Track occupancy indication without using control software

[Harry Lime]

Having used DCC for many years with Loftus Road, I am now experimenting with an aspect of DCC that is new to me: track occupancy. However, I do not want to go down the path of buying computer software. Instead, what I am trying to do is to see whether it is possible to set up track occupancy sections with LEDs on control panels instead.

 

I currently have a Lenz system (LZV100 v3.6). I also have a single LDT RS-8 unit. I have now created a three section, single track (one rail isolated) above as per the LDT instructions on their website (page 1132).

 

Where I would appreciate some guidance is on what to some, may be an easy thing to do. I would like to run a series of green LEDs that indicate whether a specific section is occupied or not. Ultimately, these will be built into the control panels for my layout. These would either be off, indicating no occupancy, or on, showing the presence of an engine (for the purposes of testing).

 

Is there a way of adding in these LEDs to what I already have, or do I need to do more in order to make this happen?

 

Thanks in advance of anyone who can provide any suggestions.

 

[Harry Lime]

Have built some brass kits over the years, so I will say yes.

[Phil S]

As usual, there are about 3 options - depending on your preferred interpretation of 'train detection'

1/ And many think of this as the first and only option - electrical detection by creating a series of electrically isolated sections whose current passes through sensitive sensor coils;

To my mind, this goes totally against the principles and benefits of 'dcc', and protentially fails to meet its description:  It requires lots of isolated sections - whereas dcc technically required only 1 - thus it affects your total dcc wiring implementation!   Secondly: it only detects 'current consumers' - which may be a loco, or one end of a multiple unit, perhaps. I DO  believe in 'sectioanlising' dcc power districts - but for the purpose of easy-fault-finding of shorts etc., and not following the requirements of either analogue sections of the past, or train detection of the future.

 

Merg members may use a TOTI (Train ON Track Indicator) kit for this method.   The results can be fed to your preferred follow-up system to collate and display results - such as latching LEDs.

 

2/ Magnetic, and using Reed Relays in the track for detection of an item of stock fitted with a suitable magnet.   Roco provide 'subsitute sleepers' with an inbuilt reed-relay for their un-ballasted track, and reed-relays can clip into their Geoline track, for example.  This method has been long established, and provides a precise trigger from suitably fitted stock - eg for stopping trams in the correct place.

 

3/ Optical Detection: The ONLY method which will fully-detect a whole train - even with plastic wheels -  the incarnation might use  a simple beam-break, or require reflection by the presence of stock.

One caution though: Some plastics which might appear opaque to 'us', may be transparrent to the Infra-Red.  Also, some designs may be affecrted by variations in overall lighting level (eg day/night)

[Nigelcliffe]

I think Dutch Master meant electrical/electronic soldering, but brass kit building is a reasonable start.  

 

With that in mind, you need a simple and cheap occupancy detector.  There are basically two types; those which measure current flowing through a track section (and thus detect a decoder equipped loco, or a coach with lighting inside it),  and those which detect at a single spot (eg. reflecting an infrared beam, or a magnetic switch from a magnet under a piece of stock).    I'd guess the first type from the LDT module you already own.

 

MERG (UK hobby group) publish a few designs and kits for measuring the current flowing through the track, the simplest is a single detector for £1.50, containing a handful of components.    I think membership is required to gain access to these designs.

[Harry Lime]

That makes for a better set of choices. My original reading around the subject only revolved around current detection and using DCC, so I have probably become rather fixated on this approach. I am minded though to look at Phil S's suggestion of optical recognition as this seems to offer a sensible way of achieving what I aiming to do.

 

Thank you to everyone for their input.

[Gordon H]

1/ And many think of this as the first and only option - electrical detection by creating a series of electrically isolated sections whose current passes through sensitive sensor coils;

To my mind, this goes totally against the principles and benefits of 'dcc', and protentially fails to meet its description:

 

Surely layouts ought to be based on prototypical methods wherever possible, and current detection is certainly one of the nearest approximations we have to that.

Too bad if it doesn't suit the frankly ludicrous idea of a layout being wired as one whole section with DCC. To me that simplistic approach makes for lazy designs and almost impossible fault finding.

 

MERG (UK hobby group) publish a few designs and kits for measuring the current flowing through the track, the simplest is a single detector for £1.50, containing a handful of components.    I think membership is required to gain access to these designs.

 

As a MERG member yourself you know full well that membership is required to obtain kits, so why pretend otherwise?

[Crosland]

Plenty of ideas for detection circuits on Rob Paisley's site http://home.cogeco.ca/~rpaisley4/CircuitIndex.html

 

Andrew

[Phil S]

1/ And many think of this as the first and only option - electrical detection by creating a series of electrically isolated sections whose current passes through sensitive sensor coils;

To my mind, this goes totally against the principles and benefits of 'dcc', and protentially fails to meet its description:

Gordon commented:

Surely layouts ought to be based on prototypical methods wherever possible, and current detection is certainly one of the nearest approximations we have to that.

Too bad if it doesn't suit the frankly ludicrous idea of a layout being wired as one whole section with DCC. To me that simplistic approach makes for lazy designs and almost impossible fault finding.

 

Phil responds: Franlky I consider trying to combine an artificial 'occupation related' series of sections - which need to be numerous and complex around junctions to be a ludricrous scenario when also totally against the needs of dcc, and counter to effective dcc-sectionalising which would aid diagnostics: I do believe in easy diagnostics being built in to layouts - but the sensible divisions are unrelated to anachronistic artificial boundaries inherited from analogue control days.

TO keep in line with Gordon's advocacy:  a Series of Smoke detectors mounted above the layout ahead of each signal would be more in keeping with 'the prototype!'  8-) but would STILL only detect the engine - as do most electrical occupancy detectors - UNLESS all coaches have lighting or other current consumers built-in (admittedly, I can now say that all our H0 and G-scale coaches, and many of our 00 caoches DO have internal lighting powered from the track ...but it still does not apply to freight wagons, but does to some of our brake vans - offering the possiblility of rear-of train detection.

By using a TOTALLY INDEPENDANT form of occupancy detection - in my case optical - I do not cause ANY fault in one system to bring down any other system on the layout - for fault-finding this is therefore a great improvement. Remember also, that a model railway is not a life-saving device, in legal terms. 8-)  My dcc sectioning for my loft layout uses 'geographical' location as its basis to help locate any fault quickly - rather than track topology,

 

However:  out in the G-Scale garden, where the risks and priorities are different, I do  have just 1 circuit, with many track feeds - to ensure reliability of supply, and all 24 sets of points are actually powered off the track bus, rather than 1 or 2 separate additional circuits - here the KISS principle is used 'instead':- points are remotely operated, usually multiple times, so that they can be heard to operate cleanly, and are not blocked by a recent twig or catkin fall 8-(   ... audible diagnistics.

 

Following mention of MERG kits: membership would bring Access to 'Hector' optical detector kits , as well as Gordon's own range of information reporting modules RPC, and other ideas.

[Crosland]

Phil responds: Franlky I consider trying to combine an artificial 'occupation related' series of sections - which need to be numerous and complex around junctions to be a ludricrous scenario when also totally against the needs of dcc, and counter to effective dcc-sectionalising which would aid diagnostics: I do believe in easy diagnostics being built in to layouts - but the sensible divisions are unrelated to anachronistic artificial boundaries inherited from analogue control days.

 

I don't see the problem. Adding detection is not sectioning the layout in the same way one would do for diagnostics, by which I assume you mean independent power districts using multiple boosters or block cut-outs.

 

Detection simply requires droppers be fed through a passive detector. Most of the droppers should be there anyway and it's not difficult to arrange for gaps in the track to create each detection zone. It can be as simple or complex as you want to make it. For indication on a control panel as in the OPs case, I don't see the need for it to be too onerous.

 

Andrew

[LNERGE]

I have used the 8 way MERG boards coupled to their four way relay drivers and i am quite happy with the result. Here is an early incarnation of the box diagram..

 

 

All signal controls are relay based though i have found my signal spacing is a bit critical (i can only just drive two trains on greens on my roundy roundy (continuous 4 aspect signalling))

 

To maintain track occupancy with prototypically arranged track circuits i've had to fit 'treadles' to apply a resistor across the rails rather that fit every wheelset with a resistor.

[Harry Lime]

That looks exactly like what I am trying to achieve!

[Dagworth]

1/ And many think of this as the first and only option - electrical detection by creating a series of electrically isolated sections whose current passes through sensitive sensor coils;

Detection doesn't have to use coils, simple solid state electronics works equally well. A couple of pairs of diodes wired back to back and with the input side of an opto-isolator in parallel gives a very sensitive current detector that will pick up as low as a single 10k resistor across the track (resistive wheelsets are easy to add to any vehicles that need to be detected that don't have motors/lights). If the output of the opto is then fed into a pulse stretcher circuit the detection becomes fast acting and slow release, exactly what I need for my own detection

This version uses a short-circuited bridge rectifier as the diode drops.

 

Andi

[Gordon H]

The main downside to diode based detectors is their inherent voltage drop, which could cause variation of speed if your decoders are unable to compensate for it. Of course this doesn't apply if all your track is detected this way using the same circuitry. Coil based detectors have no voltage drop as they only use the track feed wire which is there anyway

Another potential advantage of the current transformer (coil) method which few have exploited as yet (myself included) is that you can run the feed of more than one piece of track through the same coil while maintaining complete electrical isolation between them.

By doing so it becomes possible to assign, say, a single coil to cover detection over a diamond crossing, which will work for both routes. Combinations of such coils could be employed at complex junctions in the same way. Just needs a bit of logical planning.

[Dagworth]

The main downside to diode based detectors is their inherent voltage drop, which could cause variation of speed if your decoders are unable to compensate for it. Of course this doesn't apply if all your track is detected this way using the same circuitry.

For exactly this reason I've had to fit additional shorted rectifiers into the district feeds for the fiddle yards and NCB section of Ravensclyffe which are not track circuited

Another potential advantage of the current transformer (coil) method which few have exploited as yet (myself included) is that you can run the feed of more than one piece of track through the same coil while maintaining complete electrical isolation between them.

By doing so it becomes possible to assign, say, a single coil to cover detection over a diamond crossing, which will work for both routes. Combinations of such coils could be employed at complex junctions in the same way. Just needs a bit of logical planning.

That's cunning, I hadn't thought of that

 

A