Cheap train detection for hidden staging. Ya'll critique my plan.

[grandeman]

The idea here is to detect trains without going bankrupt. [:D] The requirements are for the system to indicate which tracks are occupied and if the train is an arrival or departure.

To do this, the plan calls for a magnetic reed switch to be located on each end of the every stub track in a yard. The switch on the entry end will control a green LED and the switch on the far end will control a red LED mounted on an indication panel. Each loco will be fitted with a magnet to trip the switches. Prior to an Ops session, staged (departing) trains will have their lead loco located over the entry switch of their respective track for a steady green indication. Arriving trains will trip the green switch momentarily (and again for each loco in the consist) to indicate to the engineer that he's entered a stub track and needs to prepare to stop. As the train reaches the end of the stub, the red LED will illuminate. If needed, the train will be backed up so that a steady red indication will be apparent on the indication board. A track with no LEDs lit would indicate an empty track suitable for an arriving train.

Here's the indication panel. Since I'm not nearly the perfectionist I used to be, it was put together free hand. It's not perfect, but should do the job nicely. The LED's obviously still need to be put in as do the terminal strips and resistor board on the back side. The white area at the right labeled Track/Loco is made of dry erase board so the cab number of each lead loco can be listed after trains are staged. The board is made of 1/4" Micarta, the stuff used to make aircraft control cable pulleys!


Here's a shot of a detection magnet installed on a P2K Geep. These are some SERIOUS magnets!


Reed switches, LEDs, resistors...


Here's a simple wiring print (please forgive my "paint" inadequacies). Each LED will draw power from Rail A, go thru a resistor installed on a PC board behind the indication panel, power the LED and return to a common Rail B connection that will be wired back to the DCS 100 command station. Since each track will have two LEDs, there will be two 4 conductor phone cables run to each 4 track yard. If neccessary, a second reed switch could easily be added in parallel with the first red "stop" LED to provide a longer indication so the engineer can stop and still have a lit LED for the spur. Of course, as previously stated, the train could be backed up a little too.


Ya'll have anything to add or see anything I'm missing???

[ereimer]

i don't know why , but i have the feeling this shouldn't be hooked up to track power to run it . it's going to take someone with electronics knowledge to figure out why i'm right . where's randy ? [:)]

this is just a hunch , i could very well be completely wrong . other than that it looks like a good plan

[NZRMac]

Eric I'm with ereimer, will DCC run an LED ok? LED's are dc, so you may need a diode in series with the LED, resistor.
Other than that cool idea.

Ken.

[Anonymous]

how does the reed switch work and how it interact with the magnet? I got no idea where to get the stuff though!

[NZRMac]

QUOTE: Originally posted by dannydd

how does the reed switch work and how it interact with the magnet? I got no idea where to get the stuff though!

The magnet closes the circuit in the reed switch when close enough.(put simply)

Ken.

[Anonymous]

LED's will light on AC power. You could wire in a diode, but an LED is a diode so it seems to me like overkill. The AC current switches polarity so fast that there is no visible blinking of the LED. Your only concern is voltage, so use proper resistors.

[Anonymous]

QUOTE: Originally posted by ereimer

i don't know why , but i have the feeling this shouldn't be hooked up to track power to run it . it's going to take someone with electronics knowledge to figure out why i'm right . where's randy ? [:)]

this is just a hunch , i could very well be completely wrong . other than that it looks like a good plan

Ereimer,
Strictly speaking you are correct. As an electronics engineer I would much prefer to see this powered separately. However, that would add to the cost, but in reality a simple DC output "wall wart" type plug pack can be quite cheap from an electronics surplus type store. However I would almost put money on the fact that everybody has at least one of these things lying around the house which is no longer used.

My reservation is that the reverse breakdown voltage of the LED may be less than the voltage coming from the DCS100. When a LED conducts it is said to be forward biased, and the voltage drop across the LED is the forward voltage. Usually somewhere from 1.6 to 3 volts depending on the colour and semiconductor material of the LED.

When the rail goes to the opposite polarity, as it does in DCC, then the LED will be reverse biased and will turn off. Thats not the problem. I dont have any data to knwo what the reverse breakdown voltage is, but here is the rub. If the voltage across the LED exceeds the reverse breakdown then the diode will again being to conduct - but not outputting any light this time. But as we have the resistor in series with the LED in the first place to liit the forward current the same will happen under reverse breakdown and the current wont be so high as to upset the DCC system.

So a separate DC supply would be best. But if the OP uses this design it may work but the LED life may be compromised, or they may be destroyed first time out! I am not sure until I get back to the office and dig out some data sheets.

John

[grandeman]

Guys, the LEDs light ok when powered across the rails (with a resistor in series). A separate power supply is entirely doable, but would require additional wiring. I'll be thinking on this... Any other comments or concerns? Thanks for the info so far!

[tsasala]

If wired to the rail power, it causes extra current that might be better used to power locos and sound. I think the "extra" cost and wiring is worth it isolate the detection system from the running trains.

You know the LEDs will only be on when the loco is very near the switch, right? Trains will be in the block with no indication. Hope that's OK.

0.02

-Tom

[nbrodar]

Good start. Isolate the detection and track power. You also might want to concider installing power cut-off switches to the staging tracks. That way, you won't have a bunch of decoders sitting around drawing power, and you can't accidentally run two trains out of staging at same time.

Nick

[Anonymous]

The amount of current to light a few LEDs is pretty small, so I am thinking that isn't a problem. Clearance for the magnets under some engines would be one thing I'd worry about. Also, remeber, this is only going to tell you when the loco is right over the magnet. If you miss it, you won't know if you've passes it or not. But the idea looks pretty good for its purpose.

[grandeman]

The thought strikes me that if a separate power supply were used, it could have a low voltage and eliminate the resistor board altogether...

[rrinker]

Agree with the above - power it with a seperate power supply, do NOT connect anything to the track like that.
If you MUST connect a regular LED to DCC track power directly (ie not through a decoder function output) you must put a standard diode in the reverse direction across the LED - DCC track voltage int he reverse direction will eventually ruin an LED as it does exceed the peak inverse voltage of most LEDs (PIV). Look at an ordinary 1N4000-series diode, they have PIV's of 50, 100, 200, even 400 volts and won't be damaged by the reverse current while protecting the LED.
Another potential problem is if that track feed is part of an autoreverse setup - if it reverses phase, the LEDs won't work! More reasont o use an independent power supply.
Check the magnets periodically, any loose metal filings and so forth scattered about the layout will quickly become attached to the bottom of all magnet-equipped locos.

--Randy

[SpaceMouse]

Something that has bothered me since I first read about this.

Does it bother anyone else that a magnet will be attached to a locomotive driven with a finely balanced elctromagnetic engine? I've never done it, but I can't help but think there will be consequenses.

[grandeman]

nbrodar, independant track power switches are something that I really hadn't considered. They would add much wiring/complexity though.

rrinker, I'll go with a separate power source. There is no autoreverse loop on the layout, so that wouldn't be an issue anyway.

Space, the locos seem to run fine with the detection magnets installed. No problems have been noted yet except one of the tracks has a screw in the benchwork under the roadbed. A slow moving loco takes a little added power to clear it. That is a lesson learned. I need to run the magnet equipped locos around the entire layout and check to make sure this won't be an issue anywhere else.

[SpaceMouse]

QUOTE: Originally posted by grande man


Space, the locos seem to run fine with the detection magnets installed. No problems have been noted yet except one of the tracks has a screw in the benchwork under the roadbed. A slow moving loco takes a little added power to clear it. That is a lesson learned. I need to run the magnet equipped locos around the entire layout and check to make sure this won't be an issue anywhere else.

Obviously, no one else see the problem so it may be a non issue. And I wasn't really seeing a melt-down of the warp core caused by irregularities in the flux capacitor.

I was seeing that over a period of time, an increased magnetic field on one the side of an electromagnetic motor would cause the motor to wear unevenly and cease to function, like is common on Mantua steamers.

[grandeman]

QUOTE: Originally posted by SpaceMouse
[
I was seeing that over a period of time, an increased magnetic field on one the side of an electromagnetic motor would cause the motor to wear unevenly and cease to function, like is common on Mantua steamers.

I'll just flip the motor over every now and then...[:D]

[joeyegarner]

Hey everyone here is a link on ebay for reed switches. http://cgi.ebay.com/Latching-Reed-Switch_W0QQitemZ6007419775QQcategoryZ19127QQtcZphotoQQcmdZViewItem

[markpierce]

Not being electronically inclined, this is all too complicated to me. I think a simpler and cheaper method is to use a camera(s) connected to a closed circuit TV.

[Anonymous]

I was wondering about the magnet under the motor, too, but since it looked like he was already running that way I figured it must be working. The screw slowing the train down is interesting...

[rrinker]

Shouldn't be a problem with the motor. Trying to pass over top kadee uncoupling magnets, on the other hand....

--Randy

[cwclark]

the problem is going to be getting the magnet on the locomotive and the reed switch perfectly lined for it to work....another drawback is that reed switches don'y always work like they should...it's a mechanical device and they are prone to hang up at the contact hinge... it may be better to just bite the bullet. spend the money and purchase the train detection circuits...then all you have to install is the TDC, a trickle voltage power source, and some resistors on your rolling stock wheels to set off the occupied block LED....chuck

[grandeman]

QUOTE: Originally posted by rrinker

Shouldn't be a problem with the motor. Trying to pass over top kadee uncoupling magnets, on the other hand....

--Randy

YIKES! I hadn't thought of that! I'll check it out tonight.

[rrinker]

If this is a hidden staging area - a simple IR beam-breaker type of thing would work just fine too, and require no alteration to the rolling stock. There are a bunch of suitable circuits on Rob Paisley's site, and there are commercial items such as the IRDOT.

--Randy

[ndbprr]

I plan to use a pair of infrared transistors. one is the emitter and the other is the receiver. When the receiver senses the source infrared it closes and can light an LED or bulb. the advantage is placing the two say an inch from the end of the tracks so when a train breaks the circuit the light goes out (or on with a hex inverter). That way it only takes the two transistors for all the tracks if you then back the offending trains back a little and stop breaking the plane of the light. I intend to use a small brass tube on the emitter to help focus the light better. I tried reed switches once and wasn't happy with them but it could have been me.

[Anonymous]

QUOTE: Originally posted by markpierce

Not being electronically inclined, this is all too complicated to me. I think a simpler and cheaper method is to use a camera(s) connected to a closed circuit TV.

I am up to date on technology enough th agree with this 100% but I am old enough that it struck me as funny when I first read it. A magnet and a few switches more complicated than a closed circuit TV system? When did that happen?

[railroadyoshi]

very interesting idea. I think Infrared may be easier to provide constant light that there is occupation, but then again you will get flicker between cars. I had the same issue as spacemouse, but you seems to have found no problem with that. Hope you find a way around the uncoupling magnets.
Best of luck, and would like to see the results[tup]

[ndbprr]

very interesting idea. I think Infrared may be easier to provide constant light that there is occupation, but then again you will get flicker between cars.

No you have the wrong conception. I am placing a beam of light like an electronic eye accross the tracks at a 90 degree angle. Anything on any track that breaks the beam will light the light so you bring a train into the staging yard until the light goes on or off and then back up a little to reset it. there would be no flicker.

[grandeman]

Randy, the Kadee magnets don't seem to present any special problems. The magnet equipped loco just glides right over them. Interesting that the benchwork screw seems to have more effect.

[rrinker]

Interesting - probably because of the way the Kadee magnets are polarized, down the long axis - so along the strip of magnet you only have half attracting and half repelling, which should balance out.

--Randy