Where to put train detectors?

[Robin2]

There is no shortage of advice on the internet, including this forum, about what sort of train detection systems to install - IR-DOT, RFID etc etc.

 

However I have not been able to find any advice about where to put the detectors and how to make use of them in a computer control system. For example how does a computer get a train to stop at a station after decelerating in a realistic manner? Or how do you get the front of a train to stop close to buffer stops, or at the correct end of a fiddle-yard?

 

All advice and links to relevant material will be welcome.

 

...R

[Edwin_m]

Depends on your computer system.  Railroad and Co can work by dead reckoning between physical detectors, provided the train in question has been calibrated so the system knows the exact relationship between speed step and actual speed.  Provided it knows the train and track lengths it can also avoid doing things like changing points under a train, even if there is no detection on the points.  Other software may be less "clever". 

 

I've wired my detectors (which are of the current variety) moreorless as track circuits would be laid out on the prototype, basically with a break at each signal and in areas of pointwork enough detectors for any combination of trains to pass through with not more than one train in each detection zone.  I also put a short detection section on the immediate approach to each signal to ensure accurate stops, which RR&Co needed five years ago when I started wiring but probably doesn't now. 

[Robin2]

Thanks Edwin, that's been very helpful.

 

I would also like to hear from anyone who has installed the type of detector (such as IR-DOT) that detects trains at a specific spot.

 

Do you know how the RR&Co software interprets the information it gets from your current detectors?

 

I have downloaded the RR&Co manual and, from reading and thinking about it, it seems that it will identify when a loco enters or leaves a particular section - which is much the same as the information it would get from an IR-DOT detector at each end of a section. However it must make an estimate about where the tail of the train is. But on the plus side, it "knows" that the loco is in the section whereas IR-Dot detectors could not tell that unless the train is blocking one of them - which, of course, allows the IR-DOT detector to "see" all of the train, and not just the loco. 

 

I want to write my own software (because I like doing that) and I want to minimize the number of detectors - partly for cost reasons but also because I can read a few (say 4 at a time) with a very simple PC interface. Also, I am thinking of converting to radio control and battery power so there may be no track current. ( http://www.rmweb.co.uk/community/index.php?/topic/64616-battery-poweredradio-controlled-locos/page-4 )

[Edwin_m]

The RR&Co software is almost (but not totally) like a prototype signalling system so you can use the detectors to control the aspects of signals etc as well as displaying them on the screen.  It also drives trains automatically so needs to make sure there are no other trains in the way.  With spot detectors I presume you can flag a section as occupied when the first detector is triggered and unoccupied when the second one is "un-triggered", though it needs to take account of train direction and the possibility a train may reverse in section.  Can't comment in any detail on this as I haven't used them. 

 

I have opted to fit resistors to rolling stock axles so that the whole train will be detected by the current detectors.  Alternatively the software can work out when a particular section is still occupied if you tell it how long the sections and all the trains are.