|
Post by ertowerty1 on Dec 9, 2021 9:49:15 GMT
Just to start a discussion on the 33.33 Hz and 125Hz signal mains.
When Lots Road was first put into operation it generated at 33.33Hz, Neasden was the same. Greenwich being a tramway generator was 25Hz, later converted to 50Hz, but never supplied power to the underground until the Eastern End of the Central Line was built.
So from the outset, the signal mains were 33.33Hz, no issues with 50Hz contamination on the track or unlikely false operation as the dual element vane (DEV)* was almost bomb-proof.
The Western End of the Central Line was supplied with a Great Western Railway standard 50Hz Industrial Frequency distribution, so to keep the standard 33.33Hz LT signalling equipment, frequncy convertors were installed at each traction substation.
In the 1960's Lots Road was modernised and 50Hz generators installed, so to keep the 33.33Hz signal mains going, frequency changers were installed in virtually all traction substations.
The Piccadilly Line eastern extension was supplied from the National Grid at 50Hz and the signal mains were 50Hz, and no frequency changers installed at this time.
The eastern end of the district line was different again, it had 2 mains both at 33.33Hz the 2 mains were at 600v and 400v nominal but electrically 90 degrees out of phase, so a Dual element vane relay could be operated from each of the mains without the need for track capacitirs. (Resistor fed track circuits would operate perfectly satisfactorily.)
I think that this was due to the concern of interference with the adjacent 25kV 50Hz overhead electricification of the London Tilbury and Southend Line.
Sometime later these 5600v and 400v mains were refered to as "Clean" and "Dirty" mains, can't remember which as which. I would imagine that power such as signal aspects were from the higher capacity 600v main and the 400V main reserved for relay operation. At some stage, this ended up as standard capacitor-fed track circuits.
The frequency changers in the substations had one motor and two genetators mechanically coupled to produce the 90 degree phase shift, and the shafts were correspondingly marked so that the phasing could be maintained following major overhauls. If the generators were mechanically coupled the wrong way round the DEVs would be driven in the opposite direction with all the signals going to red, even if all the track circuits were unoccupied.
Not sure how 125Hz came into being, but guessing that if you are going to have frequency changers then ypu might asd well have a frequency that gives you some economic advantages as well, higher frequencies are more economic in transformer and generator construction, however, volt-drop in signal mains increase.
Fast Forward to SSL upgrade, the parts of the Met and District lines had to converted from 33.33Hz to 125Hz to avoid interference from the new "S" stock trains, and the track circuits were converted to audio frequency. A further 50Hz main was then installed to supply the audio-frequency cabinets.
Thus on the North end of the Met, a failure of the 33.33, 50 or 125Hz supplies would bring the service to a halt. This situation exists to this day, on the basis that it was a temporary installation and would be all swept away when the TBCT was installed a short while later (This was in about 2008)
I wonder if any other members can add any further information to this?
*The DEV is effectivly a 2-phase motor, with the operating torque being provided by a 90 degree phase shift between the 2 windings, which both must be of the same identical frequency. If windings are eneregised by different frequencies, a very feeble and eractic torque would be produced which would not cause significant rotation of the vane.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 10, 2021 10:14:50 GMT
It was only the track circuits that got rid of the 33 1/3hz all other circuits still use 33 1/3hz in those areas.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 10, 2021 10:18:53 GMT
The east end of the District from Campbell Road Sub which is between Bow Road and Upminster has 2 AC Mains 600v and 440v the 440v was reserved for track circuits everything else come off the 600v feed.
They once let a S stock run in a 33 1/3 hz with mitigations in place as it had to get to Upminster to be used as test train and when the track circuits were observed they worked as they should do with the relays operating normally.
|
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Dec 12, 2021 19:12:35 GMT
my memory is that there wasn't a 50Hz main as such, but instead local power pillars taking their feed from the mains at various suitable points. A similar arrangement was used in other places too.
There is of course the 50Hz signal main on the Bakerloo Line, which replaced the 125Hz main in the 1987-91 resignalling.
|
|
|
Post by d7666 on Dec 14, 2021 15:28:38 GMT
Re. why 125 Hz, I’ll suggest but without supporting evidence: need to avoid multiples of 50 and 33.1/3 to avoid harmonics, and avoid 60 and 75 and their harmonics as they appear in telecoms gubbins, TVs etc. 125 Hz is the next one easily arrived at by mechanical gears and or multiples or submultiples of numbers of poles on rotating machines. But this is a guess from my general knowledge I have no specific knowledge in this signalling area.
|
|
|
Post by ertowerty1 on Feb 6, 2022 14:20:53 GMT
British Railways used 83 1/3Hz, now obselete.
Extract from GKRC0758 Iss 1
The arrangement has the motor-generator 90 degree phase shift in common with the East end of the district lines 2-main system.(coincidence?)
The 83 1/3Hz a.c. track circuit is designed to be immune to interference from 50Hz a.c. and d.c. traction systems. The first types of 83 1/3Hz a.c. track circuits operated from a 2 phase 831/3Hz supply, one phase (phase U) being ‘unscreened’ and the other (phase S) being ‘screened’. A phase difference of 90° existed between the two. The 83 1/3Hz a.c. supply was generated by central motor generator sets. The ‘unscreened’ phase was generated at 650V and fed through unscreened cables to the lineside equipment housings where it was stepped down to 110V before being fed to the track and then to the control coil of the relay. The ‘screened’ phase was generated at 110V and fed to the local coil of the relay. In all cases, the power supplies and track relays were situated in equipment buildings thereby confining the ‘screened’ phase entirely to the building. This considerably reduced the likelihood of contamination of the phase by 50Hz interference.
Due to the impracticability associated with the original arrangement, in ensuring the long feed cables from a central supply point are free from 50Hz contamination and the widespread reliability implications of a single motor/generator failure, the 831/3Hz a.c. supply is currently provided by a local static frequency invertor which operates from the general 50Hz a.c. signalling supply.
Additional information concerning these track circuits and similar ones operating at 50Hz a.c can be found in GK/RC0757. The more general requirements can be found in GK/RC0752, GK/RC0753 and GK/RC0754.
Minor edits to carriage returns and spacing to improve readability - no alterations to post wording. (Tom)
|
|
|
Post by zbang on Feb 6, 2022 17:39:23 GMT
Couple of things- Many people would consider 125Hz to be an "audio frequency" (very close to B2 at 123.47Hz, nine steps below middle-C).
125Hz is a 10 pole machine turning at 1500 RPM.
|
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Feb 6, 2022 19:58:34 GMT
Equipment rooms running on the 125Hz supply are generally a lot noisier than those running on 33⅓ Hz. It was much nicer to test without the background hum of all the relays working on 125Hz.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 6, 2022 23:18:24 GMT
You learn to block the noise out of a 125hz room
|
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Feb 6, 2022 23:29:33 GMT
Oh yes. I noticed the silence of a 33⅓ Hz room far more than the hum of a 125Hz one.
|
|
|
Post by d7666 on Feb 7, 2022 12:20:00 GMT
British Railways used 83 1/3Hz, now obselete. The 83 1/3Hz a.c. track circuit is designed to be immune to interference from 50Hz a.c. and d.c. traction systems. Somewhere in my boxes of rubbish cough archives I have some old paperwork concerning testing of BR class 315 EMU, and again later class 317, over the Tilbury tracks adjacent to District tracks exactly because of these frequencies and potential for interference. 315s and 317s had to be proven as these were the first ones in the area with power traction electronics; 315s had to be proven twice over as some with GEC kit some with Brush kit. This testing was generic, not aimed, at the time of testing, that 315s or 317s would work in the area, although 317s later did. I believe 313s and 321s were subject to similar testing but I not seen any evidence to that effect.
|
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Feb 7, 2022 15:21:34 GMT
I assume the original colour light signalling on the LT&S lines used 83⅓Hz track circuits? DC tracks wouldn't have been suitable due to the Underground next door and associated risk of DC traction interference, and 50Hz would have been unsuitable due to the potential for traction interference from the Overhead Line Equipment.
|
|
|
Post by ertowerty1 on Feb 11, 2022 10:13:56 GMT
Just a quick note regarding frequencies: The substation frequncy changers (rotating machines) were (are) powered by induction motors, which always run at less than synchronous speed, so that the nominal frequencies of 33 1/3 and 125Hz were about 4% less, depending on the load. An induction machine slows down with loading and will produce zero torque at synchronous speed, thats the nature of the beast.
As an aside, does anyone test DEVs with rated frequency on one winding and 50Hz on the other to see the response? Apart from the noise, I guess the vane wouldnt do much except vibrate, possibly at the beat frequency.(Signal main Hz plus/minus 50Hz).
What was the in-service performance of the DEV 50Hz track circuits on the East End of the Picc line like?
|
|
|
Post by d7666 on Feb 11, 2022 14:23:13 GMT
I assume the original colour light signalling on the LT&S lines used 83⅓Hz track circuits? DC tracks wouldn't have been suitable due to the Underground next door and associated risk of DC traction interference, and 50Hz would have been unsuitable due to the potential for traction interference from the Overhead Line Equipment. Off top of head, I can't remember that level of detail from those reports; not even sure which box I have buried them in; I'd have to read them again to confirm or deny that assumption; no knowledge of it myself. One reason to check is that part of the LTSR route did have a 1500 V DC period before 25 kV AC; Fenchurch Street to Bow Junction was DC and done part of the GE Liverpool Street - Shenfield - Southend (GE) scheme; now while that LTSR section does not parallel LT tracks, it may have introduced frequency complexications
|
|
|
Post by d7666 on Feb 11, 2022 15:23:04 GMT
This topic lies outside my day job so this is in effect a layman comment.
Ref signalling main AC generation, I have always been under the impression it was? is? generated by DC powered motor alternators off the 630 V traction supply. And it was this use of DC motors dipping speed when the feed circuit was loaded that led to signalling main frequency drift. I have no idea where I have this from and TBH never thought about it until now. If that's wrong then learnt something from this thread.
|
|
|
Post by zbang on Feb 11, 2022 16:50:20 GMT
The substation frequncy changers (rotating machines) were (are) powered by induction motors, It seems a bit odd to me to use non-synchronous motors in that application- changing frequency, unless you're not concerned if the output is stable and not really on-freq. If that's the case, so be it- the 125Hz nominal becomes more like 120Hz, a seemingly small difference but could affect the operation of some resonant devices (e.g. vibrating reeds).
A shunt-wound DC motor is often called a "constant-speed motor" and has rather good stability unless the supply voltages drops significantly.
|
|
|
Post by d7666 on Feb 11, 2022 17:25:31 GMT
The substation frequncy changers (rotating machines) were (are) powered by induction motors, It seems a bit odd to me to use non-synchronous motors in that application- changing frequency, unless you're not concerned if the output is stable and not really on-freq. If that's the case, so be it- the 125Hz nominal becomes more like 120Hz, a seemingly small difference but could affect the operation of some resonant devices (e.g. vibrating reeds).
A shunt-wound DC motor is often called a "constant-speed motor" and has rather good stability unless the supply voltages drops significantly. "Wot e said" Or, more formally, I agree.
|
|
|
Post by zbang on Feb 12, 2022 1:34:03 GMT
Oh ya, assuming the "signal main" is carried continuously through along the entire line with multiple sources, that gets close to being an infinite bus source and causes all generators feeding it to run in lock-step (at the same frequency & phase). That then brings all kinds of interesting problem with connecting a new generator to the bus- have to have the frequency and phase correct or substantial mechanical stresses can occur. (This from my knowledge of power generation, not of the LU signaling systems.)
|
|
|
Post by brigham on Feb 12, 2022 8:42:26 GMT
Hardly 'problems'. If you don't have a synchroscope, you can do it with three light bulbs!
|
|
Tom
Administrator
Signalfel?
Posts: 4,196
|
Post by Tom on Feb 12, 2022 15:49:26 GMT
The signal main is known to have phase variations at on different sections (it's sectionalised based on substations). There is a requirement to ensure that the two coils of a Double Element Vane relay are always fed from the same main because of this.
|
|
|
Post by zbang on Feb 12, 2022 17:33:00 GMT
Hardly 'problems'. If you don't have a synchroscope, you can do it with three light bulbs! Sure, if you have them. How many small converter sets do? Anyway, as tom says, it's sectionalized so if the problem exists, it's probably small. (Off to ask some electrical engineers I know about this....)
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 13, 2022 21:27:38 GMT
Hardly 'problems'. If you don't have a synchroscope, you can do it with three light bulbs! Or 2 analogue multimeters is another trick
|
|
|
Post by d7666 on Feb 14, 2022 22:25:31 GMT
Hardly 'problems'. If you don't have a synchroscope, you can do it with three light bulbs! Or 2 analogue multimeters is another trick IME a bigger issue these days is finding someone under the age of about 40 who can actually use one multimeter to do anything ..... don't get me started these alleged engineering masters that can't. Thread drift. Ceased.
|
|
|
Post by ertowerty1 on Jan 6, 2023 0:15:44 GMT
Re synchronising of frequency changers:
These are fully automatic, just press the button and go, electronics are a lot better/quicker at detecting phase/frequency conincidences. Manual syncing was not a design feature.
The western end of the central line had 3-phase frequency changer outputs but only one phase would be synchronised at any time, speeds up the process as a choice of three phases are availible.
In the early 1970s, many substation frequency changers were run in parallel via the signal mains, until a fault occured on the west end of the district line which induced a cascade failure with supplies being lost over a large part of the railway.
On the east end of the district line a further signal main was installed for the S stock immunisation, cant remember whether 50 or 125Hz, run substation to substation. After that time, the sections were made smaller, thus making it easier to restore power after an outage.
Train service still gets screwed up but hopefully for not as long.
|
|