TI21/EBI cross bonding issue

[magicmaciej]

Hello,

NR standard for EBI states that no cross bonding  (direct or indirect) is allowed between tracks of the same frequency.
Why this is an issue? I understand that impedance bond will block the frequency used for train detection and therefore this should not be a problem. Also with AC electrified railway almost everything is either directly or indirectly bonded to earth/rsc.

Cheers,

Magic

[PJW]

Hello,

NR standard for EBI states that no cross bonding  (direct or indirect) is allowed between tracks of the same frequency.
Why this is an issue? I understand that impedance bond will block the frequency used for train detection and therefore this should not be a problem. Also with AC electrified railway almost everything is either directly or indirectly bonded to earth/rsc.

Cheers,

Magic

I am not an expert but I do know that there have been a number of problems with runaround paths such that a track has become falsely energised as a result of receiving the appropriate frequency from a different track.  London Underground certainly experienced serious problems on the Wimbledon line due to sneak paths via the air main, which is a gerally good but inevitably not perfect earth.

You'd need to read the standard carefully, but perhaps "cross-bonding" means from one rail of one track to a rail of another track in the same area for the purposes of sharing return current between parallel tracks (this sounds particularly applicable to 3rd rail do electrification to me); I'd agree that one way or another there will be some level of interconnection which cannot be avoided.

The role of an impedance bond is to differentiate between the traction return current (50Hz) and the track circuit frequency (around 2kHz) but it will not absolutely "block".  As you have said almost everything is bonded to the return conductor and although this is earthed at certain intervals, it is still possible that it can be the path for track circuit current from one track to leak into another.  The latest variant of these tracks places some digital coding on the signal so that the Rx will reject even if of the correct frequency.

So I don't actually know for sure, but presumably someone who truly understands has set the application criteria and it does seem plausible.

[dorothy.pipet]

I am no expert on bonding in any context, however, a project I am working on is adding "Cross-Bonding" to an existing railway with OLE fitted and 3rd rail d.c. power alongside. The reason is to provide better return paths for fault conditions owing to an OLE upgrade.
Tracks in the area are Reed and TI21 (EBI).

Generally a railway of up to 4 tracks will not have the same track frequencies parallel, therefore each cross-bond is between tracks of different frequencies. As PJW says adjacent Cross-bonds, both being bonded to the Return Conductor, could pass that interference along, or there is the scenario with more than 4 parallel tracks. I found out this would be a problem for "normal" Reed track receivers but TI21 and the higher spec Reed tracks we have are OK.