View Full Version : track relay ideas with dcc

December 13th, 2011, 02:33 AM
Heres the deal, im designing my signal plans for my cajon pass layout and i have an idea but no dcc system to try it on. heres the question. has anyone ever hooked a relay to the rails while DCC is on so that the dc pics up the relay? reason i ask is i work in the signal dept for bnsf and im trying to try the real way of doing things to run my signal system so that the draw of a train in the block will drop the relay.


December 13th, 2011, 02:37 AM
Check out the DBD22 from TeamDigital. Will sense block occupancy and set a signal. Can be done directly, or as a sensor input to other DCC electronics

December 13th, 2011, 03:22 AM
Id prefer to find a relay that would pick and drop on its own with a couple extra contacts in it. i want to wire the whole signal system myself and take care of the relay logic with my one eiting.

December 13th, 2011, 09:05 AM
AFAIK, the 1:1 Relays are energized until a Short occurs through the presence of an axle. As current is flowing through the axle, the relay drops. I imagine it would be rather hard to do that with DCC, as we usually try to avoid shorts at all cost. So somehow you will need to sense the current flow instead of the short.

have fun

December 13th, 2011, 09:59 AM
Michael has it right. A current sensor arrangement is the easiest and cheapest way to emulate a track relay arrangement.

You will still need a track circuit as such, complete with insulated joints, though this will be what powers the loco. Essentially, you will create the opposite of what you (and I with 33 years in signalling) expect when a track circuit is shunted. I guess like the approach track circuit for a predictor crossing where the presence of a state (in it's case a change in inductance) operates the input.

December 13th, 2011, 01:10 PM
What im wanting to do will not create a short. yes the big boys create a short but since our locos dont short normally, it wone here. What i want to do is have a relay that connects to both rails and picks up when the track has power. now when a loco gets in the block or a string of cars with resistors on the wheels, the LOAD of the loco and/or cars will make the loss over the circuit to great and not keep the relay up. on end of the circuit will have battery and the other end, the relay as im direction of traffic on cajon pass.
You should understand what im saying GATS, this is JUST like a normal track circuit.

December 13th, 2011, 03:24 PM
What you are saying (if I understand it correctly) is that under unoccupied conditions, the track voltage will pull in the relay, but when there is a load on the track (loco, resistor wheelset, etc.) it will pull the track voltage down enough to cause the relay to drop out (i.e., the relay state is dependent on the track voltage). (Also when there is a track break due to derailment, rock slide, etc.). That won't work with DCC as track voltage is constant (and a form of AC to boot). That being the case, it is better to use a current operated detection system rather than a voltage operated one. It's really two sides of the same coin, as the voltage operated approach used in the prototype depends on the track voltage being pulled down as a result of current flow through the circuit, correct?

Mike Sheridan
December 13th, 2011, 04:00 PM
The OP's idea (one end of the circuit will have battery and the other end, the relay as in a prototype system) is doable in principle, but it would be very hard. You'd need to provide filters (chokes and capacitors) to keep the 'AC' DCC power separate from the DC track circuit, and reliably shunt the track circuit even if only a single or few axles were in the block. It would need a lot of work on the locos and rolling stock. (Even the prototype doesn't have to face this amount of mixing as on electrified lines the power feed is on a separate rail or catenary, even though the ground side is common.)

As others have said, current detection makes more sense - we don't have such a need for failsafe systems ... eg. we don't generally have much problem with broken rails.

Of course, you could dump DCC in favour of battery operation, which would completely change the situation ... but that's a whole other can of worms.
(Fuse lit I now retire to a safe distance :) )

jagged ben
December 13th, 2011, 08:08 PM
If the wish is to use actual relays (out of hobby interest), why not put the relay in series with the locomotive? That is, the locomotive completes the coil circuit when it enters the block. This is a assuming a 12-16V AC relay will activate on the DCC waveform, but probably that will not be a problem. Also assumes that the draw of the relay will not adversely affect the performance of the locomotive, which would probably require some testing. I don't know what scale we're dealing with here. If N scale, maybe setting DCC on a voltage for a larger scale would work. If a large scale, I'm not sure it will work.

A big problem I see is that of matching any use of a relay a to the use of resistor wheelsets. Electronic detection systems that detect any amount of current down to a tiny amount have no problem with this. A relay requires more current than I think you'd want flowing over a resistor wheelset. Too much current through many resistor wheelsets will overload the DCC system.

Alternatively, if the wish is to use relays out of hobby interest, using a current detection system to trip the relays might be the best approach.

December 13th, 2011, 10:17 PM
The chokes wouldnt be hard as i deal with that on a daily basis in the signal dept. I bet theres relays out there that i can use. Track voltage is usually around the 1v DC on a DCC system correct? Its time to play and figure this out!


jagged ben
December 14th, 2011, 03:57 AM
Track voltage is usually around the 1v DC on a DCC system correct?

I think you made a typo. It's from 12-16V usually. The setting usually depends on the scale.

December 14th, 2011, 05:38 AM
yeah i just saw i forgot a 2. Think ill experiment with a 12v automotive relay first.

December 14th, 2011, 03:48 PM
You would need a pretty sensitive relay. The 75-150 ma drawn by a loco would be no problem, but a 10K resistive wheelset would only draw just over 1 ma. On top of that, the relay coil in series with the track power would seriously disrupt the high frequency (100 KHz) DCC data+power waveform. If you in fact want an actual relay to emulate the prototype, I would use current detectors to activate the relay. That way everybody is happy (I think).

Mike Sheridan
December 14th, 2011, 05:13 PM
Actually, adding to my previous post, the 'AC' DCC signal (I think it is about 8kHz, rather than 100) can have a DC component in the form of an offset (the loco 0 system). I expect that will probably foul up a DC track circuit.

December 14th, 2011, 06:43 PM
Look at it as the loco and relay draw power in parallel from the power source. Ill add a capacitor(shunt in the real RR world) to stop the AC and give this a shot. if i get it to work, ill build everyone cheaper signal systems!

December 14th, 2011, 07:24 PM
I was a Signal Maintainer too. I worked for NYC Transit for 31 years. I have successfully put a couple of signals on several of my T-Trak modules. The track circuit consists of an NCE BD20 current detector controlling a 12 volt DPDT dip relay. They do work the opposite of the prototype track circuit, (relay is de-energized with no train in the block) but the signals work fine. I use a separate 12 volt power source for the detectors and other relays controlling the signals. You shouldn't load up the DCC system with stuff that wasn't designed to work with it. My signals will work with either DCC or DC. The relays are cheap and don't use too much current. If you need more contacts, use another relay as a repeater. In the photo the bottom left relay is the track relay, the bottom right relay is a repeater of the track relay, the upper left relay is the home relay and the right relay is the distant relay. I've had my modules in several shows and they've drawn quite a bit of interest with the working signals.