Switching system for two tracks

Are the relays wired with a seal-in circuit, or are you depending upon the reed switch to keep the relay coil energized while the train is stopped?

Series string: one contact holds first relay on, second contact disconnects the other relay. Only one can be on at a time. Later I will try to post a copy of the circuit. The reed switches are only closed for a moment when the train passes over.

This is an interesting read on adding some basic components to prevent the relay contacts from sparking (which degrades the contacts):

https://www.ato.com/how-to-pre...rk-at-relay-contacts

Sorry the diagram is sideways. This is the first time I've tried to post something like this. The relays are 3pdt not the 4p shown. 4th pole n/c.

trainman129 posted:

...Our problem is after a few hours of operation the system stops working. You can hear the relays operate but they no longer hold themselves on. If I take the relays apart (they are 3pdt plug in units) and clean the contacts all is well for another few hours.

As I understand it, you're using pole 1 of each relay to ping-pong track power to the sidings.  Pole 2 and 3 are cross-connected to effect a latching relay triggered by the momentary reed-switch closure.

When you say "You can hear the relays operate but they no longer hold themselves on" does this mean they only click on for the duration of the reed-switch trigger... and then click off (do not latch)?

The first thought is something is going on with the pole 1 contacts that carry track power.  But if it's the latching behavior that is failing (poles 2 and 3), those are low-current, low-voltage contacts and hence curious that those are what's failing!  Can you confirm it's the pole 1 contacts that you have to clean?

Is this a recent phenomenon or has this been going on since day 1?   Has it been getting worse (you now have to clean the contacts every hour vs. every 2 hours last year or whatever).

Approximately how many relay operations per hour?

What is the 3PDT relay part #?  Approximately what is the current loading for each of the 2 trains?  What is your track power transformer?

Since these are "plug-in" units, have you actually swapped in new relays?  

----

As for an alternative, I'm curious why you didn't simply use one SPDT 12V latching relay.  The relay would always have one or the other siding powered as opposed to having the ability to disable power to both sidings.  The 2 reed-switches would then set or reset the latching relay.  The SPDT relay contacts of the relay would apply track power to one or the other siding.  These modules are available off-the-shelf with screw-terminals (no soldering) for about $10-20.  Example from Azatrax:

Or, if budget is an issue and you're the tinkering type, I believe you could do this for less than $3 out-of-pocket (free shipping from Asia) using a 4-channel 12V relay module from eBay.  In this OGR thread, I show how to use such a relay module (10 Amp contacts) to implement a similar alternating function.  That is, 2 of the 4 non-latching relays are cross-connected to implement the latching behavior similar to what you're doing.  The reed-switches would provide the momentary A and B triggers to toggle the latching SPDT relay that sends track power to the alternating siding.  You don't need the time-delay feature so all you need is the 4-channel relay module.  And you already have a source of 12V DC.

When you say "You can hear the relays operate but they no longer hold themselves on" does this mean they only click on for the duration of the reed-switch trigger... and then click off (do not latch)?

That's right. Relays operate but do not latch. You can see track power blink off-on as the train passes over a reed switch. I've been using fine emery cloth to clean all 3 poles.

Thanks for the ideas above. I'll have to give them a try. We only have access to the layouts on Tuesdays so it takes time to get much done. Work on ideas at home then take the project in on Tues. to try them out.

Jay

Went with the Azatrax unit. Seems to work well. Need to change another reed switch though. One track switch not changing properly.

An I.R. detector based approach would most likely have worked fine and be less troublesome.