Hall Disc (aka Banjo) Signals Explained

If a book on North American signaling has a chapter before semaphores that covers stuff like, flag signals, ball signals and tilting targets, Hall "Banjo" signals are often included since they not only look strange, but have also been extinct in the wild since the 1950's.  While hardly forgotten to history, most sources tend to focus on how these signals worked mechanically, as opposed to how they were employed functionally.  The result is that unless one really digs, you are unlikely to know much more about a Hall banjo signal than it was sort of like a searchlight and sort of like a semaphore.  Today I hope to clear up some of the ambiguity surrounding these dinosaurs of North American signaling.

First, how do Banjo signals work?  Hall disc signals were one of the first types to work in conjunction with the revolutionary idea of the track circuit.  Now I have seen some sources associate Banjo signals with time separation systems, and while they could certainly be used in such a capacity, I am going to focus on their use in automatic blocks.  A disc signal (patented ~1869)  is basically a searchlight mixed with a semaphore, before the advent of either electric lighting or electric motors.  Like semaphores, disc signals have a large day indication (the disc) illuminated by natural light, and a small night indication illuminated by an oil lantern.  Like a searchlight, there is a magnetic armature that moves in response to an electric current. When the signal is energized the disc and filter are moved away from the day and night signals changing them from red (or yellow) to clear (or green).  If you can't tell where the disc goes, it is moved up to the 10 o'clock position where a small corner can still be observed.

Some say this is where the term "Clear" signal came from.  Argue in the comments.

The above photo shows a pair of banjo signals in their "clear" state, the colored disc moved up and out of the way to allow daylight to show through. The two state disc signals could be made to display three signal states be using either a distant-home configuration (as is still popular in Europe) or by mounting two heads on the same mast, one displaying Red and Clear, the other Yellow and Clear.  Clear/Clear = Clear, Clear/Yellow = Approach and Red/Yellow = Stop (and Proceed).

This rare color photo shows a Hall disc signal actually doing its job.  As you can see the colored discs are actually pretty effective.  In a time before electricity, these would be a high tech alternative to mechanically operated semaphores. This also explains why Banjo signals are so often associated with the Reading system.  In the late 19th century the Reading was an extremely wealthy railroad and would be able to afford something that would be like CBTC today.  As ABS became more prevalent, disc signals fell into the same category as 2-position lower quadrant semaphores,a commodity technology which lasted into the 21st century on some former Southern Pacific lines.

Now it is time to put together the final piece of the puzzle.  Above we see an arrangement that was only possible within a fairly narrow sliver of time.  Reading Hall disc signals are on the same mast as mechanically operated semaphores.  What is going on here?  Well the disc signals that are designed to work with low power, primitive track circuits are providing an automatic block signal indication, while the mechanical semaphores are worked from a nearby interlocking tower providing the route protection. 

 

Here we see a slightly different scenario where we have two position semaphores for trains moving torwards the camera, the top mounted semaphore for a straight route, the middle semaphore for a diverging route and then the low mounted call-on signal.  Then we see the Banjo for trains moving away from the camera as ABS exit signals. The semaphores can be directly controlled from a mechanical lever frame, displaying either Clear, Medium Clear or Restricting, and any ABS modifications given by the Baljo signals when the train departs the interlocking limits.  So Medium Approach would be Medium Clear entrance to Approach exit.

I am by no means an expert on all the ways these signals were employed, but I do know that one can't simply dismiss them as proto-searchlights or glorified semaphores as is often the case in so many books and museums.  Hall disc signals needs to be understood in the context of a manual block world devoid of even the basic electronics that would revolutionize railway signaling in the 1890's.  They were cutting edge technology employed by only by those railroads that could afford to be early adopters.

Grand Central Terminal Interlocking Charts

It is often quite difficult to find good technical information on New York City region rail infrastructure due to a combination of security theatre and clever employees knowing that information equates to job security.  While some Penn Station information exists on Mark Bej's site, most of what I have on Grand Central is in the form of modern day, paper form track schematics that don't provide much detail and don't provide any historical insight.  For this reason I was quite surprised when an unrelated Google search brought up a result containing not only a complete Grand Central interlocking chart representing the late New York Central era, but also the complete set of locking sheets detailing which levers locked what.

Apparently these diagrams have been around since 2004, so sorry if I am a little late to the party, but if I can miss them I figured they could use a bit more promotion.  The diagrams themselves are in .pdf format so you'll have to click through since I don't feel like making screenshots.  For a quick overview GCT was controlled by 4 interlocking towers, U, A, B and C, each with a GRS Model 2 pistol grip type interlocking machine.  There are another pair of listed interlockings, FA and FB, in service on the loop tracks, but these may have been sections of the A and B machines. Like Buffalo Central Station, the New York Central used a common lever numbering strategy across the entire terminal complex. Lever 1 was in U tower, Lever 1099 was in B tower. This is a good tip to remember if you see photos with lever numbers above 400 and think you've found a contender for the world's largest interlocking machine.

While the diagrams don't have any specific information about the interlocking machines as one might find on an official chart, you can derive much of what you need including frame size and the number of switch and signal levers.  The diagrams also provide one of the best references as to the layout of GCT itself, better than even some of the official documentation which  isn't drawn to scale.  Enjoy!

Throw Away Society

So NS is slowly re-signaling it's former N&W lines in Virginia and West Virginia, that's not anything new.  What is interesting was something I saw in a pair of photos.  At Farm, WV, NS is replacing a number of new NS style Darth Vader masts with...newer NS style Darth Vader masts.

Just goes to show that it doesn't matter how new the signal is, how new the wiring is or even if PTC is being installed or not.  It's simply cheaper to install and test in parallel instead of cutting a signal over to a new batch of wires.  This is a bit baffling given that new signal installations have terminal boxes at the base to make that sort of changeover cheap and easy, but I guess buying new is cheaper and easier. 

Round Searchlight in an Oval Hole

Everybody knows that part of what makes a searchlight signal a searchlight signal is the round backing.  Be it the large target variety that became the Darth Vader signal of its day or the small target variety that tickled the fancy of the Michigan Central and Conrail, the whole reason the backing plate affixed to signals is called a target is because they are round (with the searchlight at the center).

Round round, get around, I get a round.

 Well some people just have to be different and in the case of searchlights this means the interesting case of the non-round target.  Typically associated with the Chicago Northwestern Railway, oval searchlight targets could also be found on certain parts of the CSX system in what I think was the old L&N territory.  The reason the CNW is known for them is that this style of searchlight appeared at many famous railfaning locations including TOWER A-2 and JB tower in Chicago, in near classic coaling tower in Nelson, IL and at the Rochelle, IL railfan park.

Tower A-2, Clear eastbound on the CNW West Line

The oval searchlights were a US&S product, but don't think that US&S only sold this type of searchlight as they were more than happy to sell the default round backing for their H-2 and H-5 searchlight offerings.  As far as I can tell, the oval backings were a modification of the backing for N-2 style color light signals as seen here in their intended function.

So why go oval?  In the case of the CNW,  much of their signaling involved (sometimes sideways) GRS color light signals with a mid-sized oval backing and the oval searchlights would provide a similar look and feel.  On the L&N their standard signal was the US&S N-series so again, an oval backing would match.  This theory is given credence by this photo of searchlights in former C&O territory near Toledo.  The C&O was almost exclusively a user of US&S N-series color lights.

Of course in other situations the oval backing may have been used for clearance or line-of-sight reasons, like this example in the wilds of Tennessee.  I can't tell if these are simply chopped down N-2 backings or something with completely different dimensions.  I suspect they are simply trimmed down on the sides.

I'll leave you with another photo taken in 1955 that shows the CNW didn't save the oval searchlights just for interlockings.  Here we see a UP passenger train rushing past a pair of what look to be brand new CTC signals on the CNW Galena Division between West Chicago and Nelson.  All 8 searchlight heads are fitted with oval backing plates.  I am uncertain how much CNW territory got these signals and how long they lasted, but it goes to show that oval searchlights can always pop up where you don't expect them.