Model Railroading > Train Detection for Signals - Which one do you use?

For those considering the installation of operating signals on a layout, Dave Abeles' book "Guide to Signals and Interlockings" is a must read.  He not only explains signals, but discusses implementation on a layout.  He even steps thru the software to set-up a new signal install.  By allowing me to "look over his shoulder" as the parameters were entered, it removed a lot of the mystery.  I am inspired to give it a try.

One of the points Dave stresses is the importance of ACCURATE detection of trains.  (If the detection system is not reliable, then trains could be anywhere... with disastrous results.)  Since the detection is at the core of a signalling system, I was curious which one(s) are being used by folks that have working signals on their layout.   

As I understand it, these are the main choices:

1) Infrared - Relatively inexpensive, but can be affected by the room light or sunlight.
2) Current Flow - Requires resistors to be installed on trucks of rolling stock to provide a current draw, so it can be detected.
3) RFID - (Radio Frequency IDentification) Requires a small computer chip to be installed on all equipment.  It not only indicates if a loco/car passed a detector, but which loco/car it is.

How would you rate the difficulty of installation?
How about troubleshooting?
If you had it to do over again, would you use the same detection system?

I'm kinda' leaning towards RFID, but the hardware is not exactly cheap. I am intrigued by the system by TrainTraxx.  They use WiFi to transmit the data to a hub, and then is sent to the signalling system.  Has anyone installed this system?
https://www.traintraxx.com/products-services/traxx-id/

Thank you in advance for sharing your experiences.  :-D

Dave

 



Edited 5 time(s). Last edit at 06/23/22 11:45 by tmotor.

I have used both 1 and 2 and they can be made to work well.  If you were doing an isolated interlocking location I would go with 1.  For a full ABS/CTC signal system I would recommend 2 and yes resistors on rolling stock.

No experience with 3.  

All detection can be a bit of a pain and I recommend experimenting first with your desired detection to get the "bugs" worked out early.  

Jeff N
Olathe

BMRC has used No. 2 for 30 years.  One recommendation for this system would be to be sure each individual block detector has its own pot for sensitivity adjustment.  Some detectors are bunched together in groups of 4 or 8 detectors with only one pot for 4 detectors.  If some blocks are different in length in one group, it's hard to get all four set right to pick up a 10K ohm resistor.  For example, if a switch is treated as a block and switches and blocks that are long runs of track are on the same 4-block detector group, it's about impossible to get them all working with one sensitivity setting.   If you're starting from scratch with this in mind, the blocks could be carefully grouped with like-sized blocks.

Our model railroad club uses the infrared system by Azatrax and it works quite well.

Bob Krieger
Cheyenne, WY

3 has some interesting applications beyond simply occupancy detection.  Each car has a unique RFID tag on it which can track the locations of each car and locomotive on the railroad.  Tied into an operations program, this can automatically generate switch lists.

Steve Davis makes extensive use of RFIDs on his KCS 3rd Sub layout.  He has some great videos on how they operate on his website: https://kcs3.webs.com/

After reading Dave's book several times, I am using method #2 as described in the book. It was straight forward and worked right off the bat. I did add pot's to my BD 20's to make adjustments easier. The added benefit of it all is that when questions arise I e-mail Dave, and he replies. 
Working on the switches now and a tie in to the CATS program.

I use method #2 on a mid-size layout with full-size CTC, using the same approach that Dave uses: Digitrax SE-8C signal drivers, NCE BD20 with NCE AIU for detection.  
If you desire CTC-like experience and dispatch then this is the way to go.  If you are only doing a single control point or interlocking, or solving a "real life" problem (like hidden staging) infrared is worth a look.

The signal block should indicate a track occupancy if any spur track switch is open within the block so a train cannot get a proceed signal into the block.
This is easy to do with current detectors by wiring a resistor shunt across the turnout diverging side.  
The other systems would need a line from the turnout control to the logic controlling signals in that block.
I started with a version of "Twin T" current detectors and have migrated to the open transformer loop (NCE BD-20) type.
There is a lot of variation in sensitivity adjustments, perhaps due to current leaking across the ballast if it has high mineral content.  (As do prototype RRs.)

railstiesballast Wrote:
-------------------------------------------------------
> The signal block should indicate a track occupancy
> if any spur track switch is open within the block
> so a train cannot get a proceed signal into the
> block.
> This is easy to do with current detectors by
> wiring a resistor shunt across the turnout
> diverging side.  
> The other systems would need a line from the
> turnout control to the logic controlling signals
> in that block.
> I started with a version of "Twin T" current
> detectors and have migrated to the open
> transformer loop (NCE BD-20) type.
> There is a lot of variation in sensitivity
> adjustments, perhaps due to current leaking across
> the ballast if it has high mineral content.  (As
> do prototype RRs.)

I have found that you can shunt the BD20 inputs to the same effect; I have two "hand throw" turnouts that are hard to reach and so they are controlled by "locally controlled" Rapido switch machines which also have relay contacts.  Wired those up to shunt the BD20's when thrown and the block shows occupied.

I'm planning on #2 and just starting. All of the layouts I recall that I've operated on (only about 7, so not a lot over the years), that detect, use #2. Seems to work well. 

I know the RFID thing is the newer option over the past few years and as mentioned, can be pricey. Guess it's just new and not a lot of information from modelers out there. Heck, I'm using 100% wifi and nothing wired for throttles. Didn't install any universal ports or anything else around the layout, just a few wifi devices. This is because I started building my layout in 2020, with 2020 technology.

Mike ONeill
Parker, CO

tmotor: before you make any decisions check out Signal Solutions Inc. The owner of this company is a prototype signal engineer and has designed some systems that are as prototypical as possible. I wish I would have found out about them before we went with Digitrax and JMRI.
Steve Kelly
California Southern MRRC
Norwalk,CA.

THANK YOU to all that replied.  :-D
 
>All detection can be a bit of a pain and I recommend experimenting first with your desired detection to get the "bugs" worked out early.  
> Jeff N
EXCELLENT advice. 
The first step is to create a compressed module with closely spaced detectors, to gain an understanding of how to install, setup, and maintain a system.
 
>BMRC has used No. 2 for 30 years.  One recommendation for this system would be to be sure each individual block detector has its own pot for sensitivity adjustment.  Some detectors are bunched together in groups of 4 or 8 detectors with only one pot for 4 detectors.  If some blocks are different in length in one group, it's hard to get all four set right to pick up a 10K ohm resistor.  For example, if a switch is treated as a block and switches and blocks that are long runs of track are on the same 4-block detector group, it's about impossible to get them all working with one sensitivity setting.   If you're starting from scratch with this in mind, the blocks could be carefully grouped with like-sized blocks.
>jdw3460
Great tip!  The track resistance will vary slightly with each block.  Being able to fine-tune it will help ensure the data from the detectors is reliable.
 
>Our model railroad club uses the infrared system by Azatrax and it works quite well.
>Bob Krieger
I like the idea that the rolling stock doesn’t need to have anything added to them in order for the detectors to work.  If a visitor brings their equipment to run, no problem.  I will give Azatrax a serious look.  http://www.azatrax.com/
 
>3 has some interesting applications beyond simply occupancy detection.  Each car has a unique RFID tag on it which can track the locations of each car and locomotive on the railroad.  Tied into an operations program, this can automatically generate switch lists.
Steve Davis makes extensive use of RFIDs on his KCS 3rd Sub layout.  He has some great videos on how they operate on his website: https://kcs3.webs.com/
>mpe383
Thanks for that link!  Steve has done some nice work, and taken the time to document it with videos is going the extra mile.  Being able to generate automatic switch lists would expedite setting up for an Operating Session.
 
>After reading Dave's book several times, I am using method #2 as described in the book. It was straight forward and worked right off the bat. I did add pot's to my BD 20's to make adjustments easier. The added benefit of it all is that when questions arise I e-mail Dave, and he replies. 
Working on the switches now and a tie in to the CATS program.
>JeffZCT
I appreciate the confirmation that Dave’s book is a winner, and the steps to a successful install are repeatable.  Pretty nice of Dave to respond to emails.  He is an example of someone trying to help the model railroad community.
 
>I use method #2 on a mid-size layout with full-size CTC, using the same approach that Dave uses: Digitrax SE-8C signal drivers, NCE BD20 with NCE AIU for detection.  
If you desire CTC-like experience and dispatch then this is the way to go.  If you are only doing a single control point or interlocking, or solving a "real life" problem (like hidden staging) infrared is worth a look.
>JUTower
Glad to know Dave’s approach is winner. 
I may use a hybrid approach to detectors, depending on the cost of the equipment.
 
>The signal block should indicate a track occupancy if any spur track switch is open within the block so a train cannot get a proceed signal into the block.
This is easy to do with current detectors by wiring a resistor shunt across the turnout diverging side.  
The other systems would need a line from the turnout control to the logic controlling signals in that block.
I started with a version of "Twin T" current detectors and have migrated to the open transformer loop (NCE BD-20) type.
There is a lot of variation in sensitivity adjustments, perhaps due to current leaking across the ballast if it has high mineral content.  (As do prototype RRs.)
>railstiesballast
Fine-tuning the detectors will be an important requirement.
 
>I have found that you can shunt the BD20 inputs to the same effect; I have two "hand throw" turnouts that are hard to reach and so they are controlled by "locally controlled" Rapido switch machines which also have relay contacts.  Wired those up to shunt the BD20's when thrown and the block shows occupied.
>JUTower
Thanks for the tip.
Turnout positions will be known, especially when first powering-up the layout.
 
>I'm planning on #2 and just starting. All of the layouts I recall that I've operated on (only about 7, so not a lot over the years), that detect, use #2. Seems to work well. 
I know the RFID thing is the newer option over the past few years and as mentioned, can be pricey. Guess it's just new and not a lot of information from modelers out there. Heck, I'm using 100% wifi and nothing wired for throttles. Didn't install any universal ports or anything else around the layout, just a few wifi devices. This is because I started building my layout in 2020, with 2020 technology.
>Mike ONeill
I’m trying to leverage whatever is the current technology, but trying not to get on the “bleeding edge” where the technology is so new it is buggy.
 
>tmotor: before you make any decisions check out Signal Solutions Inc. The owner of this company is a prototype signal engineer and has designed some systems that are as prototypical as possible. I wish I would have found out about them before we went with Digitrax and JMRI.
>Steve Kelly
I watched the 4 episodes of “Realistic Signaling for Modelers” where Kevin Rudko (founder of Signal Logic Systems) describes signals, detection, and demos his system on Trainmasters TV.  Kevin’s resume is impressive.  His system offers Infrared detectors that can fit between the ties of N-scale track.
https://signalogicsystems.com/news-and-updates/
 
 
Initially I was sold on RFID with a wireless connection to a hub, but now I’m not so sure.  There seem to be several players offering different systems.  I will need to research a bit more and narrow it down to a few contenders.  Then I will set up a test module for each detection system and try them out. 
 
Thanks again to all those that shared their valuable experience and suggestions.
Take care and God bless!
Dave



Edited 1 time(s). Last edit at 06/27/22 06:14 by tmotor.