Train Detection

AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
After 25 years, I am finally installing signals to my railway and want them to change aspect with the passage of trains. There are various ways of doing this but why doesn't anyone use track power?

I was thinking of having a short dead section of rail 6 mm or so. This short section or rail would have a wire to a relay coil and the return to the main rail on the opposite side. As a loco passes over the dead section it applies voltage through the loco chassis and will thus energise the relay. It would be weather-proof and discrete, no moving parts and maintenance free (no more than track cleaning).
The size of our rails would make it fairly easy to cut out and insert a small piece using epoxy or could use isolating track joiners.

Has anyone thought of this or have any suggestions?

Alan
 
Gavin Sowry

Gavin Sowry

Garden Railroader and Raconteur
27 Oct 2009
5,985
83
65
Hutt Valley, NZ
I wish you well in this project.

Best advice I can offer, is study the prototype, then study it again. Avoid any sort of system that works on a 'time' principle... seen too many of them on layouts that put up a green light after a period, signalling a train into an occupied block. All occupied blocks should be protected with red lights. Best practice is red light 'latches' when a block gets occupied, and stays latched at red until the signalman 'clears' the block (which he can't do if it is still occupied).

Have I lost you yet ?

Just gave some thought to what you suggested.... remember, every metal wheel that passes over your 'joint' will have the effect of energising the relay. Mind you, that could be helpfull if you were to consideraxle counter technology.
 
dunnyrail

dunnyrail

DOGS, Garden Railways, Steam Trains, Jive Dancing,
25 Oct 2009
14,442
197
71
St.Neots Cambridgeshire UK
After 25 years, I am finally installing signals to my railway and want them to change aspect with the passage of trains. There are various ways of doing this but why doesn't anyone use track power?

I was thinking of having a short dead section of rail 6 mm or so. This short section or rail would have a wire to a relay coil and the return to the main rail on the opposite side. As a loco passes over the dead section it applies voltage through the loco chassis and will thus energise the relay. It would be weather-proof and discrete, no moving parts and maintenance free (no more than track cleaning).
The size of our rails would make it fairly easy to cut out and insert a small piece using epoxy or could use isolating track joiners.

Has anyone thought of this or have any suggestions?

Alan
It ought to work but I would suggest the dead section be held in place with a couple of Massoth Isolating Clamps. Will provide all the screws that you need to attach wire rather than Soldering. You will also need a Latching Relay so that once it has thrown the Relay it will not be ‘chattering’ as every wheel breaks the Isolation. However all Metal Wheels ought to give the option of more chances for the Relay to actually throw.

Another way is to have a Micro Switch placed near the Rail so that Wheel Flanges depress the Arm. Again a Latching Relay is required. A second Micro Switch Can then be used to return the section to ‘normal’. We used this method on the Ruschbahn to show when Trains were lurking in Hidden Sidings.
 
  • Like
Reactions: Administrator
PhilP

PhilP

G Scale, 7/8th's, Electronics
5 Jun 2013
22,501
169
Tamworth, Staffs.
You do have to consider stock with insulated wheels..

So a 'treadle' (prototypical-speak) microswitch would work.. Two, and the order of firing gives direction, as well as occupancy. - Just to get your creative juices flowing.. :)
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
I am a train watcher so I like to watch the train pass and anticipate the next train. Signals that change green to red with the passage of a train is my objective. I'm not looking at using the track power to trigger positive block control that stops the trains. For me, the train passes the detector and changes the signal to red. After a delay the signal reverts to green, purely for my amusement.

Alan
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
I am familiar with industrial delay timers but they are big and expensive. I am now looking at a miniature PCB relay that triggers a 555 timer circuit. These timers can be used in different ways, such as astable, mono-stable and bi-stable. If I sound like an electronics expert, I am certainly not. Searching online for timer brings up this 555 device and the info is all there.
For me the 555 timer in mono-stable mode will change the signal to red and the circuit times down until it changes back to green without any further input. While it is timing down, further triggers have no effect, it continues to time down from the first trigger, thus it would ignore the wheels of long trains. If you wanted a more positive action, we could use the timer in bi-stable mode. This means that a train would trigger the timer when the train enters a block. The signal would remain red until the train operated another sensor as it exits the block to revert the signal to green. That would be a kind of block occupancy without a positive override of an approaching train and perhaps could be expanded to actually stop a second approaching train?
I hadn't thought about relay chatter but perhaps a capacitor across the relay coil would calm it down so that a dead wheel bridging the gap doesn't trigger the relay. My original thought was that the loco circuit and it's pick-ups would apply power to the dead track insert.

Any electronics people out there?

Also, any electronics people, would there be an alternative to relays ?

Alan
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
It ought to work but I would suggest the dead section be held in place with a couple of Massoth Isolating Clamps. Will provide all the screws that you need to attach wire rather than Soldering. You will also need a Latching Relay so that once it has thrown the Relay it will not be ‘chattering’ as every wheel breaks the Isolation. However all Metal Wheels ought to give the option of more chances for the Relay to actually throw.

Another way is to have a Micro Switch placed near the Rail so that Wheel Flanges depress the Arm. Again a Latching Relay is required. A second Micro Switch Can then be used to return the section to ‘normal’. We used this method on the Ruschbahn to show when Trains were lurking in Hidden Sidings.
Like the idea of Massoth insulated rail joiners, nice and easy!

Don't like the idea of micro switches in the garden, I had considered them but they would be an eyesore, difficult to mount & align, not damage and weatherproof. That is why I wondered why we couldn't use the track power. As in my previous post, maybe electronics could be a solution ?

Alan
 
a98087

a98087

Registered
8 Nov 2009
1,512
13
31
Wiltshire
How about stick a magnet under the loco or wagon, this can then make a reed switch glued to a sleeper, or sat Between the rails?

This can the new drive a latch relay or timer, that’s operates a signal, it’s cheap as the rare earth magnets and reed swirches needed are dirt cheap,

ANd the timer can be whatever you find that suits?

Dan
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
How about stick a magnet under the loco or wagon, this can then make a reed switch glued to a sleeper, or sat Between the rails?

This can the new drive a latch relay or timer, that’s operates a signal, it’s cheap as the rare earth magnets and reed swirches needed are dirt cheap,

ANd the timer can be whatever you find that suits?

Dan
I'm sure that reed switches work. I am reluctant to use them because all the loco's would need a magnet fixed to their underside and when i have visiting loco's the signals would not work. Also I'm not sure of their longevity and reliability.
I am reasonably aware of the various methods of train detection but wanted to discuss the possibility of using track power to a relay and if anyone has tried it before.
The advantage of a short dead section of rail, 1/4 inch long is no moving parts, no contacts to fail, weather-proof, works with any loco and is invisible apart from the wire to the signal control.

Ordered miniature relay, diodes and capacitors, I'm sure this will work.

Alan
 
PhilP

PhilP

G Scale, 7/8th's, Electronics
5 Jun 2013
22,501
169
Tamworth, Staffs.
It will work..

You will need to arrange for the capacitor to be charged, to hold the relay.. You might find it better to use a 555 timer circuit.. It would take many sets of wheels going over, to charge the capacitor enough to hold the relay for any appreciable time.
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
It will work..

You will need to arrange for the capacitor to be charged, to hold the relay.. You might find it better to use a 555 timer circuit.. It would take many sets of wheels going over, to charge the capacitor enough to hold the relay for any appreciable time.
Yes I like the idea of the 555 timer, it will switch a pair of led's from green and red for a set time, when used in mono-stable mode. With the relay, I shall use a miniature pcb relay, but it doesn't have to hold on. The relay will just give a quick pulse of 0 volts to the input of the 555 timer and the 555 timer holds it's output high until it times up. The relay would be a simple switch to apply the 0 volts to the 555 timer and I'm only using a relay to separate the track voltage from the low voltage dc that is required by the 555 timer.
But you still need a timer, or an 'unlatching' option, to clear the signal..
That's the beauty of the 555 timer, it will not require any further inputs, in mono-stable mode it is simply a timer and when it times up it will change back to green.

The 555 timer can be also be set up in bi-stable mode and then it works like a latching relay. One input to set the output high and a separate input to reset the output low. I am not using the 555 timer in bi-stable mode yet but I can see an application for it on my return loop to return loop line, but that will be a future project.

There are probably many ways to control the signals but I wanted to try the 'dead' rail method and so far it does look viable. Now I'm back to the shed to make a dead section in my test track. I'll let you know later.

Alan
 
  • Like
Reactions: Paul M
PhilP

PhilP

G Scale, 7/8th's, Electronics
5 Jun 2013
22,501
169
Tamworth, Staffs.
Have fun! :):clap::clap:
 
P

Paul M

Registered
25 Oct 2016
2,670
87
56
Royston
I've been following this thread with interest, it just shows how much knowledge there is on this forum.
 
Greg Elmassian

Greg Elmassian

Registered
8 Mar 2014
2,576
103
San Diego
www.elmassian.com
if the opto has two LEDs as shown, yhou can eliminate the full wave bridge... but with just a resistor then you might not trigger at low voltages, and the transfer ratio may not be high enough.

So, maybe put a zener in the circuit feeding the LEDs and tune the "question mark resistor" for 50 milliamps...

Greg
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
An idea for a possible solution?.......

I've left the value for the resistor as a ?, assume track input (DCC) at around 22V and as Greg just loves calculations and has some interest in optocouplers.
 
AlanL

AlanL

Registered
1 Mar 2016
34
1
Northampton
Thanks John, I wondered if there was a 'solid-state' solution avoiding relays. This is a quantum leap above my self-taught knowledge of electronics. I've got loads of questions about your circuit, primarily will it change a signal to red and then time down to revert to green?

Alan