Wiring Stub Terminal Passenger Station

[GaryHinshaw]

Instead of auto reverse can you use power (phase) routing from the wye legs? 
Jason

That could probably work Jason, but I haven't thought through how to wire it.

I did a little reading up on the DCC Specialties PSX-AR auto-reverse boards and it appears that they have a feature called "double reverse" which is exactly tailored to the situation above.  The tips manual for the board, http://tonystrains.com/pdf/dcc_specialties/psx_hints.pdf, notes the following:

PSX-AR “Double Reverse”
On some layouts there are situations that require two PSX-AR’s be connected back to back. This can cause a condition where both reversers try to fix the short by reversing the lines. They both keep trying to fix the problem and continue to flip back and forth, causing a delay. The newer PSX-AR’s have a CV that can be set to delay the action of the reverser. Only one of the two reversers needs to have this CV set to on. This allows the other reverser to change status during the short delay and resolve the conflict.

In other words you can have a fast and slow board working together.  So if you are willing to buy two new boards, this would give you complete operational freedom.

[wcfn100]

That could probably work Jason, but I haven't thought through how to wire it.

Basically, if you throw all four points of the crossover when changing routes there's never really a 'short' (okay there is, but no closed routes that create one only that the staging track could be out of phase which can be controled by the whatever throws the crossover).



Jason

[RBrodzinsky]

Oh, wow. That is a great addition to the PSX-AR. 

I am still not certain it will work reliably, but am willing to be convinced.  Using the prior diagram, lets say a train is pulling out of the right hand tracks, and crossing over to the left hand branch of the wye to head "west". If the right-hand reverser is the "fast" one, and the two sets of station tracks are opposite polarity, as the train crosses over the right hand side reverses polarity to match the left side. Now, as it continues out of the station area, what if the mainline is of different polarity of the, now, unified station tracks. We're crossing through a reversing section that cannot contain the whole train (and, per the original post, these trains are long and lighted).  Does the left hand reverser flip when the short on the wye occurs, and the right hand flip as well?  Am I overthinking this?

[GaryHinshaw]

Good point - let's make a truth table to evaluate the possibilities.    Here is the same diagram as above with some nomenclature added:



In your example the train starts in the fast (F) green block (time step 1), proceeds across the crossover into the slow (S) blue block (time step 2), then on into the west leg of the wye (time step 3), which I've arbitrarily set to fixed positive polarity.   We assume the train spans all three blocks at this step.  The truth table for each time step looks like this:

Time step		FSW		FSW		FSW		FSW
1		+++		+−+		−++		−−+
2		+++		−−+		+++		−−+
3		+++		−++		+++		−++
3'		+++		+++		+++		+++

Here, time step 3' is the state of the system after both blocks adjust to accomodate any shorts.   The interesting column is the last one: both blocks are negative and the west wye leg is positive.  In step 2 nothing happens; in step 3 the slow block flips because it is opposite the wye and there is a short in it when the train enters the wye; in step 3', the fast block flips because it is opposite the slow block, and it will flip before the slow one does again.  So I think this direction of travel is fine.  Now consider the opposite case where a train starts in blue and heads east:

Time step		SFE		SFE		SFE		SFE
1		++−		+−−		−+−		−−−
2		++−		++−		−−−		−−−
3		+−−		+−−		−−−		−−−
3'		++−		++−		−−−		−−−

Here the first two columns give you problems.  If the spanned block is the fast one and it is bordered by blocks of opposite polarity that are fixed or slow (effectively fixed), there is no solution that leaves all 3 segments with the same polarity.  Perhaps power routing is a better choice for this configuration (or the original single-block reverser solution with the operational restriction).

BTW, there are other configurations where a fast and slow reverser are desirable, just not in this spanned-block example.

[wcfn100]

I think you guys are overthinking all of this.  There's only two things that can happen coming into the yard.  Either you go straight or you cross over. All you have to do is match the yard to which leg of the wye the train is coming from which you don't even need an auto reverser for.



Yard 'A' will always match what turnout '4' (and '2') is doing and yard 'B' will match what turnout '1' (and '3') is doing.  When the cross over switches, the yards automatically switch to match the other leg and then again is you switch the cross over back.  You can also see that you can bring in two trains if you're using the straight routes because at that point they are two separate yards.

Unless I'm missing something?

Jason

[RBrodzinsky]

Jason, what you are missing is that the leads to the yard are coming off a wye. Without some reverser in play, the two outer most rails are the same polarity, yet are opposite rails. We still have to auto reverse at some point to re-enter the main.

[wcfn100]

Jason, what you are missing is that the leads to the yard are coming off a wye. Without some reverser in play, the two outer most rails are the same polarity, yet are opposite rails. We still have to auto reverse at some point to re-enter the main.

I'm not missing that part.  The two yards 'polarity' get controlled by the cross over.  When a train is leaving the yard whatever route you pick will be how that yard gets aligned. Basically the same thing that would happen if you used an auto reverser except you do it as part of the cross over sequence.

A longer way to look at it:

If a train comes in on wye leg 'y' into yard 'A' the crossover is aligned straight.  If that same train wants to go out the other leg 'z', the crossover gets switched and the phase of yard 'A' gets switched to match leg 'z'.

I'm sure could be missing something, but because of the double crossover, this isn't a true wye.  If the 4 switches get thrown at the same time, there's never a closed off route e.g. a train coming into the yard always has an open route.  You just need to get that route on the same polarity.

It's really no different than controlling a DC wye where you could control the tail leg with a toggle that would throw the switch and change the polarity of the track.  The only difference is with DC you'd have to remember to not change the direction of the locomotive because that was done with the toggle.

Jason

[wcfn100]

Maybe this picture will show it better.



The yards only need to know what leg of the wye they are routed to (which there is only two options) which is controlled by the double cross over.  No auto anything needed just wire it all with the turnout contols, but you could use one auto reverser for each yard if you didn't want to deal with the wiring.


Jason

[jagged ben]

I think Jason has it right.

Another relevant detail:  The Kato double crossover is entirely insulated in the middle.  Only the outside rails are continuous.  In order to make this work Flagler just needs to put insulating joiners on the crossover's outside rails at the end away from the station.  Then he feeds the station as outlined in Jason's posts.

One complication is that he'll need to come up with a relay method that marries the required X-block type relays to Kato turnout control.  Either that or some kind of physical interlock that only allows him to throw the crossover if the track power is swapped.

[wcfn100]

Either that or some kind of physical interlock that only allows him to throw the crossover if the track power is swapped.

I'm imagining that reversed.  The yards follow whatever the turnout does.  Whenever the crossover is thrown, the yards just automatically get reversed because they have to be.

Jason

[jagged ben]

On second thought...

Flagler will probably have to cut gaps in the middle of the outside rails of the crossover in order to make this work.  Jason's diagram more or less shows why, in the figure on the right.

[jagged ben]

  No auto anything needed just wire it all with the turnout contols, but you could use one auto reverser for each yard if you didn't want to deal with the wiring.

You could use one auto reverser total to serve both sides of the yard.  Just wire the inside rails to one rail of the reverser, and the outside rails to the other.

I'm imagining that reversed.  The yards follow whatever the turnout does.  Whenever the crossover is thrown, the yards just automatically get reversed because they have to be.

Yes.  We both could have just said "or vice versa." 

[jagged ben]

To sum up:  the simplest solution would be to take a Dremel and cut gaps in the middle of the outside rails of the crossover.  Then wire the inside rails of both sides of the station to one 'rail' of an autoreverser, and the outside rails to the other.

That's it.  Done.  No insulating joiners anywhere.

No operational restrictions other than not physically running trains into each other  .

[wcfn100]

You could use one auto reverser total to serve both sides of the yard. 

Yes you could.  *

edit:  You would have to make sure the wye legs were always 'opposite' polarity (as they are in all of the drawings) which could get over looked at some point.  I don't think you need any reversers, but if you wanted to, I would go with one for each to be safe.

Jason

[GaryHinshaw]

Yes, that's a beautifully simple solution.  I drew it out to convince myself that I was thinking about it correctly. Assuming that the crossing is fully insulated, it should indeed work just fine with a single reverser above the gaps (as long as the straight routes are gapped):



Nice.