Reduce wiring for train detection?

[Slawman]

Hi All, I am growing very tired of maintaining the web of cables needed on my system to operate train detection. I currently run a partially automated setup so the detection modules are critical and I need lots of them. I use the Massoth reed detector run back through a detection module. Is anyone aware of a solution I can run without needing wires? I like the little 8156601 Switch Decoder 1K II and have installed these in most of my points, so no wires needed back to my central station. Just looking for something similar for the detection modules?

 

[The Shed]

This is my own simple implementation on Automation.

Slightly confused by this "I am growing very tired of maintaining the web of cables needed on my system to operate train detection"
Does this imply you are running individual cables from each Feedback Module back?, in which case yes, I'd get tired and frustrated maintaining the Rats Nest of wires.

I solved the problem initially by creating Satellite Points to group together sections, with the use of the Bus Adapter Module, then running a Single Dimax Bus cable to the next Satellite Point, and then onto the CS, picture below gives the general idea.

With the move to Wireless Transmission, the problem of cabling no longer exists.

 

[Slawman]

That is a very neat implementation. I love the ferrules and coated track connections, I am sure that cuts down on maintenance significantly.

My particular challenge involves these detection reeds:



Last count I had nearly 50 of the darn things and they are all wired back to a detection module. This is in addition to isolated track sections. That is the particular wiring that I am finding problematic.

I am after something that can send it's signal back along the track to the Central Station (if it exists)?

 

[The Shed]

Ah! with you now, other than daisy chaining the GND wire to become a Common between the A & B contacts, which in itself will reduce the wire count, not significantly, but any reduction is better than nothing.

Massoth do have these available, the PZB/IR Signal decoder and the ITC/IR Receiver, the question of the sheer economics may make this impractical.

Alternatively, I moved from using Track Contacts to Track Detection, again the practical considerations of altering and fitting Isolating sections may be a step to far.

The Track Detection system, there is a reduction in the amount of cable, BUT does need careful thought and consideration from an operational point of view, not a simple task to rectify if not thought through in the first instance

Although saying the above, it DOES provide alternative scope for Automation and Animation, which the Track Contact system cannot provide, well not as they are provided by Massoth, although anyone who is prepared to self build alternative Track Contacts, could utilise alternative Automation and Animation routines.

Any method as above will require additional work and expense, unfortunately no simple solution is at present available, best to hope for, is to contain the Rats Nest of wires to a manageable level!

Given all the above, I would still always install some method for Automation and Animation, makes for interesting and surprising operating sessions!

 

[Slawman]

Continuing my research. Has anyone used RFID successfully? Conceptually it would work but in practice outdoors.....

 

[FatherMcD]

Continuing my research. Has anyone used RFID successfully? Conceptually it would work but in practice outdoors.....

Check out this thread Location detection using RFID Michael has been using it and could possibly expound on practical experience outdoors. (Though he hasn't posted since 2021)

 

[Greg Elmassian]

YEAHBBUT....

He runs a system where the intelligence is in the locos... the loco determines where it is and reads it's location as it moves around and determines what to do.

The OP is using a system where the brains are not in the locos...

so the layout needs to know where the trains are, not vice versa as in the thread mentioned.

Greg

 

[Slawman]

YEAHBBUT....

He runs a system where the intelligence is in the locos... the loco determines where it is and reads it's location as it moves around and determines what to do.

The OP is using a system where the brains are not in the locos...

so the layout needs to know where the trains are, not vice versa as in the thread mentioned.

Greg

Exactly right Greg however the detection event could still be relevant for the automation. The software I am using relies on knowing the start point of the loco and then "assuming" the next event along the track is that loco (if you have properly positioned it then it could not be anything other than that loco). Assuming the track tags could be location specific I can conceive how that might work....

 

[Greg Elmassian]

Yeah, I made some guesses on your system.

How does your system work? What software? His system gets the info to the loco, is there a way you could read it back with your software?

I'm looking at what you are asking, i.e. some form of limited automation, mainly for running multiple trains on the same track, I am DCC now, and also will be bringing up railcom support

 

[Slawman]

Yeah, I made some guesses on your system.

How does your system work? What software? His system gets the info to the loco, is there a way you could read it back with your software?

I'm looking at what you are asking, i.e. some form of limited automation, mainly for running multiple trains on the same track, I am DCC now, and also will be bringing up railcom support

I am using TrainController (Silver) for automation, connected to a Massoth 800Z via RS232. Massoth gathers the detection through traditional means :> Track detector activated by Magnet on loco :> Back to feedback module :> Into 800Z : > Into Windows based system running TrainControl.

Here is an older snapshot of the layout which has grown a bit but serves to illustrate my application:




Each of the Blocks has three detection points and associated wiring back to a feedback module. The system can run several trains fully automated and also semi automated if I am feeling like that.

Unclear how the signals may be brought back into the system at this stage however i am somewhat handy with integration so may be able to rig something up once I identify a potential solution. The signal is simple low level on/off so nothing too complex on the face of it although the fact that no-one has come up with a wireless solution to date makes me think it may be more difficult......

Best regards,

Jordon

 

[korm kormsen]

i am not sure, if i can explain clear enough, what i found out with an experiment.
using DC analogue. two transformer/controlers. both sharing one rail as common. (gave me an uninterrupted common everywhere) and the other rail interrupted and connected one section per controler.

when i installed the "AC - reed - switch (turnout or section turnoff) - back to AC", i connected one of the cables from the AC to that common rail mentioned above. so i could connect the "in" of the reeds just to the common rail beside it. (sure, from reed to switch and back to AC i needed cables as before)

that saved me at least a third of the cables.
i tested that for three or four hours, and nothing blew up.

 

[Slawman]

i am not sure, if i can explain clear enough, what i found out with an experiment.
using DC analogue. two transformer/controlers. both sharing one rail as common. (gave me an uninterrupted common everywhere) and the other rail interrupted and connected one section per controler.

when i installed the "AC - reed - switch (turnout or section turnoff) - back to AC", i connected one of the cables from the AC to that common rail mentioned above. so i could connect the "in" of the reeds just to the common rail beside it. (sure, from reed to switch and back to AC i needed cables as before)

that saved me at least a third of the cables.
i tested that for three or four hours, and nothing blew up.

Hi Korm, I do not follow what you have done but it is good nothing has blown up Do you have a diagram? What is the protocol / Interface you are using to bring the detection signal into your system?

I do not think the solution I am after has been developed (yet). RFID has potential as do some of the Wifi standards out there as well as broadband over power lines (BPL) technology. I use a micro WiFi 6 module on one of my other projects which is 2cm x 3cm and very inexpensive. I am going to have a play around to see if I can get a block set with one of these chips to communicate back through Arduino on one of these chips.

As with many of our model train specific technologies, there is quite a lag between development and implementation given the low volumes involved (in comparison to developing electronics for TVs for example). There are a few technologies with the potential here but nothing I have found specifically "plug and play" which conforms to any standard.

 

[Greg Elmassian]

What I was stating is the RFID example has the sensor IN THE LOCO, so how would you get that data to your system in the computer? Since the loco is moving, no wires.

That's why I commented in the first place.

Now if the sensors were in the track, you would be golden, and you could put RFID tags on your locos and cars, but now I think the cost of sensors is prohibitive, indeed the author explained that putting the sensors in the loco let him do it more cheaply (many fewer locos than track positions)

Greg

 

[Slawman]

What I was stating is the RFID example has the sensor IN THE LOCO, so how would you get that data to your system in the computer? Since the loco is moving, no wires.

That's why I commented in the first place.

Now if the sensors were in the track, you would be golden, and you could put RFID tags on your locos and cars, but now I think the cost of sensors is prohibitive, indeed the author explained that putting the sensors in the loco let him do it more cheaply (many fewer locos than track positions)

Greg

The RFID sensor in the loco would transmit a tag detection back to whatever you were using to receive wirelessly, which would then emulate an open/close to the 800Z (or a detection module in between). The RFID tag can be unique but no idea if this is a configurable parameter for everyday fiddling. My knowledge of RFID comes from supermarket implementations where they use RFID to track stock movements by line and the tags are in the goods and specific to that line ie. the tomato soup carries RFID tags that tell the system it is Tomato Soup. This is all just thought bubble stuff Greg, I have no practical experience to back-up my theorising and may have also not properly understood the limitation you are highlighting.

 

[Greg Elmassian]

Yeah, the obstacle would be getting the loco to transmit back to your software AND besides the RFID it scanned, it would need to also transmit the loco number.

Very nice, but how to interface? Not much data, but you would need a little micro in the loco, plus a little transmitter, and of course receiver on your software system.

Much more involved than if the sensor was under the track and wired to your computer.

Greg

 

[korm kormsen]

this is the best, i can do.
my point was, to spare the black cable from AC to reed by using the blue rail as common.
i use the LGB switchmotors and their add-on switches.

cable colours also follow LGB code.
red and blue = DC
black and white (light grey in the pic) = AC
half wave AC = orange

 

[Cliff George]

I am using TrainController (Silver) for automation, connected to a Massoth 800Z via RS232. Massoth gathers the detection through traditional means :> Track detector activated by Magnet on loco :> Back to feedback module :> Into 800Z : > Into Windows based system running TrainControl.

Here is an older snapshot of the layout which has grown a bit but serves to illustrate my application:

View attachment 312800


Each of the Blocks has three detection points and associated wiring back to a feedback module. The system can run several trains fully automated and also semi automated if I am feeling like that.

Unclear how the signals may be brought back into the system at this stage however i am somewhat handy with integration so may be able to rig something up once I identify a potential solution. The signal is simple low level on/off so nothing too complex on the face of it although the fact that no-one has come up with a wireless solution to date makes me think it may be more difficult......

Best regards,

Jordon

Hello Jordon,

I have some experience with using traincontroller, I wonder if I can help you.

Generally, there is less wiring with using occupancy detectors rather than momentary detectors, think about it, you get a notification on entry to either end of the block, for which you would need two momentary detectors. A disadvantage of occupancy detectors is possibly that you need to gap the rails, with proper planning I didn't find this an issue.

You do not need three detectors per block, with well profiled engines one occupancy detector with virtual stop and brake points is just fine, that is what I use (I am aware of what the traincontroller manual says). Even if you do require absolute brake and stopping points some detectors are redundant. For example, if a train cannot stop in a block anything more than one occupancy detector in the block is a pointless waste. If a train can only stop in one direction (i.e. at the end of a siding, or on a double track main line) only a maximum of two detectors are required.

I wanted to keep all my electronics indoors. Each of my blocks is wired up with an individual twisted track power lead and the occupancy detectors located indoors where the block power leads start. This keeps all of the occupancy detectors together indoors and makes wiring everything up simpler. I use the type of detector where one of the track power lead wires is passed through a device on the detector. This arrangement means that outdoors you get a sort of star effect with wires (which have to be fairly hefty to account for voltage loss) going to each block (and non blocked track). There is more track wiring than you would get with a non detected layout, but no detector wiring outdoors. I also use pneumatics to operate my points to avoid electronics outside.

Hope this helps. Any questions or comments are welcomed.

Regards
Cliff

 

[Dagnall]

I think you should look at the Rocrail solutions. I use hall momentary sensors (more reliable long term than reed relays that seem to break over winter) connected to a WIO (wireless I/o module). Each WIo is an esp32 based WiFi and can be diy , so it needs a power supply, but easily can support multiple io. My setup uses max 16 sensors and each WIO drives up to 4 servo driven points. So an ideal layout has a WIO at each station. (At least for me!) I just wire 12v to each station. I could extract power from the track dcc, but prefer to keep that separate.
I do not see much stuff about Rocrail, but it is well worth exploring!
This is my diy (16 io+ 4 servo) WIO board, but you can buy much better kits ) ) from links on the rocrail site.

 

[Slawman]

Hello Jordon,

I have some experience with using traincontroller, I wonder if I can help you.

Generally, there is less wiring with using occupancy detectors rather than momentary detectors, think about it, you get a notification on entry to either end of the block, for which you would need two momentary detectors. A disadvantage of occupancy detectors is possibly that you need to gap the rails, with proper planning I didn't find this an issue.

You do not need three detectors per block, with well profiled engines one occupancy detector with virtual stop and brake points is just fine, that is what I use (I am aware of what the traincontroller manual says). Even if you do require absolute brake and stopping points some detectors are redundant. For example, if a train cannot stop in a block anything more than one occupancy detector in the block is a pointless waste. If a train can only stop in one direction (i.e. at the end of a siding, or on a double track main line) only a maximum of two detectors are required.

I wanted to keep all my electronics indoors. Each of my blocks is wired up with an individual twisted track power lead and the occupancy detectors located indoors where the block power leads start. This keeps all of the occupancy detectors together indoors and makes wiring everything up simpler. I use the type of detector where one of the track power lead wires is passed through a device on the detector. This arrangement means that outdoors you get a sort of star effect with wires (which have to be fairly hefty to account for voltage loss) going to each block (and non blocked track). There is more track wiring than you would get with a non detected layout, but no detector wiring outdoors. I also use pneumatics to operate my points to avoid electronics outside.

Hope this helps. Any questions or comments are welcomed.

Regards
Cliff

Hi Cliff, I like the sound of your approach. The wiring would be reduced significantly and bringing all the gear back to a central, protected location is attractive. I am going into a research phase and I will consider and trial your approach in the process. I have also toyed with the idea of pneumatics however I have found the points switches I bought (before I was aware of a pneumatic solution) are still going strong after 7 Sydney summers but might consider this if and when they start to fail. I treat all my switches with Innox once a year and many of them still look almost new inside. How long have you been operating on this design and which software do you use for automation? Regards, Jordon

 

[Slawman]

I think you should look at the Rocrail solutions. I use hall momentary sensors (more reliable long term than reed relays that seem to break over winter) connected to a WIO (wireless I/o module). Each WIo is an esp32 based WiFi and can be diy , so it needs a power supply, but easily can support multiple io. My setup uses max 16 sensors and each WIO drives up to 4 servo driven points. So an ideal layout has a WIO at each station. (At least for me!) I just wire 12v to each station. I could extract power from the track dcc, but prefer to keep that separate.
I do not see much stuff about Rocrail, but it is well worth exploring!
This is my diy (16 io+ 4 servo) WIO board, but you can buy much better kits ) ) from links on the rocrail site.

Thanks Dagnall, I have started to research Rocrail's approach and technology. What software and central station do you operate?