|
Post by roverlei on May 22, 2019 20:43:58 GMT
I'm writing from South Harrow, where basically every train on my commute now stops for between 1-2 mins. Train op today mentioned this was because of a new computer system which regulates the timings. Is this true? Clearly in the past South Harrow eastbound was used to park trains rolling order to push more through Rayners, but it's a bit of a pain to have to hang here every journey.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 22, 2019 21:07:45 GMT
Its not that new but yes its correct the system actually works now as before there was lots of faults with the original system which when trains are not reversing the area was running like an automatic site
|
|
|
Post by goldenarrow on May 22, 2019 21:09:19 GMT
It could well be due to control now passing from Earls Court to South Kensington that the new system of train control holds each train time by the Working Timetable at each platform stop but my knowledge of how South Ken operates is very limited. I do know that South Harrow is the boundary between Rayners Lane signal cabin and South Ken (formerly Earls Court) which is why as you said, trains are usually held at this location in both directions with the Eastbound taking into account radio alarm tests aswell. I know that under Earls Court, automatic signals (Sudbury Hill to North Ealing)were just those and only signals that were prefixed such as South Harrow (WV) or North Ealing / Ealing Com (WM) had push button panels in the control room.
|
|
|
Post by PiccNT on May 22, 2019 23:06:19 GMT
That's correct. When the branch up to South Harrow was migrated to the new PICU (Piccadilly Interim Control Upgrade) system, the vast majority of semi-automatic station starters now hold trains to their timetable departure time.
|
|
|
Post by ted672 on May 23, 2019 11:31:07 GMT
That's correct. When the branch up to South Harrow was migrated to the new PICU (Piccadilly Interim Control Upgrade) system, the vast majority of semi-automatic station starters now hold trains to their timetable departure time. I wonder if this is part of the reason for slow driving on the Picc as mentioned a few times in this thread about SPADs districtdavesforum.co.uk/thread/30624/spads?page=1&scrollTo=479773. Drivers are minimizing platform delays by taking time between stations.
|
|
londoner
thinking on '73 stock
Posts: 478
|
Post by londoner on May 23, 2019 13:18:07 GMT
Its a pretty useful "delay" for the drivers because I think there are staff toilet facilities there - at least I assume this because many drivers will leave the trains momentarily. I remember in the past the train used to always pause around the North Eailing area as well, but this does not seem to happen as much nowadays.
|
|
Colin
Advisor
My preserved fire engine!
Posts: 11,310
|
Post by Colin on May 23, 2019 15:57:27 GMT
It could well be due to control now passing from Earls Court to South Kensington that the new system of train control holds each train time by the Working Timetable at each platform stop but my knowledge of how South Ken operates is very limited. It only controls the semi automatic signals. Automatic signals do exactly that & work automatically. My understanding, in simple terms, is something like this: When Earls Court was in control it used program machines. These use a roll of film with holes punched in it similar to an automatic piano that played music based on the holes in a film that was fed through it. The program machine moves forward a step, waits for a train to appear and if it corresponds it operates the signalling and steps forward. So the process repeats. Now my understanding is that program machines can step forward up to two minutes early so in theory it allows an early running train to continue moving forwards......unless the signaller intervenes. I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. Timetables are written with leeway built into them - booked stand times here and there - to account for odd little delays that may affect a given journey. Under the old pure program machine system if there were no delays you might get to the other end of the line a couple of minutes early. This new computer system won't allow to you run 10 seconds early let alone up to 2 minutes early, so you're now seeing much more of those built in timetabled stands. migrated to the new PICU (Piccadilly Interim Control Upgrade) system, the vast majority of semi-automatic station starters now hold trains to their timetable departure time. Same effect on the District between Barons Court and Ealing.......and not just station starters either. It's created SPAD traps all over the place where none existed before. And whilst in fairness you can't predict door issues, I've had it twice this week where the starter's held me for two minutes and then when they cleared I failed to get a pilot due to the stupid S stock door system which traps small items like pens, batteries and stones in the door runners. By the time I'd dealt with it I'd departed two minutes late. Had the station starters cleared a bit early like they used to, I'd of had a fighting chance of an on time departure. Hey ho.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 23, 2019 17:53:51 GMT
Programme machines had 4 main modes
Push Button - Signal Op has full control First Come First Served - trains are routed according to TD Programme Machine Mode - trains would be routed and held to time as per roll, train number has to be correct and TD has to be correct Programme Only Mode - same as above but ignores TD but machine will route and put up the TD from the roll
The machines only step with the passage of trains, signal op can cancel the train or hold one which goes into a storage.
The timing is set up to be 30 seconds fast
The only time you will see the machines in a programme only mode is usually the start and end of service usually most machines are in FCFS mode and thus heavily relies on the TD being correct.
The only way to set a train to leave early is either in push button or FCFS mode if in the other 2 modes the train will wait until it’s timetabled booked path
Oops forgot to say there is no longer any physical mechanical programme machines left they are now a mini computer with the timetable on a USB stick but all the back end cards and relays are still in operation
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 23, 2019 17:58:20 GMT
Just realised I could of got the programme machine modes the wrong way round before a signal op pulls me on this
|
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on May 23, 2019 18:47:54 GMT
It could well be due to control now passing from Earls Court to South Kensington that the new system of train control holds each train time by the Working Timetable at each platform stop but my knowledge of how South Ken operates is very limited. It only controls the semi automatic signals. Automatic signals do exactly that & work automatically. My understanding, in simple terms, is something like this: When Earls Court was in control it used program machines. These use a roll of film with holes punched in it similar to an automatic piano that played music based on the holes in a film that was fed through it. The program machine moves forward a step, waits for a train to appear and if it corresponds it operates the signalling and steps forward. So the process repeats. Now my understanding is that program machines can step forward up to two minutes early so in theory it allows an early running train to continue moving forwards......unless the signaller intervenes. I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. Timetables are written with leeway built into them - booked stand times here and there - to account for odd little delays that may affect a given journey. Under the old pure program machine system if there were no delays you might get to the other end of the line a couple of minutes early. This new computer system won't allow to you run 10 seconds early let alone up to 2 minutes early, so you're now seeing much more of those built in timetabled stands. migrated to the new PICU (Piccadilly Interim Control Upgrade) system, the vast majority of semi-automatic station starters now hold trains to their timetable departure time. Same effect on the District between Barons Court and Ealing.......and not just station starters either. It's created SPAD traps all over the place where none existed before. And whilst in fairness you can't predict door issues, I've had it twice this week where the starter's held me for two minutes and then when they cleared I failed to get a pilot due to the stupid S stock door system which traps small items like pens, batteries and stones in the door runners. By the time I'd dealt with it I'd departed two minutes late. Had the station starters cleared a bit early like they used to, I'd of had a fighting chance of an on time departure. Hey ho. In normal mode programme machines would hold trains to to time at station starters. Only if the signal operator selected one of the degraded modes (first come first served or push button) would a train be signalled early, in the case of the former automatically by the machine, and in the latter by the signal operator when he presses the button for the signal route. At Cobourg Street most machines were left in programme mode most of the time, only deviating during disruption. I seem to remember one exception - the southbound machine at Archway which was left in FCFS, as a SPAD mitigation measure of all things, to prevent trains being held by the southbound starter which had a history of start-up againsts. (Or at least I believe that was the reason, the local instruction didn’t actually say!). It may well be the case that Earl’s Court ran/run their machines in other modes more of the time, I can’t comment if that may have been due to issues with the machines in their senior years, but it’s also likely that the District and Picc don’t tend to run quite as close to book as the Northern and Victoria did/do!
|
|
|
Post by MoreToJack on May 23, 2019 19:24:53 GMT
by the signal operator when he presses Or she, or they.
|
|
|
Post by nig on May 24, 2019 10:07:06 GMT
It could well be due to control now passing from Earls Court to South Kensington that the new system of train control holds each train time by the Working Timetable at each platform stop but my knowledge of how South Ken operates is very limited. It only controls the semi automatic signals. Automatic signals do exactly that & work automatically. My understanding, in simple terms, is something like this: When Earls Court was in control it used program machines. These use a roll of film with holes punched in it similar to an automatic piano that played music based on the holes in a film that was fed through it. The program machine moves forward a step, waits for a train to appear and if it corresponds it operates the signalling and steps forward. So the process repeats. Now my understanding is that program machines can step forward up to two minutes early so in theory it allows an early running train to continue moving forwards......unless the signaller intervenes. I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. Timetables are written with leeway built into them - booked stand times here and there - to account for odd little delays that may affect a given journey. Under the old pure program machine system if there were no delays you might get to the other end of the line a couple of minutes early. This new computer system won't allow to you run 10 seconds early let alone up to 2 minutes early, so you're now seeing much more of those built in timetabled stands. migrated to the new PICU (Piccadilly Interim Control Upgrade) system, the vast majority of semi-automatic station starters now hold trains to their timetable departure time. Same effect on the District between Barons Court and Ealing.......and not just station starters either. It's created SPAD traps all over the place where none existed before. And whilst in fairness you can't predict door issues, I've had it twice this week where the starter's held me for two minutes and then when they cleared I failed to get a pilot due to the stupid S stock door system which traps small items like pens, batteries and stones in the door runners. By the time I'd dealt with it I'd departed two minutes late. Had the station starters cleared a bit early like they used to, I'd of had a fighting chance of an on time departure. Hey ho. if that's the case can anyone explain why heathrow t 5 and t4 nearly always clear early not just by a few minutes can be as soon train gets into platform even if it doesn't depart for another 15 minutes
|
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on May 24, 2019 10:53:54 GMT
It only controls the semi automatic signals. Automatic signals do exactly that & work automatically. My understanding, in simple terms, is something like this: When Earls Court was in control it used program machines. These use a roll of film with holes punched in it similar to an automatic piano that played music based on the holes in a film that was fed through it. The program machine moves forward a step, waits for a train to appear and if it corresponds it operates the signalling and steps forward. So the process repeats. Now my understanding is that program machines can step forward up to two minutes early so in theory it allows an early running train to continue moving forwards......unless the signaller intervenes. I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. Timetables are written with leeway built into them - booked stand times here and there - to account for odd little delays that may affect a given journey. Under the old pure program machine system if there were no delays you might get to the other end of the line a couple of minutes early. This new computer system won't allow to you run 10 seconds early let alone up to 2 minutes early, so you're now seeing much more of those built in timetabled stands. Same effect on the District between Barons Court and Ealing.......and not just station starters either. It's created SPAD traps all over the place where none existed before. And whilst in fairness you can't predict door issues, I've had it twice this week where the starter's held me for two minutes and then when they cleared I failed to get a pilot due to the stupid S stock door system which traps small items like pens, batteries and stones in the door runners. By the time I'd dealt with it I'd departed two minutes late. Had the station starters cleared a bit early like they used to, I'd of had a fighting chance of an on time departure. Hey ho. if that's the case can anyone explain why heathrow t 5 and t4 nearly always clear early not just by a few minutes can be as soon train gets into platform even if it doesn't depart for another 15 minutes On a related note, is anyone able to post a summary of how the PICU system works? Its another example of every line now having their own unique signalling or control systems. I have the manual as a PDF somewhere but it’s quite a long read!
|
|
|
Post by PiccNT on May 24, 2019 12:27:55 GMT
It only controls the semi automatic signals. Automatic signals do exactly that & work automatically. My understanding, in simple terms, is something like this: When Earls Court was in control it used program machines. These use a roll of film with holes punched in it similar to an automatic piano that played music based on the holes in a film that was fed through it. The program machine moves forward a step, waits for a train to appear and if it corresponds it operates the signalling and steps forward. So the process repeats. Now my understanding is that program machines can step forward up to two minutes early so in theory it allows an early running train to continue moving forwards......unless the signaller intervenes. I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. Timetables are written with leeway built into them - booked stand times here and there - to account for odd little delays that may affect a given journey. Under the old pure program machine system if there were no delays you might get to the other end of the line a couple of minutes early. This new computer system won't allow to you run 10 seconds early let alone up to 2 minutes early, so you're now seeing much more of those built in timetabled stands. Same effect on the District between Barons Court and Ealing.......and not just station starters either. It's created SPAD traps all over the place where none existed before. And whilst in fairness you can't predict door issues, I've had it twice this week where the starter's held me for two minutes and then when they cleared I failed to get a pilot due to the stupid S stock door system which traps small items like pens, batteries and stones in the door runners. By the time I'd dealt with it I'd departed two minutes late. Had the station starters cleared a bit early like they used to, I'd of had a fighting chance of an on time departure. Hey ho. if that's the case can anyone explain why heathrow t 5 and t4 nearly always clear early not just by a few minutes can be as soon train gets into platform even if it doesn't depart for another 15 minutes That area is still controlled by Earl's Court (Northfields desk) and is due to be migrated later this year I believe.
|
|
Tom
Administrator
Signalfel?
Posts: 4,098
|
Post by Tom on May 24, 2019 13:52:15 GMT
I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. It's more simple than that - the programme machines are replaced by the PICU system which does the roles of Programme Machine (i.e the timetable), control circuitry, and lever operation in one, interfacing directly to the valves which move the signal levers. As others have said the old Programme Machine circuitry was state of the art at the time, but often had local quirks which meant that it wasn't fully utilised and often ended up left in one mode. The other thing seen with PICU is routes being called later - I suspect they've been called on the calculated sighting point which is sometimes later than the physical sighting point.
|
|
|
Post by PiccNT on May 24, 2019 16:20:07 GMT
I gather the new computer system at South Kensington tells the old program machines at Earls Court when it is ok to clear a signal and step forward. Because there seems to be no leeway in the computer system - its on time or, err, on time - it now holds many more signals at danger than was previously the case. It's more simple than that - the programme machines are replaced by the PICU system which does the roles of Programme Machine (i.e the timetable), control circuitry, and lever operation in one, interfacing directly to the valves which move the signal levers. As others have said the old Programme Machine circuitry was state of the art at the time, but often had local quirks which meant that it wasn't fully utilised and often ended up left in one mode. The other thing seen with PICU is routes being called later - I suspect they've been called on the calculated sighting point which is sometimes later than the physical sighting point. The latter point is certainly correct. We noticed many signals clearing later than they used to and other strange goings on. PICU does allow individual trains to be put into a certain mode and it's useful when running out of service that through running, or whatever it's called can be selected for that train to prevent it being regulated, signalling permitted. We have had a few modifications in the past few weeks to re-tweak calling conditions. The different behaviour on any given day for example is the starter at Hammersmith WB fast. Prior to PICU, it used to clear, all things being equal, as the train entered the platform and ran over the first block joint. Now, the signal clears just as you are about to berth the train (if not running early) but on occasion it reverts back to how it used to clear. My learned friend on the green line is quite right about potential SPaD traps and I guess WD40a is one signal that now clears very late causing the driver to slow right down.
|
|
Tom
Administrator
Signalfel?
Posts: 4,098
|
Post by Tom on May 24, 2019 16:58:07 GMT
Yes, the WD40's are one group where I've heard complaints that they are more sluggish than previously. They previously cleared with a train in Barons Court platform but now often don't until the train has pulled out.
|
|
|
Post by caravelle on May 24, 2019 18:21:32 GMT
I'm sure there is a solution for this problem.
|
|
Colin
Advisor
My preserved fire engine!
Posts: 11,310
|
Post by Colin on May 25, 2019 1:02:05 GMT
My learned friend on the green line is quite right about potential SPaD traps and I guess WD40a is one signal that now clears very late causing the driver to slow right down. Absolutely bang on the money! After a couple of SPAD's there a manager looked into it and found that control of the signalling in the area had been switched from Earls Court to South Kensington. Nobody had bothered to inform the drivers of this change and so they were caught out by a change in the way the signalling works that they hadn't been made aware of. Thankfully we were at least told in advance when the Acton desk switched over so the signalling changes in that area were less of a surprise. Yes, the WD40's are one group where I've heard complaints that they are more sluggish than previously. They previously cleared with a train in Barons Court platform but now often don't until the train has pulled out. WD40a has effectively become approach controlled now. It won't clear until we're about halfway between it's repeater and the signal itself. I've looked at my speed a few times now and I'm usually at 16mph when it clears.
|
|
Tom
Administrator
Signalfel?
Posts: 4,098
|
Post by Tom on May 25, 2019 22:12:24 GMT
Yes; I was aware the move had taken place but it wasn't widely announced even to the Technical Staff. In my current role I have criticised the lack of operational change notification; previously something like that would have been publicised in the Traffic Circular (and more than just publishing the new phone numbers, there would be a description of the physical changes at Earl's Court and at each site).
It's just as important for those on the maintenance side as the way the Release works at each site is now different, too.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 26, 2019 2:21:53 GMT
Yes; I was aware the move had taken place but it wasn't widely announced even to the Technical Staff. In my current role I have criticised the lack of operational change notification; previously something like that would have been publicised in the Traffic Circular (and more than just publishing the new phone numbers, there would be a description of the physical changes at Earl's Court and at each site). It's just as important for those on the maintenance side as the way the Release works at each site is now different, too. Apart from Ealing Common on our sites which is still as was
|
|
Tom
Administrator
Signalfel?
Posts: 4,098
|
Post by Tom on May 26, 2019 9:21:59 GMT
Indeed, Ealing Common remained unmodified as to bring it in line would have required significant safety circuit mods to bring it in line with the rest of the area.
The only reason that bulletin was sent out was because I insisted on it (and ended up having to rewrite large parts of it)!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 26, 2019 11:23:40 GMT
Yer abit too late though well after the event don’t worry it confused us when we went to test out Hammersmith being the first of our sites to go over to PICU
|
|
|
Post by youngmale1967 on Jun 5, 2019 19:09:07 GMT
When Earl's Court use to control most of the PIC Line we had a lot of issues with Programme Machines (PM), Timing and worst of all problems with Train Descriptions. The CTFS which came in first when T5 opened for the Northfields desk and later went to take other desks Acton Town, Hammersmith (Local & Fast) & Kings Cross desk which PICU is replacing (Started Cockfosters with Northfields desk going in September) was covering a lot of these faults, that's why a number of wrong signals occurring everyday inc. Hanger Lane Junction, Ealing Common, Acton Town, Turnham Green just to name few places. We had constant problems with timing so we left most of the PM in First Come First Served (FCFS) that's why you had station starters clearing early. With PICU they added timing control to the Station starters as this system works as designed, so the train will be held to time. The big changes are signal PB1, PB16, WD35, WD43 these didn't have timings when controlled from Earl's Court. When control of Northfield's goes to South Kensington I would expect the following signals to have a timing added Northfields Westbound starters WR35b, WR32a, Hounslow Central both roads WT2, WT11 & Hatton Cross WW3. We still control Northfields to Heathrow and still have issues with PICU and with timetable gets corrupted, this one of the reasons why signals clear very early at Heathrow T5/T5 & T123 but there is a notice displayed locally to remind drivers to depart on time. This will no longer happen once South Ken takes control. Also PICU the train becomes the PM so if a train runs empty runs early or engineering train South Ken change the train data to First Come mode and also puts trains up as NIS.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 5, 2019 19:18:33 GMT
WD43rt1 was usually left in Auto with the lever on site remaining reverse all day unless something needed crossing over to the Picc a very rare move these days.
CTFS was a pain at times as this would override the traditional TD & if the TD was put up by the Programme Machines especially for District's, as CTFS ruled from West Kensington going west
|
|