Page 1 of 2
| How does interlocking know its own arrangement? 03/03/2014 at 21:27 #56486 | |
maxand
1637 posts |
Steamer wrote here: Quote: Thanks for the explanation, I've always meant to ask how the interlockings actually work.I agree with Steamer that this deserves its own topic. See link above for earlier discussions regarding TORR, axle counters, first and last wheel replacements that led to his post. Log in to reply |
| How does interlocking know its own arrangement? 03/03/2014 at 21:37 #56488 | |
headshot119
4869 posts |
In E10K interlocking. Turning the entrance button, and then presses the exit button for that route begins the route call process. The interlocking knows which route to call by the combination of the entry and exit buttons. The interlocking checks no other routes are set from that signal (UR relays down). The interlocking checks all points are in the correct position, or are free to move. (WZR/LR mesh) The interlocking then checks for any conflicting routes (UR relays from other signals). If all of those are true, then the route will call, and attempt to energise one of the relays in the signal head (HR, GR [Couple of other types]) these then go through another process to check for things like approach control etc. It will also display the route lights, move and lock points, and lock out any conflicting routes. That's the basics at least, other forum members are much more versed in E10K interlocking. "Passengers for New Lane, should be seated in the rear coach of the train " - Opinions are my own and not those of my employer Last edited: 03/03/2014 at 21:42 by headshot119 Log in to reply The following user said thank you: Steamer |
| How does interlocking know its own arrangement? 03/03/2014 at 21:38 #56489 | |
Steamer
3923 posts |
It might be worth adding that I was referring specifically to Geographical interlocking, since Geoff's description stated that relays are arranged similarly to the track layout. The question may not apply to other types of interlocking.
"Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
| How does interlocking know its own arrangement? 03/03/2014 at 21:45 #56490 | |
|
Steamer
3923 posts |
" said:How does this work? Is there a big set of AND gates (or relay equivalent) for each possible route, or is there a neater way of doing it? "Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
| How does interlocking know its own arrangement? 03/03/2014 at 21:49 #56491 | |
|
headshot119
4869 posts |
Turning the entry button, and then pressing the exit button, energises the UR for that particular route, the entry button is connected electrically to the UR for that route via the set of relays mentioned above, and the exit button. (Or that's a simple way of explaining it in a way I can get to make sense)
"Passengers for New Lane, should be seated in the rear coach of the train " - Opinions are my own and not those of my employer Log in to reply The following user said thank you: Steamer |
| How does interlocking know its own arrangement? 03/03/2014 at 21:53 #56492 | |
|
Forest Pines
525 posts |
" said:" said:In relay circuitry, the equivalent of an AND gate is just two relay contacts in series anyway - or, in this case, two switch contacts.How does this work? Is there a big set of AND gates (or relay equivalent) for each possible route, or is there a neater way of doing it? Log in to reply The following user said thank you: Steamer |
| How does interlocking know its own arrangement? 03/03/2014 at 22:03 #56493 | |
|
Steamer
3923 posts |
" said:Can't believe I didn't spot that! Trust me to look for complications where there aren't any... Is there a book/pdf about how interlockings work available for the general public? I apologise if I'm starting to sound like an irritating kid with all these questions- we've hit a major intersection between my railway and electronics interests. "Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
| How does interlocking know its own arrangement? 03/03/2014 at 22:42 #56495 | |
GeoffM
6284 posts |
The IRSE Green Books are old but the information is still relevant. I can't actually find them on www.irse.org but they must be there somewhere. There's a price for members and a price for non-members. Even better, but harder to find, is a book called Railway Signalling, edited by OS Nock, ISBN 0-7136-2724-7. That's also an IRSE textbook. There's a section on geographical interlockings too 
SimSig Boss Last edited: 03/03/2014 at 22:42 by GeoffM Log in to reply The following user said thank you: Steamer |
| How does interlocking know its own arrangement? 03/03/2014 at 22:49 #56496 | |
|
Ron_J
329 posts |
I'll try to make this as easy to follow as I can. For a Scottish Region geographical interlocking with an NX panel interface there are four levels involved in route setting. These are Initiation, Completion, Locking and Proving. There are five main types of geographical relay sets - main signal, shunt signal, double points (crossover), single points (turnout) and ground frame. When setting a simple route from main signal to main signal the following takes place in the interlocking... 1. Initiation: When the signaller presses an entrance button, the PNR in the entrance signal relay set picks, causing the RPNR to pick and stick over its own front contact. This sends an "initiation voltage" feed to all available routes from that signal. This voltage proves any points in the route are available for selection (via the front contacts of the WMR relay for trailing points and NZLPR or RZLPR relays for facing points) and picks the UKER in the relay set of each available exit signal. The UKER being picked prevents that signal being used as an entrance when the button is pressed. The button indication for the entrance signal will flash with the UR dropped and the RNPR picked (i.e. the button has been pressed but no exit selected). 2. Completion: At this stage the signaller chooses the exit signal by pressing the exit button. This causes the UNR to pick and sends "completion voltage" back to the entry signal set via the points NWUR or RWUR relays - if points are required to move for the route, this is what moves them. When the completion voltage arrives back at the entry signal set, the UR latches up which cuts the feed to that set's UCR. When the signaller releases the exit button the PNR drops, causing the UKER to drop in all the other possible exit signal relay sets. The button indication for the entrance signal will become steady with the UR and the RNPR both picked. 3. Locking: This level causes the points in the route to be locked and locks out any opposing routes. Each track circuit in the route has a relay known as an SR for each direction of travel - a track circuit on a bidirectional line will have Up and Down SRs. With the UR picked (see step 2 above) the ULR will drop, which will in turn drop the first SR and so on throughout the route. The sequence is held at each set of points until the appropriate LR has also dropped (the LR energised proves the points are free to be moved). At facing points the NWR or RWR is used to select the next SR to drop. At this point the route lights will illuminate on the panel. This stage is also where approach locking is applied in that the ULR contact that drops the SRs is in series with the front contact of the APR, thus ensuring the route is approach locked once the aspect has cleared. 4. Proving: This level proves all is in order to clear the entrance signal; this is similar to the completion level in that a voltage is fed back over the route from exit set UNR (already picked in level 2) to the entrance set via various relay sets in the route for intermediate points, track circuits, signals etc. When the feed reaches the entrance unit it energises the HR, thus sending a feed to the lineside equipment to clear the entrance signal. Pulling the entrance button up causes the RNPR to drop, causing the UNR and the HR to drop in turn. This puts the signal back to danger and holds the route through the APR being down. I'll admit I did have to look at my course notes for a bit of this! Last edited: 04/03/2014 at 00:02 by Ron_J Log in to reply The following users said thank you: Steamer, DaveBarraza, Hawk777, Forest Pines |
| How does interlocking know its own arrangement? 04/03/2014 at 00:05 #56500 | |
BarryM
2158 posts |
Here is a sample of Route Relay Interlocking diagrams. For the uninitiated, they may be very confusing. Barry Barry, Sydney, New South Wales, Australia Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 03:22 #56503 | |
DaveBarraza
88 posts |
Thanks Ron_J for the well written description. Apologies for entering late. Question: In the US there are Non-Vital and Vital parts of interlocking logic. Roughly, your steps 1 and 2 would be non-vital, and steps 3 and 4 would be vital. Is this the case in the UK? There are similarities between the high-level and low-level machines, 1 and 3 both involve a cascade away from the entrance, and 2 and 4 the "return". 1 and 2 being "non-vital" allows them to take place remotely if required. Can the geographic xlkg's being discussed here be easily remote controlled? If the initiation and completion were taking place elsewhere, would the field end of things begin with step 3? Another point of interest for me is the way you describe "voltage" being sent out in the direction of available exits, checking the availability of switches in the process. I'm more used to this being done with actual non-vital relays than with straight current. Even the office systems that I use are still based on the model of the relay based selection. More generally it's astonishing to me that the very complicated layouts I see in SimSig and Quail Maps could be distilled down to just 5 elements... How are crossings of tracks in interlockings (not associated with a switch) taken into account? When a switch needs to move to provide flanking protection but it is not in the path of the train (which I would assume is the path the voltage takes) how is that dealt with? Questions questions. If anyone has NYC subway signal questions I could answer in trade!! Thanks, Dave Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 05:14 #56507 | |
|
GeoffM
6284 posts |
" said:Question: In the US there are Non-Vital and Vital parts of interlocking logic. Roughly, your steps 1 and 2 would be non-vital, and steps 3 and 4 would be vital. Is this the case in the UK?Well... see my other response as well, but I think all interlocking types use the equivalent of vital relays for the entrance/exit push buttons (as well as the rest of the interlocking, of course). However, those entrance/exit relays could be driven by non-vital relays from the panel. " said: Can the geographic xlkg's being discussed here be easily remote controlled?Yes, remote interlockings are controlled by TDM or FDM links. I think in simple terms this could be described as a constant barrage of 0s and 1s representing relay states in the office, with indications coming back from the field similarly. Wiring at the field end obviously varies at least slightly between local (to the panel) interlockings and remote interlockings but as far as geographic interlockings are concerned, it doesn't care - same plug. " said: More generally it's astonishing to me that the very complicated layouts I see in SimSig and Quail Maps could be distilled down to just 5 elements... How are crossings of tracks in interlockings (not associated with a switch) taken into account? When a switch needs to move to provide flanking protection but it is not in the path of the train (which I would assume is the path the voltage takes) how is that dealt with?Not sure what the 5 elements refers to... Westpac IV has 24 module types which include a few different signal types, single-ended points, double-ended points, diamond crossings, isolated track sections, overlap marker... Flank protection is, I think, always custom wired outside of geographical units. SimSig Boss Log in to reply The following users said thank you: DaveBarraza, Steamer |
| How does interlocking know its own arrangement? 04/03/2014 at 05:51 #56508 | |
|
maxand
1637 posts |
Steamer wrote: Quote: we've hit a major intersection between my railway and electronics interestsPerhaps the ideal middle ground is to study how model railroad buffs design their switching. Admittedly there seem to be more US than UK publications and the requirements are simpler in that I've never seen any of their signalling include overlaps or complicated interlocking, but wouldn't it be a hit if you could work that into your model railway layout! I'm not a railway modeller myself, but enjoy downloading Model Railroad Hobbyist magazine (free!) each month and drooling over the lengths these guys go to to weather their rolling stock. Obviously SimSig does an excellent job of simulating real-life interlocking. All you need to know is the conditions that need to be met for each connection; headshot119 and Ron_J, above, seem to have spelt out the basics. You sound as if you already have enough digital electronics under your belt to attempt to replicate this. Maybe with an Arduino or a slice of Raspberry Pi and some over-the-counter HO turnouts you could create a really nice simulation. 
Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 11:10 #56510 | |
|
kbarber
1712 posts |
" said:Steamer wrote: Unless I'm much mistaken, neither overlaps nor flank protection (nor indeed trapping) form part of US practice. In fact, they're pretty much confined to UK (and UK-derived) systems. That's one reason why UK signalling appears to be so restrictive at first sight (and maybe answers some of your questions/frustrations Max) - it is restrictive! (In fact it's a good deal less restrictive than it used to be. Imagine, in the old days of mechanical signalling, having to ascertain you had 1/4 mile clear ahead of your first stop signal before you could give permission for the box in rear to clear his signals. Imagine, where boxes were close together, not being able to give that permission until any preceding train had passed clear of the box in advance and you'd received the 'Train out of Section' signal - always having to check for 'normal needle' ahead before giving line clear. 183 metres standard overlap? Folk these days don't know they're born! [/Monty python]) It's also one reason, I suspect, that non-UK manufacturers have struggled to succeed in the UK signalling market - their computer-based interlockings weren't designed from the outset with tricky things like trapping & flanking and overlaps in mind and have proved very difficult to adapt. SSI, on the other hand, was built from the ground up to do such things; it's been easy to sell abroad because it's easy to specify an overlap of 0 and no trap and flank requirements around a route. Log in to reply The following user said thank you: maxand |
| How does interlocking know its own arrangement? 04/03/2014 at 11:44 #56514 | |
|
sloppyjag
480 posts |
This thread reminded me of an old Open University programme about Railway Signalling which, in fairly basic terms, explains how interlocking works. (This has probably been posted here before and is currently available on Youtube - Part 1 and Part 2.) It doesn't go into any great detail about relays and was made in 1989 when SSI's and IECC were still under development - the Liverpool Street workstations are shown under test at Derby RTC. Planotransitophobic! Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 12:14 #56518 | |
|
Jan
889 posts |
" said:Flank protection is, I think, always custom wired outside of geographical units.You probably shouldn't quote me on that, but I think that at least a certain amount of flank protection is provided by default in German geographic interlockings. One reason for this might be that in the UK, crossovers almost always seem to be provided as double-ended points, which automatically provides a certain amount of flank protection by itself. In Germany on the other hand, points are mostly (always ?) configured as single-ended points, so even in a simple crossover, the other point end explicitly needs to be moved/locked in order to provide flank protection. " said: In fact, they're pretty much confined to UK (and UK-derived) systems.Don't forget Germany, although that didn't stop Siemens from messing up the Portsmouth resignalling. :whistle: The Dutch on the other hand don't have overlaps. Two million people attempt to use Birmingham's magnificent rail network every year, with just over a million of them managing to get further than Smethwick. Log in to reply The following user said thank you: Finger |
| How does interlocking know its own arrangement? 04/03/2014 at 13:34 #56521 | |
|
maxand
1637 posts |
kbarber wrote: Quote: Unless I'm much mistaken, neither overlaps nor flank protection (nor indeed trapping) form part of US practice. In fact, they're pretty much confined to UK (and UK-derived) systems.Thanks Keith for pointing that out. Explains a lot. Any Aussies here tell me if Australian systems employ them? Last edited: 04/03/2014 at 13:35 by maxand Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 15:17 #56524 | |
|
Stephen Fulcher
2026 posts |
Thanks for that information about Geographical Ron. Not something I have ever had the fortune of having anything to do with, aside from trying to make sense of the WestPac rigs at Derby Training School nearly ten years ago during a lunch break of the SSI course nearly ten years ago, but I was always intrigued as to how it worked. Perhaps I should try and teach myself a little about it, not that it will be of any use where I work now. As for E10k, I could elaborate on what Karl said above, but to do so would probably only serve to confuse most people on the forum, so I won't bother as what he said is largely correct. Max, I did contemplate building a model railway whereby the trains reacted to routes set using a purpose built E10k Western Region Relay Interlocking, along the basics of those in use at Bristol, Exeter and many other places, but the thought of having to buy nearly two hundred relays for the layout I had in mind, and then hard wire them all, put me off a little, although I did get as far as to "write" most of the interlocking diagrams. Last edited: 04/03/2014 at 15:19 by Stephen Fulcher Log in to reply |
| How does interlocking know its own arrangement? 04/03/2014 at 23:33 #56548 | |
|
mrtux
19 posts |
maxand, here in Melbourne and probably throughout Victoria, we have overlaps done in the same way as they seem to be in the UK - little or none for shunting and a suitable amount for passenger-carrying lines. I don't know what's used on the goods/freight lines. I don't know about what's used in the rest of Aus, but I wouldn't be surprised if they also use overlaps. Log in to reply The following user said thank you: maxand |
| How does interlocking know its own arrangement? 05/03/2014 at 00:54 #56550 | |
|
DaveBarraza
88 posts |
" said:Steamer wrote:Easy! I took a copy of the NYCT typical interlocking circuits and built this!   
Log in to reply The following users said thank you: maxand, Hawk777 |
| How does interlocking know its own arrangement? 05/03/2014 at 01:31 #56553 | |
|
DaveBarraza
88 posts |
Quote:NYCT uses overlaps, and flanks, and that's why SimSig works the way I'm used to! I'm inferring from the conversation here that certain interlocking platforms are specifically designed for UK signaling practice. A supplier was blamed for "screwing up signaling" apparently based on their platform... What I'm used to in the US is processor interlockers that are a completely blank slate. The executive code gives you a box that responds to inputs, drives outputs, communicates with other locations and solves logic. The entire application logic of an interlocking is the responsibility of the engineer who designs that plant, and the engineer who checks him. Do processor interlockings in the UK come with some pre-coded, pre-vetted, "canned" logic that can be applied in a modular fashion by the application engineer - like a software version of the geographic interlockings? How much is left for the designer of each interlocking? Interlocking design is highly contextual, so how much can one reasonably rely on chunks of pre-written logic? My personal experience from "automating" the design process is that the higher the level of automation the greater potential for mistakes. When I went back to pencil for first draft, I found my logic to be more carefully considered. Maybe I'm missing a fundamental difference in approach. Generally US interlocking logic designs are all derived right out of relay circuit logic. Is the UK different? Thanks, Dave Log in to reply |
| How does interlocking know its own arrangement? 05/03/2014 at 16:52 #56572 | |
|
GeoffM
6284 posts |
RE previous picture: Wow! " said: I'm inferring from the conversation here that certain interlocking platforms are specifically designed for UK signaling practice. A supplier was blamed for "screwing up signaling" apparently based on their platform...If it's the one I'm thinking of then yes, it was not entirely without issue, to put it mildly. For something like a year the line had to revert to a very basic signalling system with massively reduced capacity. To be fair, it didn't take that long to fix, but the windows of opportunity to get in and do the work were far apart. " said: What I'm used to in the US is processor interlockers that are a completely blank slate. The executive code gives you a box that responds to inputs, drives outputs, communicates with other locations and solves logic. The entire application logic of an interlocking is the responsibility of the engineer who designs that plant, and the engineer who checks him.There are interlockings that can be written this way. Westrace (Westinghouse->Invensys->Siemens) is a ladder-logic system that's used quite a bit on metro systems (including sections of London Underground) but as far as I know does not exist on the Big Railway. An "add-on" to Westrace is a templates tool - basically, you want a point end, here's a template and it'll draw the diagram and fill in the contact/relay IDs. I think you can then modify the produced data as it's a one-way process. " said: Do processor interlockings in the UK come with some pre-coded, pre-vetted, "canned" logic that can be applied in a modular fashion by the application engineer - like a software version of the geographic interlockings? How much is left for the designer of each interlocking?SSI, and thus Smartlock and Westlock which are generally compatible variations, are heavily designed for UK main line requirements. That's what it was designed for, and it does it very well (but like all processor-based systems, slower than relays). It's good enough to be used elsewhere in the world too - Belgium and Spain come to mind for starters. " said: Interlocking design is highly contextual, so how much can one reasonably rely on chunks of pre-written logic?Yes and no. The safety assessors and suchlike prefer more automation to reduce mistakes. I think it depends on the automation though, as to how likely one is to make mistakes. " said: Maybe I'm missing a fundamental difference in approach. Generally US interlocking logic designs are all derived right out of relay circuit logic. Is the UK different?This I find odd as probably 75% of US interlockings follow one of two patterns - pairs of universal crossovers, or single-to-double track! If anything could gain from automation and/or ready-made units, this would be it! Yet it appears that the designs are written over and over again. No doubt there is a lot of replication (risking copying errors), especially with stuff like Microlok boolean/ladder logic. SimSig Boss Log in to reply The following user said thank you: maxand |
| How does interlocking know its own arrangement? 05/03/2014 at 18:13 #56582 | |
kaiwhara
584 posts 
|
Wow, just Wow! I love the mimic CTC style panel, with those thumb switches. So many interlocking's over here in New Zealand had panels like that, now very few left!  Andrew Sorry guys, I am in the business of making people wait! Log in to reply |
| How does interlocking know its own arrangement? 06/03/2014 at 01:46 #56600 | |
dmaze
88 posts |
" said:Unless I'm much mistaken, neither overlaps nor flank protection (nor indeed trapping) form part of US practice.I'm used to seeing interlocking home signals pretty much right in front of the switches. It looks like there's about 200 feet in my userpic, but the signals are also placed so they're not obstructed by a bridge. But you can see Swift interlocking, the last mainline junction on the MBTA Fitchburg line, really does have the signal within about 5 feet of the first switch. Log in to reply |
| How does interlocking know its own arrangement? 09/03/2014 at 04:53 #56756 | |
|
BarryM
2158 posts |
" said:maxand, here in Melbourne and probably throughout Victoria, we have overlaps done in the same way as they seem to be in the UK - little or none for shunting and a suitable amount for passenger-carrying lines. I don't know what's used on the goods/freight lines.They are used in New South Wales. For those interested, here is a link to Transport NSW Government Signalling Principles PDF.(5 MB). Read Barry, Sydney, New South Wales, Australia Log in to reply |