KATO 20-652 extend a Automatic Crossing Gate detection track

[JanW]

I had an issue with the detection tracks that KATO supplies with the KATO 20-652 Automatic Crossing Gate. These are only available as 62 mm tracks. 
In my lay-out the rail crossing is located in the tracks leaving the station. The barriers should close when a train leaves the station. If I installed a single detection track after the turnout before the signal, the barriers would close too late. Because the crossing is quite close to the station, I needed two detection tracks, one at the end of each platform such that the barriers close when either train A or B leaves the station (see photo). 
This requires the left detection track to be 64 mm long. I first try to use a 62 mm track and squeeze the platform in between. That proved to be too tight an forces the outer edge of the track upwards. So I needed to make the detection track longer... 

After ample consideration I dared to disassemble the detection track and saw one of the ends off. Using the end of the ballast bed and the rails of a 64 mm track I managed to extend the detection track to 64 mm. See photo's for the result. Now all fits well around the platform.

[cteno4]

Nice work! Great to get in there and get it solved!

 

jeff

[Yavianice]

So you can hook up 2 detection tracks per side out of the box? Or did you alter this too?

[JanW]

21 hours ago, Yavianice said:

So you can hook up 2 detection tracks per side out of the box? Or did you alter this too?

Yes you can!  Each track in the crossing comes with 4 connectors for detectors: 2 for each direction. There is some fuzzy logic that closes the barriers when one of these detectors detect a train (or your finger for that matter) and opens the barrier when the track is cleared again. If, in the mean time, another track commands the barriers to close, they remain closed until also the other track cleared. Sometimes the logic is confused and the barriers remain closed. Then it is time to reset (a small button you can activate near the barrier with a pin).
Interestingly, the detectors also get confused by direct sunlight. Apparently there is enough infra red in the sunlight to activate the detector.  

[kvp]

The detectors are counting the entry IR reflection passes for each direction. The exit detector substracts from this count and when both directions reach zero, the crossing is cleared. There could be problems from curves and ambient light. Curves put different parts of the same train over the entry and exit sensors, so they produce different patterns, while direct sunlight has enough infrared (heat) power to blind the sensors. It's easy to test the sensors by swiping a finger a number of times above the entry sensor and then the same times above the exit sensor (in the same direction).

[JanW]

Something I just found out: De detector tracks cannot be in a tunnel. The tunnel ceiling will reflect the IR and immediately upset the crossing.. 

 

[kvp]

15 hours ago, JanW said:

Something I just found out: De detector tracks cannot be in a tunnel. The tunnel ceiling will reflect the IR and immediately upset the crossing.. 

They can be, just make the tunnel ceiling high enough or IR absorbant. Reflecting the beams sideways at an angle with any IR reflecting material could also work.

 

ps: Having decorations, like trees, catenary poles or highway overpasses above IR detector tracks is also a problem. For exhibition layouts, even a visitor bending over the layout, halogen lights or skylights on the ceiling or the flash of a camera could trigger a detector.