 |
Re: Doppler based RF signal activation?
Author: OldPoleBurner
Date: 10-05-2016 - 20:24
We are are probably referring to the same thing with different semantics; a two element AC vane relay vs a two phase AC vane relay.
In such a relay as I was referring too, there are two coils oriented 90 degrees from each other; the magnetic fields of which both act upon a round metal disk called a vane, to make it try to rotate, much like the old fashioned PG&E kilowatt-hour meter, which actually does rotate in analog to the energy (KWH) passing through it. In fact, there is much physical similarity between the two.
When correctly energized, the vane rotates in the correct direction working against gravity, pulling up the heavy contact armature connected to it, thus in one unified mechanical action closing the front contacts of the relay after opening the back contacts.
One coil, called the local coil or sometimes the reference coil, is connected directly to a local power source, typically PG&E in much of California. The other coil, called the track coil, is in the simplest installations, connected through a coupling transformer directly to the track at one end. The other end of the track is connected through a current limiting load resistor to the secondary coil of a "lossy" transformer (similar to a doorbell transformer only much bigger); the primary of which is connected to the same PG&E source as the local coil, in the specified polarity.
The combination of the resistance of the load resistor and the reactance of the transformer, along with the inductance in the track itself; significantly shifts the phase of the track coil (up to 90 Degrees). Thus there are two phases acting upon the vane, causing it to rotate in a specific direction. But if the phases are backwards, the relay will not pick up, thus polarizing the AC current traversing the rails. Moreover, shunting by a train itself also shifts the phase of the track current.
So not only is train shunt measured by how much current is being diverted away from the relay during occupancy; but also it is being measured by how much phase shift is being presented to the track coil. Therefore, in order to actually rotate the vane to pick up the relay, all the electrical properties presented to the relay must be proper, energy level, phase angle, and polarization must all be as designed.
This set up provides exceptional sensitivity to shunting (even an exponential response), relatively strong immunity to momentary loss of shunt (bouncing occupancy) even on non-shiny rail, rejection of foreign current (from imposter sources), excellent detection of circuit faults such as electrical component failures, bad insulated joints, broken rails, etc. In fact, any failure of any component of the track circuit, including the track itself, will cause a safe default to the occupied state. All these complex interactions must be working as designed to pick up the relay and post the block as clear.
Hence, in my view, they are the best technology there is for train detection. The down side? Track circuit length is typically limited, with circuit design being harder and harder the longer they get.
In such a relay as I was referring too, there are two coils oriented 90 degrees from each other; the magnetic fields of which both act upon a round metal disk called a vane, to make it try to rotate, much like the old fashioned PG&E kilowatt-hour meter, which actually does rotate in analog to the energy (KWH) passing through it. In fact, there is much physical similarity between the two.
When correctly energized, the vane rotates in the correct direction working against gravity, pulling up the heavy contact armature connected to it, thus in one unified mechanical action closing the front contacts of the relay after opening the back contacts.
One coil, called the local coil or sometimes the reference coil, is connected directly to a local power source, typically PG&E in much of California. The other coil, called the track coil, is in the simplest installations, connected through a coupling transformer directly to the track at one end. The other end of the track is connected through a current limiting load resistor to the secondary coil of a "lossy" transformer (similar to a doorbell transformer only much bigger); the primary of which is connected to the same PG&E source as the local coil, in the specified polarity.
The combination of the resistance of the load resistor and the reactance of the transformer, along with the inductance in the track itself; significantly shifts the phase of the track coil (up to 90 Degrees). Thus there are two phases acting upon the vane, causing it to rotate in a specific direction. But if the phases are backwards, the relay will not pick up, thus polarizing the AC current traversing the rails. Moreover, shunting by a train itself also shifts the phase of the track current.
So not only is train shunt measured by how much current is being diverted away from the relay during occupancy; but also it is being measured by how much phase shift is being presented to the track coil. Therefore, in order to actually rotate the vane to pick up the relay, all the electrical properties presented to the relay must be proper, energy level, phase angle, and polarization must all be as designed.
This set up provides exceptional sensitivity to shunting (even an exponential response), relatively strong immunity to momentary loss of shunt (bouncing occupancy) even on non-shiny rail, rejection of foreign current (from imposter sources), excellent detection of circuit faults such as electrical component failures, bad insulated joints, broken rails, etc. In fact, any failure of any component of the track circuit, including the track itself, will cause a safe default to the occupied state. All these complex interactions must be working as designed to pick up the relay and post the block as clear.
Hence, in my view, they are the best technology there is for train detection. The down side? Track circuit length is typically limited, with circuit design being harder and harder the longer they get.
