Tom
Administrator
Signalfel?
Posts: 4,196 
|
Post by Tom on Jul 1, 2021 15:25:43 GMT
Not strictly right the train does indeed get its position from the many thousands of tags, but you failed to mention it reports back to the VCC via the radio antennas  I'm sure once I've actually been on a course for it I'll have a textbook answer, but in the meantime the only part missing from my previous post was the bit about the medium used for it, which is as you say radio to the wayside Access Point. And even don't some track circuits, of some types, and used in some places by LU, have microprocessors (AIUI Texas 9995s) in frequency tuned tracks tuner units ? Not in the Tuning Units but they're certainly in the receivers. I understand LU currently holds the world spares stock of said processor. |
|
|
|
Post by d7666 on Jul 1, 2021 15:29:49 GMT
The primary location system for CBTC is the VOBC on the train (computers) reporting their position to the VCC (computers), using the RFID tags on the track for position information. The back up uses axle counters which again feature a microprocessor based system (and yes, if you want to interrogate it, you need a computer). One of the weaknesses with an axle counter system is that you can leave a detection head off the rails and it will still show the section as unoccupied (at least until the first train). Not strictly right the train does indeed get its position from the many thousands of tags, but you failed to mention it reports back to the VCC via the radio antennas And thence through the MRU?s WRUs? (or whatever the TLAs are for those black boxes) etc etc, and they be computers too. |
|
|
|
Post by d7666 on Jul 1, 2021 15:32:10 GMT
Not strictly right the train does indeed get its position from the many thousands of tags, but you failed to mention it reports back to the VCC via the radio antennas I'm sure once I've actually been on a course for it I'll have a textbook answer, but in the meantime the only part missing from my previous post was the bit about the medium used for it, which is as you say radio to the wayside Access Point. And even don't some track circuits, of some types, and used in some places by LU, have microprocessors (AIUI Texas 9995s) in frequency tuned tracks tuner units ? Not in the Tuning Units but they're certainly in the receivers. I understand LU currently holds the world spares stock of said processor. OK, thanks, I knew they were used somewhere, and that is also my understanding on the world stock holding (it is how I found out 9995s were so used; I wanted a small number for another purpose and found another part of the same empire had a monopoly). |
|
class411
Operations: Normal
Posts: 2,743
|
Post by class411 on Jul 1, 2021 15:54:55 GMT
Thanks to everyone who answered my questions.
Having been shown to be 100% wrong about not using computers for train protection (although possibly not in the way it might appear) I spent a very pleasant couple of hours working out a possible systems architecture (what I have been doing professionally for years) that would provide a system as near safe as it's possible to be.
The epiphany I had is that as every single node on the entire system can be a computer (or ASIIC [Application Specific Integrated Circuit]) it is possible to have every communication channel - radio, fibre or copper - be absolutely certain that every signal it gets is from where it is supposed to come from - something you cannot guarantee with older connection techniques. Then if any asset believes that if does not have contact with sufficient other properly functioning assets to guarantee system integrity, at any point or in any area it can halt whatever is necessary to ensure the system's safety.
|
|
|
|
Post by d7666 on Jul 1, 2021 16:10:22 GMT
Thanks to everyone who answered my questions. Having been shown to be 100% wrong about not using computers for train protection (although possibly not in the way it might appear) I spent a very pleasant couple of hours working out a possible systems architecture (what I have been doing professionally for years) that would provide a system as near safe as it's possible to be. The epiphany I had is that as every single node on the entire system can be a computer (or ASIIC [Application Specific Integrated Circuit]) it is possible to have every communication channel - radio, fibre or copper - be absolutely certain that every signal it gets is from where it is supposed to come from - something you cannot guarantee with older connection techniques. Then if any asset believes that if does not have contact with sufficient other properly functioning assets to guarantee system integrity, at any point or in any area it can halt whatever is necessary to ensure the system's safety. Such is done on CBTC by the certifcate authority server, or whatever it's called, it has a TLA which I forget. Simple terms, possibly over simplified, each train borne VOBC talking by wi-fi is assigned a unique certficate to communicate with VCC central equipment (but that sentence skips through a heap of kit in between). I have no idea if or where ASICs might come into this, didn't go that deep into h\w when I did this stuff 5 years ago. Possibky they do. |
|
|
|
Post by d7666 on Jul 1, 2021 16:15:43 GMT
What I meant is you could remove the rail it is mounted on and leave it in the cess, and the system would still function, as has happened. REALLY  !?!?!? ohh f ff fff flips sake. |
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Jul 1, 2021 19:25:00 GMT
REALLY !?!?!? ohh f ff fff flips sake. Of course, one should never assume it hasn't been done deliberately; I know of at least one instance where it was. (I also know of others where it wasn't...) |
|
|
|
Post by t697 on Jul 1, 2021 19:31:41 GMT
What I meant is you could remove the rail it is mounted on and leave it in the cess, and the system would still function, as has happened. Not a possibility unique to axle counters either. I seem to recall an incident at King's Cross main line some years ago where the track circuit had at least 3 ends through pointwork and a rail was left out in error but the track circuit transmitter and receiver ends were still present and the missing rail went undetected until an alert train driver noticed it and stopped in time to prevent a derailment. |
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Jul 1, 2021 19:40:15 GMT
Yes indeed - an inherent weakness in 'single rail' track circuits. Broken rail detection has often been a fringe benefit which cannot be relied upon - I've been to two broken rails where the track circuit didn't detect them.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 1, 2021 20:23:47 GMT
Yes indeed - an inherent weakness in 'single rail' track circuits. Broken rail detection has often been a fringe benefit which cannot be relied upon - I've been to two broken rails where the track circuit didn't detect them. It’s very common |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 1, 2021 20:25:52 GMT
WRU - Wayside Radio Unit
MRU - Mobile Radio Unit
Both are the same kit just programmed differently depending on its use when they are unprogrammed they are known as SDR - Software Defined Radio
|
|
|
|
Post by d7666 on Jul 1, 2021 21:44:46 GMT
WRU - Wayside Radio Unit MRU - Mobile Radio Unit Both are the same kit just programmed differently depending on its use when they are unprogrammed they are known as SDR - Software Defined Radio Wot he said. I did all this stuff 5 years ago on pilot courses then politics and reorganisations cut in and got shunted away from it; so I done the theory and no practice at it. I couldn't remember what the W of WRU stood for! |
|
class411
Operations: Normal
Posts: 2,743
|
Post by class411 on Jul 2, 2021 7:52:03 GMT
What is the range of ranges over which a train communicates with a (thanks to DistrictSOM) WRU?
|
|
|
|
Post by spsmiler on Jul 2, 2021 11:32:21 GMT
Not a possibility unique to axle counters either. I seem to recall an incident at King's Cross main line some years ago where the track circuit had at least 3 ends through pointwork and a rail was left out in error but the track circuit transmitter and receiver ends were still present and the missing rail went undetected until an alert train driver noticed it and stopped in time to prevent a derailment. wow, just wow ... oh the possibilities, even at low speed |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 2, 2021 15:54:51 GMT
What is the range of ranges over which a train communicates with a (thanks to DistrictSOM) WRU? Thats a very close guarded secret between Thales and someone at LU |
|
class411
Operations: Normal
Posts: 2,743
|
Post by class411 on Jul 2, 2021 16:21:15 GMT
What is the range of ranges over which a train communicates with a (thanks to DistrictSOM) WRU? Thats a very close guarded secret between Thales and someone at LU I suppose the distance between the wayside RFID's is similarly guarded. At least from those who are not LU employees with tape measures! |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 2, 2021 17:00:31 GMT
Thats a very close guarded secret between Thales and someone at LU I suppose the distance between the wayside RFID's is similarly guarded. At least from those who are not LU employees with tape measures! There is no real set distance as such just at least 4 within 100m but there will be more around stations and points |
|
|
|
Post by yerkes on Jul 3, 2021 21:43:04 GMT
Well, my original question took off in a way that I wasn't quite expecting...
Anyway, thanks to those who responded to my question with some useful gen. I think I have a copy of the May 63 'Modern Railways' so I will look it up.
|
|
roythebus
Pleased to say the restoration of BEA coach MLL738 is as complete as it can be, now restoring MLL721
Posts: 1,275
|
Post by roythebus on Jul 5, 2021 9:02:01 GMT
Are any of these computer controlled systems any cheaper to run than a man in the signal box with levers, or the programme machine with relays? Are they cheaper to install? Are they more reliable or safer? ISTR when the Vic was opened reading that it required more ATOs than guards for one-man operation. The ARO's pay was higher than that of a guard.
Axle counters in open sections are susceptible to "things on lien line" such as the local yobbos pressing them up and down to give false readings (several cases in Lincolnshire on the main line). Can computers deal with things a driver might have to contend with such as someone who chucks a load of barbed wire on the track that gets tangled round shoe gear? Remove dead animals stuck between juice rails and running rails causing a dead short on the 750volt side? Real life situations that I had to deal with as a BR driver.
|
|
|
|
Post by spsmiler on Jul 5, 2021 9:32:09 GMT
Are any of these computer controlled systems any cheaper to run than a man in the signal box with levers, or the programme machine with relays? Are they cheaper to install? Are they more reliable or safer? ISTR when the Vic was opened reading that it required more ATOs than guards for one-man operation. The ARO's pay was higher than that of a guard. Axle counters in open sections are susceptible to "things on lien line" such as the local yobbos pressing them up and down to give false readings (several cases in Lincolnshire on the main line). Can computers deal with things a driver might have to contend with such as someone who chucks a load of barbed wire on the track that gets tangled round shoe gear? Remove dead animals stuck between juice rails and running rails causing a dead short on the 750volt side? Real life situations that I had to deal with as a BR driver. Absolutely issues that primarily apply to open air sections of railway. Things like fallen trees, supermarket trolleys and even bed mattresses could be added to that list. Then there are human trespassers. Perhaps full unattended automation should be reserved for tunnels and (perhaps) fully new-build services. Such as in cities like Lille France and Vancouver Canada. Maybe when they get their new trains the underground sections of the Central, Piccadilly and Bakerloo lines could be converted to full automation, but with staff at the front of the trains on surface sections? Of course it will be expensive - for instance: every station served by driverless trains would need platform doors (or half-height gates). This is something that has been done on the Paris Metro, so the concept is already known. Individual line characteristics and sheer practicality suggests that perhaps Southgate station would be excluded from automation? Of course, if our beloved decision makers are *that* keen for a London Underground line which includes surface sections of railway to be 'driverless' then I think more people would agree to it if it was a brand new Underground line where the surface sections could be designed from the outset to be suitable for fully automated trains. Although not on-topic for this thread I feel sure that there are paper crayonista plans for a brand new service that links Brent Cross with west London. |
|
|
|
Post by selbiehouse on Jul 19, 2021 11:00:43 GMT
Going back to the original point there is a photo on the LT Museum website 1998/83908 which shows an R stock test train at South Ealing on 19/12/1962. The caption reads as follows:
Experimental R-stock District line automatic train seen in the vicinity of South Ealing. The train was designed to pick up safety signalling codes and automatic driver commands from the railway tracks via quills located at the front of the train bogie. The experimental trains were operated on a one mile stretch of track between South Ealing and Acton Town in late 1962. A train incorporating the automatic equipmentent entered service on 8 April 1963. A section of District line eastbound track, between Stamford Brook and Ravenscourt Park, had been specially adapted for the automatic rolling stock.
|
|
metman
Global Moderator
5056 05/12/1961-23/04/2012 RIP
Posts: 7,421
|
Post by metman on Jul 19, 2021 21:18:09 GMT
One wonders what modifications to the R stock controls were made in order for the test to have been carried out. I think the converted 1960 tube stock had the twin start buttons like the 1967 tube stock but could be wrong.
|
|
|
|
Post by bigvern on Jul 19, 2021 23:22:34 GMT
the 1960 stock were fitted with the trial Victoria Line equipment pick up coils, safety Box, auto driver box, frequency generator and mechanical governor, and yes they did have start buttons and door controllers in the cab.
The side cab doors were panelled in and droplight windows fitted.
|
|
|
|
Post by t697 on Jul 20, 2021 8:02:59 GMT
One wonders what modifications to the R stock controls were made in order for the test to have been carried out. I think the converted 1960 tube stock had the twin start buttons like the 1967 tube stock but could be wrong. The article in May 1963 Modern Railways isn't clear exactly what additional controls were provided and where located. There is mention of "the master button ... pressed to re-start the train", implying just a single Start button perhaps. There are 3 photos. One shows a general view of the train and cab front on the South Ealing Test Track. One can see the pick-up coils on the leading bogie, similar to but not the same as 1967TS ones. There is also a huge maybe 5ft, vertical dipole antenna mounted on the offside cab corner. Looks like it would foul the adjacent car if it was a middle car. Another photo shows the pick-up coils closer up. Seems to be an earlier shot as the coils are cleaner as if just fitted and the dipole antenna does not appear even though its lower tip should do, judged from the first photo. Lastly there is a photo of a panel of electrical connections, indicator lamps etc. It's not clear where this is on the train. 'Cabooses' were fashionable on test trains in those days so possibly it's a closed off section between the cab and first passenger doorway. The principles of operation described seem similar to original Vic line ATC. The article mentions the Stamford Brook to Ravenscourt Park trial as still being 'shortly'. |
|
|
|
Post by jimbo on Jul 23, 2021 6:39:35 GMT
There is a page of details in Piers Connor's short history 'The R Stock Story'. Talk of ATO dates back to 1955. The most distinctive change to the test train was the large cab-mounted aerial, which apparently was not for ATO but for radio communication with the test engineer's hut at South Ealing during early trials. Tests started there on 3 Dec 1962, with use in public service at Stamford Brook from 8 April 1963 for 12 months. It was withdrawn for modification and trial on the South Ealing test track from time to time, in the light of experience. Interest then moved on to the 1960TS shuttle trials on the Central line in preparation for the Victoria line build.
The R stock trial automatic start sequence is detailed. The driver closed a special automatic control master switch. To start the train he pulled and then pushed a special double-action starting button, intended to prevent inadvertent operation. The button showed a white, blue or yellow light to indicate the code being received from the track, either full speed, 20mph, or stop.
I wonder why the trials settled on the R stock from Stamford Brook to Ravenscourt Park only. I guess a modern surface stock train had more space for a trial installation. Presumably it was near to the Signals Dept offices at Acton, and included eastbound workings from both Ealing and Richmond in an automatic signalling area, but the train would make few passes in a day. Did an engineer have to go out to observe each trip? Were all District drivers trained in the special operation of this one train? Any old-timers out there who can recall those days?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 23, 2021 11:37:45 GMT
Going back to the original point there is a photo on the LT Museum website 1998/83908 which shows an R stock test train at South Ealing on 19/12/1962. The caption reads as follows:
Experimental R-stock District line automatic train seen in the vicinity of South Ealing. The train was designed to pick up safety signalling codes and automatic driver commands from the railway tracks via quills located at the front of the train bogie. The experimental trains were operated on a one mile stretch of track between South Ealing and Acton Town in late 1962. A train incorporating the automatic equipmentent entered service on 8 April 1963. A section of District line eastbound track, between Stamford Brook and Ravenscourt Park, had been specially adapted for the automatic rolling stock.
|
|
metman
Global Moderator
5056 05/12/1961-23/04/2012 RIP
Posts: 7,421
|
Post by metman on Jul 23, 2021 12:55:58 GMT
Thanks yes I have Pier’s book and that aerial certainly was sizeable!
|
|
DWS
every second count's
Posts: 2,487
|
Post by DWS on Jul 23, 2021 13:13:45 GMT
There is a page of details in Piers Connor's short history 'The R Stock Story'. Talk of ATO dates back to 1955. The most distinctive change to the test train was the large cab-mounted aerial, which apparently was not for ATO but for radio communication with the test engineer's hut at South Ealing during early trials. Tests started there on 3 Dec 1962, with use in public service at Stamford Brook from 8 April 1963 for 12 months. It was withdrawn for modification and trial on the South Ealing test track from time to time, in the light of experience. Interest then moved on to the 1960TS shuttle trials on the Central line in preparation for the Victoria line build. The R stock trial automatic start sequence is detailed. The driver closed a special automatic control master switch. To start the train he pulled and then pushed a special double-action starting button, intended to prevent inadvertent operation. The button showed a white, blue or yellow light to indicate the code being received from the track, either full speed, 20mph, or stop. I wonder why the trials settled on the R stock from Stamford Brook to Ravenscourt Park only. I guess a modern surface stock train had more space for a trial installation. Presumably it was near to the Signals Dept offices at Acton, and included eastbound workings from both Ealing and Richmond in an automatic signalling area, but the train would make few passes in a day. Did an engineer have to go out to observe each trip? Were all District drivers trained in the special operation of this one train? Any old-timers out there who can recall those days? No not all drivers and guards would have been trained, it was only an experiment. Test train crews had to be able to man the trains (no female test train crews in 1962 ) The test track between Acton Town and Northfields was on the Westbound in them days. Of course the project was to enable the Victoria Line to operate ATO and not have guards. |
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Jul 23, 2021 15:45:03 GMT
The driver closed a special automatic control master switch. To start the train he pulled and then pushed a special double-action starting button, intended to prevent inadvertent operation. The button showed a white, blue or yellow light to indicate the code being received from the track, either full speed, 20mph, or stop. Sounds very like the old 'Westalite' style push buttons you find in some signal cabins. I wonder if any photos survive? |
|
|
|
Post by spsmiler on Jul 23, 2021 20:01:32 GMT
 Close-up of the front-left of a 1967 Tube Stock train, showing what I believe to be the receiver unit which detected the codes (from the running rail) used by he ATO system. The rail nearest the camera is the live power rail, it is rusty because this train was on a little-used depôt track. I no longer remember exactly where / when this was filmed ... 1983 Acton Depot open day, perhaps. |
|
!?!?!?
