Semaphore Queries

[tpatrick]

Semaphores came in different flavors, so to speak, but why the variations? For example some arms were pointed while others were squared; some are red with a white chevron while others are yellow and black; there were upper and lower quadrant signals, too. What is the meaning of all this? And can you point out other variations as well?

Tim 

[CShaveRR]

I'll try to tackle this one.

First of all, lower-quadrant semaphores were generally used earlier than upper-quadrant semaphores.  In both cases, a horizontal semaphore was a stop indication.  But on a lower-quadrant semaphore, the blade had to lift up to indicate Stop--not very fail-safe. 

Now, a block system of lower-quadrant semaphores, each with only two positions available, would often have two blades to give what passed for an approach indication.  On each of these signals, the top blade gave the aspect for the first block--either clear or not clear.  The lower blade moddified this by showwing what conditions were in the second block.  Thus a clear signal was given when both blades were down, an Approach was the top blade down and the bottom blade up.  And to further differentuiate between the two blades, the top one was painted red with a white stripe, and the bottom one was yellow with black.  They also had a different shape:  on an intermediate block signal, the top blade would have a pointed blade, and the lower one would have a "fishtail" blade.

The different shapes indicated a different type of signal.  In an upper-quadrant system, here's what was possible.

Flat blade:  an absolute signal at an interlocking.

Fishtail blade:  a distant signal for an interlocking, not part of a block system.

Pointed blade:  an intermediate signal in a block system.

Rounded blade:  a train-order signal, not part of any block system.

This may have varied from railroad to railroad, and also over time, but this is the general idea of what was once out there.

[MP173]

I was fortunate today to catch a NB CSX passing the blades south of Romney, In. 

ed

[tree68]

Carl - with proper counter-balancing, I'm sure the lower quadrant blades were failsafe as well.

[rrnut282]

I'm afraid not.  The bell-crank that would make your lower quadrant semaphore 'fail-safe" could jam, or the connecting rods could break , leading to a false "clear".  On the upper quadrant semaphore if anything broke the blade would fall down into the "stop" indication.

[CShaveRR]

Larry, I have to admit that I've never seen lower-quadrant semaphores in action, so I can't tell how they perform, or behave.  Having seen upper-quadrant semaphores, though, when trains "knock them down" (I've still heard it called that in color-light CTC-controlled interlockings), I can say that they literally do drop into the horizontal position--and sometimes bounce a time or two afterward.

[al-in-chgo]

Much of the semaphore tradition goes way back in history.  If at all possible, I'd recommend you get ahold of Brian Solomon's excellent book, RAILROAD SIGNALING.  It's still in print but also widely available used.  It is more than just photos and really discusses ops a lot.

Eleven Amazon amateur reviewers (I am one of them) gave the book an average of 4.5 stars, which is a very high score.

Most of you probably already know that B. Solomon has written and illustrated a number of really, really good books about railroad lines and railroad operations.  -  a.s.

 

[Anonymous]

The railroad I worked for used the lower quadrant semaphores at places where operators could hand up train orders.  I have always assumed every road used them only at train order stations, maybe not a valid assumption.  If the operator was off duty, it had to show 'proceed'; that is, the blade(s) could not be horizontal.

They were manually operated, and the cords that raised or lowered the arms were between the interlocking machinery and the outside wall next to the track. In a depot, the cords were usually just above the bay window.

When you saw a mast, between the tracks and the tower/station with two lower quadrant blades, one on each side of the mast, you knew it was a train order point.

Art  

[Deggesty]

Yes, lower quadrant signals were used as lineside signals; the only road that I remember for certain that used them was the SP.

Ice was another hazard that could cause a false Proceed signal, as it could outweigh the counterbalance.

[Railway Man]

The standard signal in the semaphore era on the Harriman Roads (UP, SP, IC) was the US&S Style B lower-quadrant semaphore.  But not all semaphores on the Harriman Roads were lower-quadrant or US&S; the LA&SL used upper-quadrant semaphores supplied by GRS and FRS, and the OSL had some GRS upper-quadrant semaphores too.

RWM

[Tulyar15]

I'm afraid not.  The bell-crank that would make your lower quadrant semaphore 'fail-safe" could jam, or the connecting rods could break , leading to a false "clear".  On the upper quadrant semaphore if anything broke the blade would fall down into the "stop" indication.

In Ireland they still use lower quadrant signals; they have weights on the spectacle plate to ensure that the arm returns to the horizontal in the event of a wire breaking. In Britain the Great Western Railway carried on using them despite a directive by the UK govt. in the 1920's requiring railways to convert to upper quadrant. The GWR simply ignored the directive, as have its successors first the Western Region of British Rail and now the Western zone of Network Rail. Indeed as recently as 2004 a new lower quadrant signal was erected at Shrewsbury! You can see it on my brothers website at: http://www.roscalen.com/signals/Shrewsbury/index.htm

[tree68]

I'm afraid not.  The bell-crank that would make your lower quadrant semaphore 'fail-safe" could jam, or the connecting rods could break , leading to a false "clear".  On the upper quadrant semaphore if anything broke the blade would fall down into the "stop" indication.

As noted by others, it's all in the balancing.  If a jammed rod could stop a lower quadrant semaphore from showing "stop", it could do the same for an upper quadrant.

It's always been my understanding that semaphores and crossing gates require a certain amount of current to keep them in the "clear" position.  If that current disappears, they move to the "stop" position, creating the failsafe mode.

The semaphores used at stations were generally called train order signals.  Some had two aspects (clear and stop), others had three, indicating no orders, orders that could be picked up on the fly, and orders that the crew had to sign for.  They were controlled by the station agent, usually with an armstrong lever. 

[wjstix]

Railroads did sometimes use different signals to differentiate train order signals from block / interlocking signals, like using upper-quadrant semaphores for block signals and lower quadrant one for train order signals at stations etc.

Lower quadrant semaphores were the standard in the UK and I suspect many other countries where the British built the first railways. In the UK upper quadrant ones existed there but were pretty rare.

The problem with lower quadrant ones that I've heard is that snow and ice could accumulate on the blade and force it down, so that it shows a false "clear" aspect. With an upper quadrant one, a similar failure would cause the signal to drop to "stop", it couldn't drop any farther than horizontal. If at all possible, it's preferred that the default or "fail safe" aspect of a signal is it's most restrictive one.

[CShaveRR]

The semaphores used at stations were generally called train order signals.  Some had two aspects (clear and stop), others had three, indicating no orders, orders that could be picked up on the fly, and orders that the crew had to sign for.  They were controlled by the station agent, usually with an armstrong lever.

Train-order signals are technically the only ones that should be called "boards".  The ones from my youth (with the round-ended blades) were controlled by levers, but not the "armstrong" levers in the terms most of us think of.  They were connected to the signals directly through the ceiling of the bay window.  They would have to be hooked to display a clear or a Form 19 indication; if left unhooked they would drop into the horizontal position.

On the CNW, the order boards I saw were all searchlight signals, but the background was triangular in shape, differentiating them from ordinary searchlight signals.

[mudchicken]

Save the T2's!

[tree68]

Train-order signals are technically the only ones that should be called "boards". 

True enough.

The train order boards at our Thendara station were (there's only up there now) were operated by two three foot high levers, complete with grip releases.  I'll have to get a picture when I go up Thursday.

 

[Deggesty]

Here's another variation on train order boards--about 1965, the IC installed flashing lights on the order boards in Brookhaven, Mississippi. I don't remember for certain, but it may have been to differentiate them from the block signals (the IC did use a flashing yellow for advance approach), but any engineer should have been able to the difference without the flashing light.

[jeffhergert]

The semaphores used at stations were generally called train order signals.  Some had two aspects (clear and stop), others had three, indicating no orders, orders that could be picked up on the fly, and orders that the crew had to sign for.  They were controlled by the station agent, usually with an armstrong lever.

Train-order signals are technically the only ones that should be called "boards".  The ones from my youth (with the round-ended blades) were controlled by levers, but not the "armstrong" levers in the terms most of us think of.  They were connected to the signals directly through the ceiling of the bay window.  They would have to be hooked to display a clear or a Form 19 indication; if left unhooked they would drop into the horizontal position.

On the CNW, the order boards I saw were all searchlight signals, but the background was triangular in shape, differentiating them from ordinary searchlight signals.

There were train order signals that had the "armstrong" levers built into the telegraph desk in the bay window.  On the Rock Island there were both the desk level levers and chains that went to levers near the ceiling in the bay window that connected to the rods.  When looking at a picture of a station with a semaphore type T O signal, if the rods going to the bell cranks are at ground level, they would have a lever arrangement.  If the rods come out at about the interior ceiling level, they would have the chain arrangement.  I've seen a picture of the interior an ex-CGW depot where they used a small lever attached to the wall with heavy wires running to the semaphore. 

The RI had both upper and lower quadrant T O Signals, as well as Flashing Light ones.  The Flashing light signals were like Carl said of most CNW's, a triangular shape to stand apart from normal block signals.  (The only CNW color light without the trangular shape heads that I've seen were inherited from former CGW or MStL lines.)  The musuem at Hampton, Iowa has a CNW lower quadrant T O signal.  I believe they were called Sanborn types, home made with scrap rail for the mast.

Jeff

        

[Tulyar15]

wjstix

Lower quadrant semaphores were the standard in the UK and I suspect many other countries where the British built the first railways. In the UK upper quadrant ones existed there but were pretty rare.

In Britain, upper quadrant signals became the standard from the the 1920's. (see my earlier post). Whilst lower quadrants do remain on former Great Western Railway lines, elsewhere upper quadrants are the standard. On former GWR lines that were transferred to other BR regions in the 1960's lower quadrants were replaced by upper quadrants. The most recent example of this happening has been at Yeovil Pen Mill station earlier this year, as a result of it having been transferred from Network Rail's Western zone to the Southern zone.

 In Ireland, lower quadrants have remained the standard, the only installation of Upper Quadrants being at Larne in Co. Antrim by the British owned Northern Counties Railway; elsewhere on its network the NC continuesd to use summersault type signals (another solution to the problem of a broken wire!). Examples can still be found at Castlerock and Portush.

[tomikawaTT]

Speaking only of the cable-operated manual-from-a-distance lower quadrant semaphores I have personally seen (in Japan half a century ago) there are multiple, redundant fail-safes.

first, the arm itself.  The blade was lightweight wood.  The spectacle casting was a substantial chunk of metal.  Left entirely to itself, the blade would always assume the horizontal position. 

rrnut282

I'm afraid not.  The bell-crank that would make your lower quadrant semaphore 'fail-safe" could jam, or the connecting rods could break , leading to a false "clear".

To address this in parts:

• The rod up the mast had to be lifted to clear the signal.  If it broke free at the top - see above.  If it broke free at the bottom, or just plain broke, the weight of the rod was on the spectacle end of the arm, not the blade end...
• The bellcrank, pivoting free, would have its rod-connection end drop down since it was heavier than the other end.
• In order to clear the signal the cable from the lever would have to raise the weighted end of the driver arm - and the weights were substantial chunks of iron.  If the cable lost tension, the driver arm came down and the signal went to its most restrictive aspect.

The semaphores I saw were used with a staff and ticket block control system - about as 19th Century as you could get.  The 'Start' and 'Home' signals (absolute stop) had red blades with square ends and vertical white stripes at the free ends.  The 'Approach' signals had swallowtail ended yellow blades with black chevrons.  They were mechanically interlocked so an approach signal could not be cleared unless the home signal to its rear had already been set to clear.

Unlike motorized U.S. semaphores, the cable-operated Japanese variety moved from horizontal to 60 degrees down (or vice-versa) faster than the eye could follow - with a clearly audible Clank! as the metal parts hit the stops that limited their travel.

Chuck (modeling Central Japan in September, 1964 - with lower-quadrant semaphores)

[tomikawaTT]

rrnut282

I'm afraid not.  The bell-crank that would make your lower quadrant semaphore 'fail-safe" could jam, or the connecting rods could break , leading to a false "clear".  On the upper quadrant semaphore if anything broke the blade would fall down into the "stop" indication.

I'm afraid I have to disagree - based on personal observation of a real lower-quadrant semaphore signaling system in Japan half a century ago.  There were multiple, redundant fail-safes:

• The blade end of the semaphore arm was a light piece of wood.  The spectacle casting was a substantial chunk of iron.  Left entirely to itself, it would assume the horizontal position.
• The operating rod that connected the sector link to the arm was attached to the spectacle end of the latter, and had to be raised to clear the signal.  Any defect would pull it down.
• The sector plate, all by itself, was unbalanced.  The rod-connector end was much heavier than the other end, so the plate would naturally assume a 'rod down' position unless something else prevented it from doing so.
• The 'something else' was a bellcrank with three arms.  One, at right angles to the other two, operated the sector plate through a roller like an old-fashioned metal skate wheel.  The shorter of the other two was connected to the operating cable, which was always under tension.  The longer straight-line arm was ballasted with a couple of sizeable chunks of iron.  The cable had to raise those weights to clear the signal.  If tension was lost, the weights dropped and the arm went to its most restrictive aspect.

The only thing that could clear a signal to the 60 degree down position was a pull on the operating cable, from the lever at the lever frame that controlled it.

The semaphores I saw were Absolute Stop (square-ended red blade with white straight stripe about 40mm in from the free end) and Approach (swallowtail yellow blade with black chevron.)  The absolute stop Home signal had to be cleared before the mechanical interlocking would allow the Approach signal to be cleared.  Once a signal was cleared, all the turnouts were locked until the signals were reset to their most restrictive aspects.  Being cable driven, they moved faster than the eye could follow, but the Clank!!! as they hit the stops was clearly audible.

Chuck