I meant to say Azatrax. I am going to try the Atlas switch method first.
Thanks to all.
Dave
Gents:
I am researching an Alastranz turnback relay with an infrared sensor. I would prefer an automated process but it starting to get complicated.
Lastspikemike ATLANTIC CENTRAL With DC, if you plan to reverse the loop polarity, you need to stop the train, reverse the loop polarity, reverse the throtte polarity, then restart the train. So yes, you can create a semi automatic situation where the turnout position controls the entry/exit polarity, but at some point while the train is in the loop, it has to stop and let the turnout be thrown and the throttle direction reversed for the other direction on the mainline. Otherwise, when you throw the turnout, the train will change direction. Sheldon This turns out to be incorrect. Took me a little while to figure out. The key aspect is to provide sections of main line powered track inside the actual geometric reversing loop long enough to hold the entire locomotive consist. This requires the reversing section to be shorter than might otherwise fit within a loop created by one turnout. For our back to back loops which basically have a long double isolated X section these main line powered sections are very long so there's practically no limit to train length, but for our one single turnout loop we sacrifice a little train length capacity to allow for these main line powered sections. A section of track long enough to contain the locomotive consist just before the exit to the loop is all that is required. For a two way loop you need the same length of main line powered track at the other end of the reversing section. Each one of these sections forms one of the two routes out of the turnout. The turnout also sets polarity in the reversing section to match train direction entering the loop. The train proceeds into the loop and passes onto the reversing section proper. When the whole powered part of the train (which is the whole thing if you use metal wheels) is within the reversing section, I.e. has cleared the entry section of track powered by the mainline, you switch polarity of the main line, as the locomotive consist clears the reversing loop onto the second section of main line included inside the loop only then do you throw the turnout which changes the polarity in the reversing section. The locomotives don't change direction and the trailing car wheels don't short out the isolating gaps. It's all in the timing of the polarity changes. Should work fine for DCC in fact. I think for DCC the mainline sections aren't needed but I'm still thinking about that. DCC allows you to flip " polarity" inside the reversing loop whenever you like. The steel wheel effects from the trailing cars might be an issue though. This all results from the Möbius Strip aspect of a reversing loop where one rail is always a continuation of the opposite rail. A reversing loop is made from one rail. Magical in a way. You never need to stop the train.
ATLANTIC CENTRAL With DC, if you plan to reverse the loop polarity, you need to stop the train, reverse the loop polarity, reverse the throtte polarity, then restart the train. So yes, you can create a semi automatic situation where the turnout position controls the entry/exit polarity, but at some point while the train is in the loop, it has to stop and let the turnout be thrown and the throttle direction reversed for the other direction on the mainline. Otherwise, when you throw the turnout, the train will change direction. Sheldon
With DC, if you plan to reverse the loop polarity, you need to stop the train, reverse the loop polarity, reverse the throtte polarity, then restart the train.
So yes, you can create a semi automatic situation where the turnout position controls the entry/exit polarity, but at some point while the train is in the loop, it has to stop and let the turnout be thrown and the throttle direction reversed for the other direction on the mainline.
Otherwise, when you throw the turnout, the train will change direction.
Sheldon
This turns out to be incorrect. Took me a little while to figure out.
The key aspect is to provide sections of main line powered track inside the actual geometric reversing loop long enough to hold the entire locomotive consist. This requires the reversing section to be shorter than might otherwise fit within a loop created by one turnout. For our back to back loops which basically have a long double isolated X section these main line powered sections are very long so there's practically no limit to train length, but for our one single turnout loop we sacrifice a little train length capacity to allow for these main line powered sections.
A section of track long enough to contain the locomotive consist just before the exit to the loop is all that is required. For a two way loop you need the same length of main line powered track at the other end of the reversing section. Each one of these sections forms one of the two routes out of the turnout.
The turnout also sets polarity in the reversing section to match train direction entering the loop. The train proceeds into the loop and passes onto the reversing section proper. When the whole powered part of the train (which is the whole thing if you use metal wheels) is within the reversing section, I.e. has cleared the entry section of track powered by the mainline, you switch polarity of the main line, as the locomotive consist clears the reversing loop onto the second section of main line included inside the loop only then do you throw the turnout which changes the polarity in the reversing section. The locomotives don't change direction and the trailing car wheels don't short out the isolating gaps.
It's all in the timing of the polarity changes. Should work fine for DCC in fact. I think for DCC the mainline sections aren't needed but I'm still thinking about that. DCC allows you to flip " polarity" inside the reversing loop whenever you like. The steel wheel effects from the trailing cars might be an issue though.
This all results from the Möbius Strip aspect of a reversing loop where one rail is always a continuation of the opposite rail. A reversing loop is made from one rail. Magical in a way.
You never need to stop the train.
That is really the same thing with gaps in different places. You are still reversing the main and throwing the turnout in any case. And you failed to explain that little wrinkle of the mainline polarity extending into the loop to the OP, I bet that was not on his radar.
You are simply controlling the polarity of the reversing section with the turnout position ahead of the mainline reversing switch. Nothing new there. You then of course need a throttle location reverse switch to back up inside the reversing section.
And this is all great if you can see the train easily to know it has moved from the reverse section to the mainline fed section. Many great layouts I have operated on have hidden reverse loops and they easily work fine with two separate switches, using minimal detection or train locating methods.
For all that "timing" you are suggesting one might as well do it the conventional way and simply have separate reversing switches for the loops.
Set them correctly to enter the loop, reverse the main while the train is in the loop, throw the turnout and exit the loop, all easily done in the time it takes and normal train to travel around any usable sized reverse loop.
But since I don't have time to draw a bunch of diagrams today, or this week, once again I will leave you to it.
And, for what it is worth, my layout is full of X sections. They are used at every mainline interlocking to eliminate unnecessary blocks.
A simple center-off, DPDT toggle switch does the trick nicely, changing the polarity in the loop just as Mike described.
Who makes a good relay for analog train operation?
Regards,
Dave K.