Lower Qudarant signal bounce

[bordercollie]

Hi

I am wondering if I need worry about introducing signal bounce for lower quadrant semaphore signals. I have seen several opinions and have seen videos of bounce. This bounce seems to be related to, at least some extent. the maintenance of signalling. Would maintenance have been better in 1920s GWR against recent videos of surviving signals? Given the small amount of movement in 4mm scale would it be worth doing anyway?

Regards

[Phil Bullock]

Beyond any slack in the operating mechanism I don’t worry …. Happy just to have working signals! 

[kevinlms]

Change them to upper quadrant and you don't need to bother!

 

Perhaps you'd be better off asking in the GWR group.

[ikcdab]

I was a heritage railway signalman for 30 years and yes, LQ signals do bounce when returned to danger.  I think there are two factors.  Firstly the length of the wire run. A shorter run allows more bounce, and secondly the signalman.  If you just throw the lever back into the frame then expect more bounce (and post wobble).  If you are more careful and replace the lever slowly, then no bounce. 

Much model railway stuff is exagerated for effect (eg thickness of sheet metal) so a bit of exagerated bounce is fine and does add authenticity.  Very easy to add it with the right servo drivers.  Both MERG and megapoints allow it to be added. 

Ian

 

[The Stationmaster]

Don't forget that GW LQ signal had the arm linked to the balance weight by a metal rod.  so in order for the arm to bounce the balance weight - with as much as a 40 lb weight on it (although it was usually less)  - also had to bounce;  not the same as having a piece of multi-strand wire between the two.

 

If all the pivots were worn than the arm could move to a very limited extent independently of the balance weight.  But the main effect of throwing a signal lever back hard was for the post, especially the tubular metal posts, to shake a little - in model railway terms that would barely be perceptible.  Did GW arms bounce?  well if eb verything was g badly worn and sloppily maintained the answer is  'possibly' but basically, yet again hardly to the extent that would be perceptible on a 4mm, or even 7mm, scale model.  

 

I've seen probably hundreds of Western signals returned to danger in every day use - quite a few of them by me - and I've never seen any noticeable bounce - but plenty of slightly shaking/vibrating signal posts or even bracket structures.

[Pacific231G]

There used to be a lower quadrant semaphore signal on the Greenford Branch just north of where it crosses the A40. I did once video it to analyse its movement and could detect no bounce. It also took 15 frames so 15/25 seconds to move. This particular signal was a good distance from Greenford E box but was wire operated. The GE  box is still a mechanical frame (the last in Greater London?)  and there are maybe half a dozen semaphore arms close to it but unfortunately not used often enough for me to have ever seen them moving.

[DCB]

I have wasted much time trying to video semaphores working with little success.  There is no warning when they are about to move,  they don't change as the loco passes like a colour light, and shouldn't be re set until the whole train has passed,  Their default setting is Horizontal, On, Stop.   As Pacific231G says they take about 3/5th of a second to move from off to on.   I have never seen one bounce but the signal post usually shakes or vibrates as they are pulled off and put back.    Especially put back as they accelerate as the lever goes back, whereas being pulled the arm signalman sometimes struggles to get the lever all the way over in the frame.  The Dapol bounce looks all wrong, the arm should hit the stop and make the  whole structure shake  but the Dapol starts juddering  some way short of the stop. and  works too slowly to my way of thinking, though it may be the bounce makes it seem to be too long.    My attempts at working semaphores won't work slowly enough for the prototypical time but the whole thing judders as it works, it's motor operated but the armature only revolves 270 degrees

 

I believe GWR lower quadrant arms are a lot heavier than upper quadrant arms as they need a heavy cast spectacle housing to balance the arm and ensure it returned to danger if the operating rod  snapped.  The counterweight towards the base of the signal is to keep tension on the signal wire rather than to balance the arm.  

[roundhouse]

I am using the megapoints servo controller on Blackmoor to operate the lower quadrant signals. However the bounce doesnt work that well, I think due to the signal linkage being a bit stiff. One is scratch built and the others are lattice posts. Hopefully over time they will loosen up. Also I have now built my own extended programmer panel so that I dont have to go under the layout to adjust the controller board. I will see how well it works this weekend at the Nailsea show.

[jim.snowdon]

Signal bounce is down to a combination of how fast the pull on the wire is released, ie how hard the signal arm hits its end stop, and how flexible the signal post is. The faster the arm is allowed to return to its on position dictates how much kinetic energy it has when it hits its end stop. That energy has to go somewhere (see Newton's laws on the conservation of momentum, etc), and the only place for it to go is in bending the post (as the least stiff part of the whole signal structure/mounting). The post is elastic and behaves like a spring, so as it recovers it passes the energy back to the arm, which then bounces away from its end stop. Gravity then repeats the whole process. Bounce on timber posts will generally be less than on steel posts, as the timber also damps out the motion.

[bordercollie]

Hi

If I understand correctly the maintenance of the signal is a major factor as to whether bounce,occurs. As my layout is a GWR branch c1929,  could I assume, or not, that at this earlier date general maintenance would have been superior to later in time? If so would the generally better maintenance extend to a branch line situation. At this stage I am thinking that I may programme varying degrees of bounce from none to a very small amounts on the various signals. In a period of good maintenance would 3’ arms have any detectable bounce?

Regards

[The Stationmaster]

4 hours ago, bordercollie said:

Hi

If I understand correctly the maintenance of the signal is a major factor as to whether bounce,occurs. As my layout is a GWR branch c1929,  could I assume, or not, that at this earlier date general maintenance would have been superior to later in time? If so would the generally better maintenance extend to a branch line situation. At this stage I am thinking that I may programme varying degrees of bounce from none to a very small amounts on the various signals. In a period of good maintenance would 3’ arms have any detectable bounce?

Regards

There would have beena local S&T Dept Signal Lineman responsible for maintenance in particular places and he, in turn, would be responsible to a signal Dept District Inspector for the standard of his work. (the latter should ne be confused with the operating/traffic Dept's District [Signalling] Inspectors who were responsible fr ensuring the safe working of signal boxes etc.

 

Thus there was somebody nominated to maintain all signals and somebody else responsible for ensuring that he did so to the required standard.

[57xx]

On 21/03/2023 at 09:50, ikcdab said:

Much model railway stuff is exagerated for effect (eg thickness of sheet metal) so a bit of exagerated bounce is fine and does add authenticity.  Very easy to add it with the right servo drivers.  Both MERG and megapoints allow it to be added. 

 

As long as it is bounce and not waggle. I have seen videos of terrible implementations where the arm just waggles quickly and was called "bounce". Too fast and no natural decay.

[ikcdab]

2 hours ago, 57xx said:

 

As long as it is bounce and not waggle. I have seen videos of terrible implementations where the arm just waggles quickly and was called "bounce". Too fast and no natural decay.

this is one of mine...

 

 

[57xx]

1 hour ago, ikcdab said:

this is one of mine...

 

 

 

That looks great. Which system is that with?

[jim.snowdon]

2 hours ago, ikcdab said:

this is one of mine...

 

 

If anything, possibly a bit slow and exaggerated. At least you have a straight pull to the off position, not the two-stage pull that many modellers seem to think is necessary.

[ikcdab]

9 hours ago, 57xx said:

 

That looks great. Which system is that with?

That uses MERG servo4 units, programmed with their sema4 GUI.

I like it a lot. 

[ikcdab]

8 hours ago, jim.snowdon said:

If anything, possibly a bit slow and exaggerated. At least you have a straight pull to the off position, not the two-stage pull that many modellers seem to think is necessary.

Maybe.  I have been around a lot of signals and I know how much it can vary, depending on the resistance on the wire.  

With the sema4 software, it's easy enough to slow it down a bit.

I'm not a fan of the two-stage pull either, but I have seen the arm go above the 45 angle then drop back to the clear position. I haven't done that on this one.

Ian

Edited April 23, 2023 by ikcdab