no such animal...

Depends

What really needs to be asked is when you want to reverse polarity, like the Prof says you can't sense the short the way DCC does and automatically reverse the polarity.

However you can sense when a locomotive has reached a certain point by an isolated track section or through the use of photocells (other methods as well) and reverse the polarity of the mainline automatically to probably accomplish the goal.

If you want to build something I would look at Rob Paisley's web site, I think you could adapt his autoreverse circuits for reverse loop operation. http://home.cogeco.ca/~rpaisley4/CircuitIndex.html#63

For a commercial solution I would look at Circuitrons' offerings, their AR-1 should do nicely.  Their website is at: http://circuitron.com/

They don't have their products online, but you can download their pdf catalog from the website, which includes how to wire reverse loops, check on PDF pages: 34 and 42.

These ideas reverse the polarity of the mainline automatically while the train is in the reverse loop, but they can not handle the situation of entering the reverse loop, reaching the trigger point and then backing out.  Once you have passed the sensor point in the loop you have to complete running through the loop.  The direction you enter and leave the loop does not matter as the mainline will have the polarity reversed while in the loop.

This is not Automatic...

... however id does work fine for reversing loops and reversing cutoffs. I have used it on several layouts from N to G.

The track power in rectified by the diode bridge so the locomotive will travel one way through the loop. The controllers reversing switch can be thrown without stopping the loco while it is in the reversing loop. If the loop is only used one way you do not need the optional reversing switch. The bridge rectifier needs to have peak inverse voltage (PIV) of at least 30V and a peak current at least twice the current draw of your highest drawing loco or MUed locos. For HO I usually use a 50V 6 Amp rectifier.

There is a voltage drop of about 1.4V through the bridge rectifier so your loco will slow a bit but as you are probably on sharp radius in the loop I don't find this a problem.

The note below should actually read " From Last Block Before Reversing Loop Turnout" !

Great Minds...

Dear Graeme,

Google "Dynamite Canyon HOn30",

http://www.zelmeroz.com/album_model/members/klyzlr/DynamiteCanyon.pdf

almost the same solution, except with an additional set of normal silicon diodes as a protection "stop block" which will hold the train until the "Mainline" throttle polarity is set to "receive the train" back from the loop. (check the YT clip for live-fire example).

Add the "no moving parts" one-way turnouts and it is about as "automatic" as one could hope for...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

The Dynamite Canyon Tramway way- No moving parts +Timer

G'day Idahomoose,

Professor Klyzlr beat me with the first part of my reply about the Dynamite Canyon Tramway.  

Now to "automate" this - First some assumptions for stuff that isn't clear in the original post:

• As the Professor​ said there is no "automatic" off-the-shelf reverser that does not depend on other electronic stuff such as detectors and relays to change the polarity of the tracks in all or part of the balloon loop and the turnout controlling access to and from the balloon loop.
• Does it matter which leg of the balloon loop is used? - is "always in" on either the straight or diverge leg, must different legs be used alternately, or it doesn't matter?
• Is the operation a simple shuttle between 2 balloon loops or does each balloon loop have to be controlled individually with subsequent manually-controlled routing to other tracks? 

Assuming a simple shuttle between two balloon loops:

1. For a run between 2 balloon loops under DC control for a simple shuttle operation, if you are using the same throttle and a given locomotive and power setting then it may be possible to simulate "fully automatic operation" using diodes as laid out on the Dynamite Canyon Tramway It should take the train the close to the same time to run between the two balloon loops.  You can use a simple adjustable timer to drive a latching relay that reverses the polarity of the mainline. This will automatically reverse the feed direction into the blocking sections of both balloon loops  
2. if you change locomotives or power settings, then the timer may have to be adjusted to suit the new trip duration.
3. If you get the timing too short, then the loco won't make it into the balloon loop and the train will back-up on the main.This is probably not a desirable outcome.
4. If you get the timing too long, then the train will come to a stand in the balloon loop at the blocking section until the timer ticks down and triggers the relay to reverse the polarity. Only then will the loco haul your train out of the balloon loop and back onto the main. This outcome may not be a show stopper but your train will not be in continuous motion.
5. If you get your timing "just right"  ie the loco has made it into the balloon loop, but has not reached the end of the blocking section, then operation will appear to be "automatic" in then when the polarity changes, the direction of rotation of the motor does not change, and the motion of the train appears to be continuous. 
6. Something like this  may be a suitable timer than can act as a timed Flip/Flop to control the polarity change to the main line.
7. If you don't want to permanently fix your point blades as in the Dynamite Canyon example, you can still use the Dynamite Canyon diode set-up on each balloon loop with point motors to control the route. If you can find a 4-Pole/Double Throw 12 Volt relay like this one then with a timer flip/flopping the 4PDT relay, 2 poles control track polarity to the main line and the other 2 poles control the point motors. These are wired so that when the polarity changes when the timer fires on one balloon loop with the train on it, the route is set for diverge for the train to come out of the balloon loop, but the other balloon loop is set simultaneously to "straight in". This means that your balloon loops are are always set to receive the next train when your train is set to leave the other balloon loop. 
8. If you have other tracks and other routes for your train, then the timer idea cannot work for you as an automatic control system because the times taken to get from one the balloon loop to the other will vary too widely and either #3 or #4 above wil occur.  

Thanks

Thanks for the info. That gives me a few things to think about.

Circuitrons' AR-1

You might just have it! In looking at Circuitron's web page it looks like the AR-1CC might do it. The layout is a hybrid Atlas Apex and Hypotenuse and the photo cells might be the answer. Thanks 

So by now with the different

So by now with the different ideas and solutions, you probably realize that to give an accurate solution what you want to do has to be a more detailed.

I think the AR-1 plain is all that you need, as the CC version just adds two more photocells, which I don't think you'll need with standard DC.  I'm thinking you could put one AR-1 controlling each reverse loop.  You'll notice in their diagrams they have the insulating gaps into the loop and the sensors before the insulated track section so the loop polarity gets reversed to match the direction of travel.

There was such an animal.

In the early 1980's I tried auto reversing, on DC rails, with a circuit using reed switches triggered by boxcars ,each with a small magnet fixed to the bottom of the boxcar, and relays that were tripped, by the magnet when the boxcar passed over the reed switch.  The turnout springs were adjusted so the loco could pass through, easily, and the turnout would reset it's self.  It worked.

The circuit was easy enough to figure out, however, the wiring took some time. I still have those relays, but, lost my schematic diagram.  The entire test track was 8 feet long with a turnout at each end.  It sat on top of a book shelf, and would go back and forth, for as long as the power was turned on and each time the loco would reach the end it would uncouple a boxcar and pick it up , again on the return trip, after dropping another boxcar at the other end. (I had Kadee couplers on the boxcars with Kadee magnets imbedded in the cork for uncoupling)

My intention was to install the entire project on my (then) layout just to make the rail yard look busy. Unfortunately, it never did get installed.  I had to take the layout down to expand my electronic shop as business was brisk and I needed the space. 

So, yes, there was such an animal.

G. T.

Circuitrons' AR-1

Jim you are right. I was misreading the the info. It does look like the AR-1 does have two pairs of sensors not one. Jim does "Command Control" only deals with DCC or does it sometimes also deal with DC? My track is primarily for the grandkids so I'm keeping it DC since I'm into the model stick building. Thanks for your time!

Jim you are right. I was

Jim you are right. I was misreading the the info. It does look like the AR-1 does have two pairs of sensors not one. Jim does "Command Control" only deals with DCC or does it sometimes also deal with DC? My track is primarily for the grandkids so I'm keeping it DC since I'm into the model building.

Yes DC Command Control does Exist

The current one available is RailPro by Ring Engineering, I currently use this. Years ago and probably available if you look for it was Rail Command by CVP, I also used this. Both use(d) constant 14v or so DC on the track with a receiver that controls direction independently of track polarity like a DCC receiver does.

And yes Ring has an autoreverser available for such purposes.

I made one for myself using a track trigger and a DPDT relay tied into the turnout motor contacts (Tortoise) so it reversed the track polarity of the loop while in the loop (under the feet of the engine so to speak)

Espee over Donner

Is "automated" really necessary?

G'day Idaho

Is "automated" really necessary? Assuming you wire each balloon loop as for Dynamite Canyon - with a bridge rectifier-fed section on the "straight in road, followed by the "blocking section" with a diode feed in from each end, then with a simple but robust (short-protected) centre-off "turn-the-knob- either way" controller,. All your grandkids have to do is:

1. When the train stops in the balloon loop: - change the turnout (hint - a robust hand throw mechanism will help here. Another option is to use http://www.iascaled.com/simulate-a-spring-switch/  for the exit of each balloon loop if you don't want/trust the grandkids to change the turnout. This takes care of Step 3 as well)
2. Turn the knob on the controller the towards the way out of the balloon loop 
3. When the end of the train is clear of the turnout, stop the train and line the turnout back to the "main", (just like the real train crew have to do).

It is as easy as 1-2-3.for well under $10 for the basic "electronics" with a grandkid-driven turnout. 

As the grandkids get older, you can tell them about the" black magic of electricity, diodes and a whole flock of other stuff about model trains. Have fun with grandkids,

AR-1 DCC vs DC

The only difference between the two versions of the AR-1 is the extra set of photocells.  The reference to command control is referring to DCC, but there other forms of command control as well, but I would not worry about them.

As noted the AR-1 is an expensive solution, so If you do want to go with cheapest solution that would be a one double poll double throw toggle switch.

In the above when the train is in either reverse loop you throw the DPDT switch to change the polarity of the main line.  If you want to reverse the direction of the train at anytime you use the reverse switch on the train power pack.  To automate the above you put the AR-1 in place of the DPDT switch and put one photocell in the middle of each loop then when the train covers the photocell in the loop the mainline is reversed automatically.  The only issue is that if after the mainline polarity is reversed if you back the train out of the loop onto the mainline you would have to throw the DPDT switch to continue backing down the mainline.  Even though this shows the train exiting the loop one way you can go in and out in either direction, although for simple operation it should be going in the same side of the loop i.e. either enter both loops on the straight or both on the curve.

Remember you still have to align the turnouts for proper operation.

I am sure the AR-1 type circuit is out there on the web you could build yourself much cheaper, suprising it is not on Rob Paisley's site.  Or one could use an Arduino, photocells, and a relay to implement this as well and using the Arduino you could throw the turnouts as well.

Works, but...

Dear MRHers,

Completely understand that the Circuitron AR-x system will work for "fully automated" systems,
(believe I have an AR-1 in the warehouse somewhere...)

but the issue I fear with such "active relay switching" of the Mainline polarity is that it effectively breaks the intrinsic "throttle direct switch VS Mainline direction" relationship many operators and modellers rely on to make-sense of the world, esp if it's their first time on the layout in question.

Sure, mess with the polarity on the Loop trackage all you wish to "make it work",
but mess with the MainLine <> Throttle Direction switch relationship at your own peril...

The benefit of the dumb "bridge rectifier" solution (and "Dynamite Canyon" stop-block addition) is that the direction-switch on the throttle maintains a constant relationship to the Mainline,
(EG "throttle direction switch left = train goes to the left on the Mainline" Everytime, All the time),

there is no relay between the throttle and the MainLine which could lead to "throttle direction switch left = train could move in either direction" ambiguity.

Ambiguity for the human,
where circuitry is "keeping track and is full-control of things"
(IE a "fully automated" system),
is one situation,

letting the circuitry handle the "hard bit" of the loop,
while maintaining full-and-direct Human control over the "mainline bits",
is a seperate and significantly different situation, requiring a different diagnostic and solution headspace...

With a minimising of the "bridge rectifier track section" length,
(see the page 5 of http://www.zelmeroz.com/album_model/members/klyzlr/DynamiteCanyon.pdf )

it is even possible to run thru the loop's "junction turnout" (DCT "one way turnout"?) on the straight route, and manually reverse back out again (as many times as the operator wishes),
as if the turnout was a straight piece of track,
and thus use a part of the loop as manual-drive switching/terminus area...

Happy Modelling,
Aim to Improve,
Prof Klyzlr