Routing Question on Derby

Everywhere else, routes had to be restored behind each train (by pulling up the exit button).

kbarber in post 114800 said:

Everywhere else, routes had to be restored behind each train (by pulling up the exit button).

On every panel I've ever seen you pull out the Entrance button to cancel a route.

Phil-jmw in post 114822 said:

kbarber in post 114800 said:

Everywhere else, routes had to be restored behind each train (by pulling up the exit button).

On every panel I've ever seen you pull out the Entrance button to cancel a route.

Except on a Turn and Push panel where you would turn the entry switch :)

kbarber in post 114800 said:

Everywhere else, routes had to be restored behind each train (by pulling up the exit button).

On every panel I've ever seen you pull out the Entrance button to cancel a route.

Of course we now have the situation where some remote interlockings controlled by large powerboxes have been converted to SSI while others remain as relay interlockings; I suspect that means a mix of TORR and non-TORR in the same box and sometimes even on the same panel. Kings Cross comes to mind, where I believe Finsbury Park and Hitchin (the latter controlling as far south as Langley Junction) are now SSI. (There may also be an SSI - or its modern-day equivalent - Westrace?... Westlock? - controlling Belle Isle, since the commissioning of the Thameslonk Junction.) The remainder, though, are still relay interlockings. So the bobbies at the Cross are presumably used to having to pull up behind every train at one signal but letting TORR do it for them at the next.

kbarber in post 114825 said:

Of course we now have the situation where some remote interlockings controlled by large powerboxes have been converted to SSI while others remain as relay interlockings; I suspect that means a mix of TORR and non-TORR in the same box and sometimes even on the same panel. Kings Cross comes to mind, where I believe Finsbury Park and Hitchin (the latter controlling as far south as Langley Junction) are now SSI. (There may also be an SSI - or its modern-day equivalent - Westrace?... Westlock? - controlling Belle Isle, since the commissioning of the Thameslonk Junction.) The remainder, though, are still relay interlockings. So the bobbies at the Cross are presumably used to having to pull up behind every train at one signal but letting TORR do it for them at the next.

Birmingham New Street was similar in that the remodelled Proof House Junction area was TORR enabled while the rest of the NX panel wasn't. Signallers on the Centre Panel are very disciplined in quickly killing dead routes to keep the panel clean. A member of this group has raised a request for a no TORR option on the New Street Simulation!

The only other example I have come across with a split operation is Wakefield Kirkgate where the current control area is a mix of re control (without TORR) and re signalling (with TORR).

Implementing TORR in Solid State Interlocking is a much easier task (there may be someone knowledgeable here who can confirm whether easy enough to call it trivial).

Of course we now have the situation where some remote interlockings controlled by large powerboxes have been converted to SSI while others remain as relay interlockings; I suspect that means a mix of TORR and non-TORR in the same box and sometimes even on the same panel.
[...]
So the bobbies at the Cross are presumably used to having to pull up behind every train at one signal but letting TORR do it for them at the next.

kbarber in post 114825 said:

Implementing TORR in Solid State Interlocking is a much easier task (there may be someone knowledgeable here who can confirm whether easy enough to call it trivial).

Making TORR possible is part of the SSI core code. I haven't read the SSI manuals in detail (I think Geoff has) but I suspect that adding it to a signal is basically a matter of either adding a flag to each affected route or listing the track circuits required for TORR on the route. So a little bit of design work (and testing) but much less than adding a route in the first place.

kbarber in post 114825 said:

Of course we now have the situation where some remote interlockings controlled by large powerboxes have been converted to SSI while others remain as relay interlockings; I suspect that means a mix of TORR and non-TORR in the same box and sometimes even on the same panel.
[...]
So the bobbies at the Cross are presumably used to having to pull up behind every train at one signal but letting TORR do it for them at the next.

I would be very surprised if it was *not* possible to disable TORR on a given signal.

Actually, I have a real example: the right hand (Ely) end of Cambridge panel - the part that uses huge squares instead of the normal little tiles - is controlled by three SSIs in the relay room in Cambridge PSB. But I'm pretty sure there's no TORR on that part any more than on the rest of the panel.

clive in post 114833 said:

Actually, I have a real example: the right hand (Ely) end of Cambridge panel - the part that uses huge squares instead of the normal little tiles - is controlled by three SSIs in the relay room in Cambridge PSB. But I'm pretty sure there's no TORR on that part any more than on the rest of the panel.

I don't know if you mean SSI or something else. With SSI it's easy - just omit either of the two items I mentioned above. Quite a few permissive shunt routes do not have TORR these days.

GeoffM in post 114841 said:

clive in post 114833 said:

Actually, I have a real example: the right hand (Ely) end of Cambridge panel - the part that uses huge squares instead of the normal little tiles - is controlled by three SSIs in the relay room in Cambridge PSB. But I'm pretty sure there's no TORR on that part any more than on the rest of the panel.

I don't know if you mean SSI or something else. With SSI it's easy - just omit either of the two items I mentioned above. Quite a few permissive shunt routes do not have TORR these days.

I meant SSI - they're real SSI units. Here's a photo I took of them; they're in a corner of the relay room for

- And I forget the bottom module.

clive in post 114858 said:

GeoffM in post 114841 said:

clive in post 114833 said:

Actually, I have a real example: the right hand (Ely) end of Cambridge panel - the part that uses huge squares instead of the normal little tiles - is controlled by three SSIs in the relay room in Cambridge PSB. But I'm pretty sure there's no TORR on that part any more than on the rest of the panel.

I don't know if you mean SSI or something else. With SSI it's easy - just omit either of the two items I mentioned above. Quite a few permissive shunt routes do not have TORR these days.

I meant SSI - they're real SSI units. Here's a photo I took of them; they're in a corner of the relay room for

Maybe they wanted consistency with the rest of the panel then.

For those interested, each cubicle has the following:
- 2 panel processor modules - basically comms devices with the workstation, ARS, and anything else interested. Dual so if one fails, the other can be used seamlessly.
- 3 interlocking modules. If one disagrees with the other 2 then it gets shut down.
- And I forget the bottom module.

I'm not S&T so may be talking out of turn, but isn't there also a need for an integrator module (is that what they call it?) to actually do the checking that all 3 interlocking modules are in agreement, or to decide which of the 3 is out of order?

(There may also be an SSI - or its modern-day equivalent - Westrace?... Westlock? - controlling Belle Isle, since the commissioning of the Thameslonk Junction.)

to avoid an RBC data change, it was requested that the new westlock covering the belle isle to canal tunnels junction should not TORR for routes used by trains under ETCS, as otherwise the RBC would see the route canceled under the train and give it an emergency stop when the route TORRed as currently to save data I believe the sub route relays aren't repeated to the RBC only the route relays so it would falsely see the route locking lost under the train if TORR were put in. Though I hopefully assume now that it has been agreed to change the interface and the RBC to take the subroute latches and allow TORR, but if when Kings cross is finally remodeled those particular routes don't TORR, that would be why.)

What is the standard operating procedure for cancelling routes behind trains? Does the signaller have to wait for the train to clear the route fully before cancelling behind? Will the potential limitations of subroute relays not being repeated to the RBC impose additional restrictions on manual cancellation of these routes?

I think that rather than having an extra box to check, instead, the state of the currently operated on part of memory of the other two is fed into each of the three boxes and its up to hardware rather than software component to notice that the state this processor is in has a different value to the other two and blow a so called safety fuse.

Though I also have a feeling that there's something in the panel processor for checking the processors are in agreement though so it might be that that does it.

(I'm not even sure how subroutes are implemented in GEC geographical interlocking if they are at all.)

What is the standard operating procedure for cancelling routes behind trains?

(I'm not even sure how subroutes are implemented in GEC geographical interlocking if they are at all.)

(There may also be an SSI - or its modern-day equivalent - Westrace?... Westlock? - controlling Belle Isle, since the commissioning of the Thameslonk Junction.)

Westrace is used heavily on the London Underground (Central line and I think some other places) but I don't think on the big railway (though Nairn comes to mind)

GeoffM in post 114897 said:

Westrace is used heavily on the London Underground (Central line and I think some other places) but I don't think on the big railway (though Nairn comes to mind)

Westrace gets around on the big railway a bit more now than it used to, often in so called "modular signalling" schemes (Shrewsbury -Crewe, Lincoln SCC (outside the station area), and north wales coast, being the three that are out there so far), but also many recent Westlocks get installed with a Westrace alongside as an "Axle Counter Processor" to preform the axle counter reset/restore functionality as I believe the logic needs much less data in a Westrace than a Westlock (or at the very least there isn't enough spare capacity in a Westlock to also do axle counter data and still cover the area it was designed to). And as you say Nairn is still out there too.

What do these interlockings boil down to internally? Are they a microcontroller-type device that executes a program, or are they FPGA-type devices that are essentially a large array of logic gates?