Page 2 of 4

Re: Swaynton Signal Box

Unread postPosted: Mon Mar 27, 2017 7:56 pm
by Chris Osment
davidwoodcock wrote:
On this website http://www.trainweb.org/railwest/railco/sdjr/signals.html, if you scroll down to "Ground Signals", Chris Osment has provided sketches which illustrate these basic types of signals without actually showing every

To give the correct credit, the 'sketches' came from JH :-) Ironically that page has been modified since your comment was posted and one sketch now replaced by an actual photo!

Somehow I've managed to miss this topic previously, so I've had some interesting catching-up to do. Moon {sic) Valley??? - sounds fun! Seriously though, wooden dolly posts intrigues me.

As regards the allocation on Push-Pull levers, I would agree that there does not appear to be a convention as to which does which. My logical (?) mind suggests that it ought to be Pull for signals coming towards you and Push for signals going away.......

Re: Swaynton Signal Box

Unread postPosted: Sun Jul 16, 2017 8:57 pm
by Brightspark
An update.
Once again thanks for all the comments and advise. Since my last post I have been busy at the drawing board and I am now in a position to start cutting metal. Posted are pictures of the MSE frame. This has been extended to 17 levers with 13 set at midpoint for the push/pull.
Photos on RMweb. http://www.rmweb.co.uk/community/index. ... fe-on-sea/

Updates will follow.

Andy

Re: Swaynton Signal Box

Unread postPosted: Sat Aug 12, 2017 10:22 pm
by Brightspark
They say that a little knowledge is a dangerous thing. Also assumption is the route of all evil. Also get an expert to check...etc, etc.
There is also the problem that modellers suffer regularly, the Sod's law that says that as soon as you finish a model someone will show you a picture that shows you got it all wrong.
This is not so bad as I may be able to rework what I have. So not finished, but at a point where I would rather not be. However I have another question to see if I have to go back to the drawing board and cut some more metal.
The question relates to locking.

I have assumed that locking only affect the points (turnouts if you want) ahead of the signal. So, for example, looking at my diagram (published on the link above) at the running signals on down line.
I currently have, when pulled off: Signal 16 (Home) locks 11, Signal 15 (Starter) locks 10 and 7 and Signal 14 (Advanced Starter) doesn't lock anything. (they will of course all release 17 as well when in combination)
The problem is that his week I was handed an extract from a book with an illustration of a junction and the locking tappets (marked fig6, if anyone can identify where it is from).
This states that the points behind the signal are also locked.
So that would mean that;
Signal 16 locks 11 and is unchanged.
Signal 15 locks 7 & 10 and also locks 11.
Signal 14 now locks 7 & 10 and possibly 9.

So my question is, bearing in mind the period and location, do I have to amend my locking chart and add the extra tappets?

I look forward to your considered opinions.

Regards

Andy

Re: Swaynton Signal Box

Unread postPosted: Sun Aug 13, 2017 7:43 am
by RDNA
The problem is that his week I was handed an extract from a book with an illustration of a junction and the locking tappets.....
This states that the points behind the signal are also locked.


What you refer to is 'each way' locking of trailing points in rear of a Stop Signal. The idea being that after a stop signal is cleared for a train to go forward the signalman should not be able to change the position of points behind the train, as this could possibly allow a second train to approach the stop signal before it had been replaced to danger.

So on your diagram -

3 reverse should lock 6,9 and 10 'either way'

14 reverse should lock 7 and 10 'either way'

15 reverse should lock 11 'either way'

Exemptions were sometimes allowed where there was an exceptionally long distance between the points in rear and stop signal ahead and having to wait for a train to pass the signal before, for example putting a crossover normal, would cause delay to parallel moves.

Full sequential locking would of course mitigate against the perceived danger but it was by no means universal, or provided at all by some companies.

Hope this helps and does not complicate your project too much!

DB

Re: Swaynton Signal Box

Unread postPosted: Sun Aug 13, 2017 8:28 am
by davidwoodcock
A quick check of an actual LSWR locking table confirms (unsurprisingly) that this was indeed LSWR practice.

Re: Swaynton Signal Box

Unread postPosted: Sun Aug 13, 2017 3:27 pm
by Chris Osment
I can't find any actual locking table to check what you have already, but - as a precaution - do not forget that signals also lock other signals behind them if they are for opposing moves. Eg 14 will lock 13PUSH and vice-versa, and 13PUSH and 4 will lock each other too.

I have also encountered cases where, for example, if 7 is reversed then you need 13PULL reversed in order to pull 14, tho' I've never quite worked out why ?

Re: Swaynton Signal Box

Unread postPosted: Tue Aug 15, 2017 8:08 pm
by Brightspark
Thank you for the replies.
Hangs head down and sighs.

Chris, I didn't publish a table. If I had done then...and I also did not consider the conflict of the 14-13Push & 13Pull-4. Good call.

The requirement for 13Pull to lock 14 when 7 is reversed is interesting. Is it because of the distance or should this also be applied to 3 and signals 4&5?

The comments by David and DB also make me wonder who did not adopt each way locking and did it become standard practice?

Anyway I shall compile a locking chart and post in the other forum.

Thanks

Andy

Re: Swaynton Signal Box

Unread postPosted: Tue Aug 15, 2017 8:26 pm
by StevieG
RDNA wrote:
The problem is that his week I was handed an extract from a book with an illustration of a junction and the locking tappets.....
This states that the points behind the signal are also locked.

What you refer to is 'each way' locking of trailing points in rear of a Stop Signal. The idea being that after a stop signal is cleared for a train to go forward the signalman should not be able to change the position of points behind the train, as this could possibly allow a second train to approach the stop signal before it had been replaced to danger.

So on your diagram -

3 reverse should lock 6,9 and 10 'either way'

14 reverse should lock 7 and 10 'either way'

15 reverse should lock 11 'either way'

Exemptions were sometimes allowed where there was an exceptionally long distance between the points in rear and stop signal ahead and having to wait for a train to pass the signal before, for example putting a crossover normal, would cause delay to parallel moves.

Full sequential locking would of course mitigate against the perceived danger but it was by no means universal, or provided at all by some companies.

Hope this helps and does not complicate your project too much!

DB
To possibly supplement DB's excellent explanation Brightspark, I believe, from reading a rather long time ago, some of the Institution of Railway Signal Engineers' famous old 'green books', that this form of locking (not specific to only the LSWR, by the way) was, I think, referred to somewhere within them under a heading of "Holding The Route", for the reasons outlined by DB.

Plus, I also considered that this locking had an additional potential benefit when there may not have been a second train to be signalled over those points in their opposite position to follow the first, of preventing, or at least reducing, the risk of the signalman accidentally prematurely moving the points to their opposite position before the ('first') train passed clear of them.

AFAIK, this 'each way' locking became a widely applied principle, though possibly not universally, until means, other than the signalman's eyes, were developed and installed [typically track circuit(s) ] to positively detect when movements passed clear of such points.
Doubtless though, there were boxes where this was done but the time, trouble and expense of altering the mechanical locking to remove the 'each way'-ing was not seen as justifying being made a pressing requirement for carrying out immediately, and perhaps some such locations were in fact never done.

Re: Swaynton Signal Box

Unread postPosted: Tue Aug 15, 2017 9:23 pm
by Brightspark
Cheers StevieG,

The Locking table is now posted. I managed that remarkably quickly which means that it is probably full of errors.

Andy

Re: Swaynton Signal Box

Unread postPosted: Wed Aug 16, 2017 10:14 am
by Andrew Waugh
I'm a bit hesitant to comment, as 1) I'm not an expert in LSWR interlocking, only what I've read in the UK textbooks and seen in Australia, and 2) too many commentators might make things confusing. But, I've looked at your locking table and it seems considerably more complex than I'd expect of a roadside station that was probably interlocked around 1890 and never significantly altered. It also misses locks that I'd expect.

My take follows. I assume that the ground discs only read over the adjacent points reverse (as was the old practice). Considerable conditional locking would be necessary if the ground discs applied over the points in both directions, which was exactly why discs only read over the points in one direction.

Note that I've only included direct locking. Direct locking is important, because it is the only locking that actually physically exists on the frame. Often levers will be indirectly locked (or released) by other levers - see lever 4 particularly. Lever 4 only has only one piece of direct locking, but reversing lever 4 indirectly locks up the whole frame except for levers 3 & 11.

1 Released by 2, 3
2 Locks 6, 10, & 11 normal. Releases 1. (Note 1)
3 Locks 11 normal, 6, 9, & 10 both ways, and locks 2 when (10 reverse and 9 normal). Releases 1. (Note 2)
4 Released by 9 (Note 3)
5 Released by 6
6 Locks 2 & 10. Releases 5
7 Locks 10 & 15. Releases 13 from mid position (Note 4)
8 (Spare)
9 Released by 10 and locks 14. Releases 4
10 Locks 2, 6, 7, 15 & 16. Releases 9 (Note 6)
11 Locks 2, 3, & 16
12 (Spare)
13 Released from mid position by 7 reverse, and 13 push locks 14 normal
14 Locks 9 & 13 push normal, and 7 & 10 bothways. Locks 3 when 10 reverse & 9 normal. Releases 17 (Note 5 & 7)
15. Locks 7 and 10 normal & 11 bothways. Releases 17
16 Locks 10 and 11 normal. Releases 17 (See Note 4)
17 Released by 14, 15, & 16.

Notes:
1. I'd expect 2 to lock 11 - essentially equivalent to the modern overlap. Signal engineers would not expect that a running move will necessarily stop precisely at the next signal. Particularly an unfitted goods. Shunting moves, however, are expected to stop at the next signal.
2. This is holding the road, ensuring that the signalman does not prematurely put back the point levers under a departing train and derail it. It is not necessary for 3 to lock 7, as 7 is locked by 10 reverse for any moves towards 3
3. No other locking is necessary. 10 is locked reverse by 9 reverse. 7 is locked normal because 10 is locked reverse, and 13 is locked in the mid position because 7 is locked normal. 10 reverse locks 6, and all the home signals except for 3.
4. I would *not* expect 7 to lock 16, as 7 is a fair distance beyond 15 (i.e. outside the overlap). But it would be quite ok to do so.
5. This would prevent a train from starting past 14, while a second train was shunting into or out of the siding over 9/10 reverse. But this would be so unlikely to be a restriction that I doubt they'd put in the conditional locking to avoid this restriction.
6. Trailing points lock each other. This point locking simplifies the locking considerably.
7. Technically, 14 locking 3 when 9 normal (& 10 reverse) is redundant, as 14 locks 9 normal. But this makes the conditional locking on 3 and 14 reciprocal and both can be implemented by one set of locks.

What have I forgotten?

Re: Swaynton Signal Box

Unread postPosted: Wed Aug 16, 2017 6:15 pm
by Chris Osment
Brightspark wrote:Cheers StevieG,

The Locking table is now posted. I managed that remarkably quickly which means that it is probably full of errors.

Andy


Might we see it here too please? It will be a lot easier than having to try to find it elsewhere......

Re: Swaynton Signal Box

Unread postPosted: Wed Aug 16, 2017 8:05 pm
by Brightspark
Might we see it here too please? It will be a lot easier than having to try to find it elsewhere......


I would love to, but I don't seem to have the means to do it.
I did try writing it out, but the message had timed out by the time I had finished.
So here is a link to where it is. It is a word file attached to the post.
http://www.rmweb.co.uk/community/index. ... _id=884752

This may even open the attachment directly.

Andy

edit. ps Oh it does open it directly. Cor Blimus!

Re: Swaynton Signal Box

Unread postPosted: Wed Aug 16, 2017 10:34 pm
by Ashley Hill
Hello,sadly I cannot open your link. Is it just me? :?

Re: Swaynton Signal Box

Unread postPosted: Thu Aug 17, 2017 12:02 am
by Andrew Waugh
Ashley Hill wrote:Hello,sadly I cannot open your link. Is it just me? :?


On my computer the link does not automatically open the Word file, it just downloads it.

Re: Swaynton Signal Box

Unread postPosted: Thu Aug 17, 2017 1:59 am
by John Hinson
There isn't a one-fits-all situation. Whether or not it opens directly depends on whether each individual's computer is configured to do so, and whether the file can be opened at all (immediately or after downloading) depends on whether the end user actually has the right version of Microsoft Word to read docx files.

This article may help:
http://ccm.net/faq/317-how-to-read-a-docx-document

John