timetable is that the 1928 timetable shows fewer passenger trains to Santa Cruz than the 1932 timetable. That's really surprising; by 1932, the depression had reached bottom, and the SP was closing stations and cutting trains. I'm curious why there were so few passenger trains in 1928; did the SP cut back because of competition from the new highway over the mountains, then increase trains as fewer people were able to drive? It's just another mystery to explore.

Now, that's one way to find some projects...

Now that I've started this blog, I want to make frequent updates both so it'll be interesting for readers and so it'll keep me working on models. I rushed out tonight to take pictures of one of an almost-complete building (PG&E's generator plant in San Jose), brought the photos inside... and just got annoyed. Lots of details in the photos were wrong - a telephone pole was crooked, the SP tool shed looked like it was sitting in front of the weeds (instead of actually in use, etc.

I gave up on those photos and decided the scene deserved a bit more work. Actually, it gives me some nice inspiration; I've got a building I want to photograph and a scene that doesn't look right. Why don't I just promise myself I'll spend a couple evenings getting the scene better and ready to photograph?

Stay tuned to see if I actually fix up the scene...

Prepping for an Operating Session

My layout lives in a garage. It's not a great environment for a model railroad. The garage holds a car, so the garage door gets opened and closed a lot. The garage door slides sideways, so sealing the bottom isn't possible. The layout gets dusty, and the spiders that find their way in spin webs across the tunnel portals to catch the occasional locomotive. Trying to ensure the layout operates well has been a struggle. Here's some of the work I've had to do.

When I started trying to host operations at my layout, I found this was a great way to frustrate friends. My small steam locomotives, tight curves, and garage environment meant that I had tons of stalls. Poor track caused derailments. Seth and Byron probably spent half their time that first night nudging the locomotives over dead spots.

After that session and a couple other unsuccessful ones, I got religion about making the layout operate smoothly. I'd originally used Shinohara track switches with Caboose ground throws to hold the points in place. This worked for about six months, but after a while, dirt got in the way and power didn't make it to the switches. I ended up replacing every ground throw on the layout with a Tortoise switch machine, and used the contacts on the Tortoise to make sure power got to the switch correctly. Then I found that the Tortoises would switch power too quickly and cause shorts, so I had to crawl under the layout, remove every switch machine, open it up, and grind the PC board traces down so the power wouldn't get switched til the points moved. If you didn't understand any of that because you're not a model railroader, don't worry - just understand I did a huge amount of painful work because I wanted to make sure people enjoyed operations on my layout.

I also figured out that my small brass steamers weren't behaving great either. In these locomotives, the wheels on the right hand side of the locomotive would pick up power from one rail, and the wheels on the left hand side of the tender would pick up power from the other rail. If dirt and uneven track kept a couple wheels from making contact, it wasn't hard to have the locomotive stall. I found Lenz's power modules could let a locomotive run for a couple seconds without power, so I put one in a locomotive and enjoyed the lack of stalls. (That locomotive moves beautifully and smoothly; the only problem is that it doesn't have a sound decoder so it's a bit quiet. Come and see it anyway!) Mark Gurries also pointed me at a web page which describes how to connect a capacitor to common decoders to get similar stall-proof behavior. Both helped, but the operating sessions still had stalls.

Finally, I realized I'd have to be strict on layout cleanliness. I ended up writing up a checklist of cleanup tasks before an operating session. I assume everyone with an operating layout has a similar list, but I've never seen anyone talk about their list. My list (which takes a few hours to complete) includes:

• Sweep the garage, and vacuum the sections of indoor-outdoor rug around the layout.
  
• Go on cobweb and dust patrol on areas around the layout.
  
• Vacuum all track on the layout, hidden and not hidden.
  
• Vacuum particularly dusty parts of the layout.
  
• Wipe down all track with a cloth soaked in alcohol. Keep wiping til the rag comes off clean.

  
• Clean all locomotive wheels with alcohol, then wire brush them with either a Kadee driver cleaner or with a wire brush in a dremel tool. Without the wire brush step, I haven't been able to get reliable operation.
  
• Clean all plastic car wheels by running them across a paper towel soaked in alcohol. Usually, the metal wheels are clean, but some of the plastic wheels are always amazingly dirty.
  
• Check that every track switch moves correctly when the switch on the fascia is thrown. Check that all the frogs are powered by manually touching the switch points the wrong way and listening for a stall.
  

For a final test, I'll take a balky locomotive (usually a small switch locomotive), and set it loose on the layout. If it stalls, I figure out what happened and clean the track there.


Then there's the general straightening up:

• Remove unrealistic or inappropriate buildings on the layout.
  
• Remove damaged or unrealistic trees.
  
• Move garage clutter out of the garage.
  
• Hang the drapes hiding the clutter under the layout.
  
• Check that all layout signage is in place and readable.
  

What's in your cleanup checklist for your layout?

Scenes from Wrights

Here's some photos of Wrights, both the real location and my model. The first two show the final station site on the north side of Los Gatos Creek (across the bridge over Los Gatos Creek from the tunnel under the summit. The second two pictures show the area in front of the tunnel. The real photo shows the original station site in front of the tunnel, and, across the tracks from the station, the general store, post office, and hotel. The model photo shows the same location (from a different vantage point) looking across the tracks at the general store.

The Perils of Modeling a Prototype

One of my favorite parts of my Vasona Branch layout is researching what the area was like in the 1930's, then building models based on photos, track plans, and other details about what was really there. Now, I can't actually match reality; building a model railroad in this way requires occasional compromises to figure out what will fit on the layout, what makes the layout fun to operate, and what makes for an interesting scene.

Occasionally, I get caught and learn that my idea of reality was completely wrong. The last time this happened, I found out my nice shed-style building for Borcher Brothers building materials near the old San Jose Market Street station had been torn down and replaced with a nice Spanish revival facade in the 1920's. Oops.

My latest historical crisis came from History San Jose's Cannery Life on-line exhibition. Del Monte's Plant #3 on Auzerais St. finally closed down in San Jose in 1999l, and was torn down and replaced by housing last year. The article describes the history of the plant. It was once the largest cannery in the world, and some claim that fruit cocktail was invented there.

Del Monte used "the whole hog" when it processed fruit. The choicest pieces of fruit were sliced and canned in their own cans. Smaller bits and parts of damaged fruit went into fruit cocktail. Even the pits from apricots had value, so Plant #3 also had an apricot kernel processing plant which is visible on Sanborn fire insurance maps for the time.

The Sanborn Fire Insurance maps have a small problem. These huge books were used to help insurance companies figure out how much to charge an industry looking for insurance. If a business was next to a dangerous industry or if the containing building was shown not to have fire protection, the insurance company could charge more. Some of these books still exist (San Jose Public Library has several), and University Microfilm also sells access to images from these books. However, it's hard to find a "1930" Sanborn map because most books were updated with glued-in sections as facts changed. University Microfilm's collection, for example, has 1915 and 1950 versions of San Jose, so I have to guess what was there during my era of 1930-1935.


History San Jose kindly included this image from the 1930 map showing the kernel processing part of the plant. (It's labelled here as a separate company.) Unfortunately, there's also the nice pencilled note mentioning that the building burned down in 1932. Now, I don't model the kernel processing buildings, and the corner where the buildings existed holds the end of a siding, but should I be modeling these buildings? Should I be modeling the empty space?

Ugh... a little knowledge causes a lot of pain.

A Few Lingering Projects

My new job has been cutting into my modeling time, but I managed to find some time last weekend. I ended up working on one lingering project, and one bit of detail.

First, the detail. My model railroad layout's gotten to a point where some of the scenes look complete - scenery, buildings, and track are in. Now, I've got time to add little details to make things look realistic. I've been installing working signals along the track. In the Santa Cruz Mountains, many of the signals were semaphore signals where an arm moves up and down to indicate whether a train can stop or go. Semaphores are real attention-getters, so the block boundaries where signals appear tend to get attention. Now, a real railroad signalling system is more than just the signals; there are also the telegraph poles that carry the wires between signals, and equipment cabinets, and, hidden in the weeds, battery boxes.

Battery boxes hold lead-acid batteries that power the signals when regular power goes out or isn't available. These days, there's probably a car battery stuffed in an equipment closet somewhere, but back in the old days, you'd see rectangular concrete vaults half-buried near signals. These vaults held batteries made from a glass jar and metal plates. The signal maintainers would stop by occasionally to replace the plates, refill the water, or add extra battery acid to keep everything running fine.

I thought these would make a neat project, so I made several from a couple pieces of styrene - a base, a slightly larger piece of sheet for the battery vault lid, and a strap to mimic the steel rod that would lock the vault lid on. In the picture of the signals at one end of the Glenwood siding, you can see one of the battery boxes in the dirt behind the signal. I still need to build the equipment closet that holds the relays that make the signal work, but that'll be a project for another rainy Saturday.


My second project for the weekend was building a searchlight signal. The model semaphore signals need a motor and mechanism underneath the layout, but I didn't always leave room for all the machinery. Luckily, the real Southern Pacific occasionally used color light semaphore signals in the Santa Cruz Mountains in the 1930's, so I can always model one of these signals instead.

When I last went to the hobby shop, the only signals I could find were some beautiful kits from Class 5 in Santa Rosa. These kits are a combination of plastic castings and etched brass, and make very detailed, realistic signals. Unfortunately, they don't have any way of actually lighting them -- there's no space in the scale light housing parts for an LED or light bulb. I've spent way too much time getting the signals to work under computer control and correctly change colors as trains go by, so I knew I had to light one of these signals.


After a couple false starts (and leaving the kit on my workbench for months just to annoy myself), I finally dived in with my soldering iron and a two-color LED. The pictures show the results. Rather than use the kit's light housing, I filed the LED down until it would fit through the hole in the brass target. I then soldered one of the LED's leads to the brass rod holding the signal up; the other got soldered to a thin wire that ran down the tube. The rest of the signal was assembled according to the kit's instructions. Unfortunately, the LED didn't quite sit in the same position as the kit's light housing, so the bracket for that is floating in air with nothing to support as seen in the side view.

Note the number plate on the signal indicating (in SP tradition) the number of miles and tenths of miles from San Francisco. (Here in the Santa Cruz Mountains, 3 digits were enough for the number plates; in Texas, the signals had 5 digit numbers because of the distance from the home office.) Another number plate went on the signal near Glenwood as seen in the first photo.

This signal got installed near the station at Wrights. Now, I only need to buy a few more signals for a couple other locations on the layout.