Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Persistent Auto-Reverser Problem

6260 views
23 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    January 2008
  • 520 posts
Persistent Auto-Reverser Problem
Posted by mreagant on Monday, May 14, 2012 9:10 AM

I continue to have a problem with the operation of my AR.  It is wired from the main buss to the rails in the reversing section.  This is a direct connection with no interveining switches, toggles, etc. There are no other power leads into the reversing section.  The reversing section is properly isolated at both ends.

A locomotive (DCC/sound)  enters from the main section and transverses the reversing section smoothly.  At exit of the reversing section, it stops, shuts everything down on the layout, including other locomotives, and after 2-3 seconds light and sound come back, but it does not resume running unless I nudge it.  The AR is an MRC AD520.  The AR emits an audible "click" at the moment of shut down, and then will continue clicking until the loco is completely out of the reversing section.

I have test run a BLI F7 A/B which ends up half in, half out during the shut down.  I've also test run a Kato SW7.  Results are the same.

I need a fresh idea to trouble shoot this.  I've got one or two more ideas to try, but then I'm running on empty.  Help!

  • Member since
    July 2003
  • From: Sierra Vista, Arizona
  • 13,757 posts
Posted by cacole on Monday, May 14, 2012 9:29 AM

You may have a faulty auto-reverse module,  If it has a sensitivity adjustment, I have never had any luck getting them to work equally well at both ends of a reverse loop, and replaced them with a different brand --  Digitrax, or even better a totally electronic one such as the PSX-AR.  A module with a mechanical relay is too slow to respond.

 

  • Member since
    December 2004
  • From: Bedford, MA, USA
  • 19,814 posts
Posted by MisterBeasley on Monday, May 14, 2012 9:53 AM

What happens if you run the train the other way, going in the "exit" and out the "entrance?"  Does it still go in smoothly and short at the exit?

Does this happen only with a full train, or is just an engine enough to make it fail?

It takes an iron man to play with a toy iron horse. 

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 22,643 posts
Posted by selector on Monday, May 14, 2012 11:03 AM

It is often recommended that the gaps be staggered transversely across from each other by about 1/8".  Did you do that?   Instead of, say, using a rail saw and cutting at 90 deg so that both gaps are directly opposite each other, you should place one of them before the other by a fraction of an inch.  This may not affect the analog mechanical ones, but Digitrax recommends that practice for their digital reversers, such as the PSX-AR.

Crandell

  • Member since
    March 2007
  • From: South Carolina
  • 1,719 posts
Posted by Train Modeler on Monday, May 14, 2012 11:04 AM

The audible click sounds like the MRC unit may not be fully electronic, don't know for sure.   But, cacole has a good idea if you have the money for the PSX-AR.    BTW, we still use DPDT in certain areas--no problems.

Richard

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, May 14, 2012 11:54 AM

 One thing not mentioned, what is the DCC system? If it's one of the ones on the lower end of the power scale, it may be tripping the command station before the AR reacts.

  Verify absolutely that no feeders are crossed anywhere - disconnect the AR and check for any voltage between either rail outside of the loop and either rail of the loop track. One feeder connected to the main when it should be connected to the AR can make really wierd things happen.

 The Digitrax AR-1 is adjustable with a potentiometer to adjust the trip current. Generally works with anything because of this. Too slow? If you count microsecodns I suppose. They flip plenty fast enough to avoid sound dropouts. And plenty reliable. The relay contacts are rated for enough operations that if it clicked once per second 24 hours a day, 365 days a year, it would still eb going 10 years later.

                         --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    December 2004
  • From: Bedford, MA, USA
  • 19,814 posts
Posted by MisterBeasley on Monday, May 14, 2012 12:16 PM

mreagant

The AR emits an audible "click" at the moment of shut down, and then will continue clicking until the loco is completely out of the reversing section.

This sounds like the auto-reverser is unable to resolve the short by flipping, so it keeps toggling back and forth.  The things that could cause this are a misplaced feeder, incomplete gapping of the reverse section, or a train or freight car truck crossing the "entrance" gap while the engine is crossing the "exit" gap.

Just to correct a previous mis-statement, the PSX-AR is made by DCC Specialties (Tony's Trains,) not by Digitrax.

It takes an iron man to play with a toy iron horse. 

  • Member since
    January 2008
  • 520 posts
Posted by mreagant on Monday, May 14, 2012 9:35 PM

Thanks for all the helpful ideas.  I have not had time today to go to the layout and follow up.  That is tomorrow's priority, and I'll report.  I can say that I've used a multi-meter to check continuity/polarity and that is not an issue so I don't think this is a "crossed wires" problem.  I'll recheck this, but the AC voltage on the main tracks entering the loop  is about 15.6v and at the other end at the exit point I get about 1v less.  The main track at the exit from the reversing section loops through a hidden staging yard and the linear track footage through that yard is far greater than the track footage from the entering point back to the DCC system.  Unless I'm completely mis-remembering my high school (and college) physics,  the distance from the power supply should not create a voltage drop. The system is an MRC Prodigy Wireless.

There are at least half a dozen sound equiped locomotives  sitting at idle on the layout so I wonder if the problem could be a system overload when the AR tries to reset.  The AR is rated at 5 amps, but maybe the cumulative effect of these active sound/light systems at the moment of exit is causing the problem.  Is that possible?

Mike

  • Member since
    December 2004
  • From: Bedford, MA, USA
  • 19,814 posts
Posted by MisterBeasley on Tuesday, May 15, 2012 6:22 AM

Are all these idling engines actually in the reversing section?  The steady-state power going to them should not be a problem, but you need to consider "inrush current" when a lot of sound engines are simultaneously powered up.

Still, I don't think this is your problem.  However, just to eliminate this as a possible problem, take the idling engines off the layout and repeat your testing.

It takes an iron man to play with a toy iron horse. 

  • Member since
    December 2001
  • 1,932 posts
Posted by Stevert on Tuesday, May 15, 2012 7:08 AM

mreagant
  Unless I'm completely mis-remembering my high school (and college) physics,  the distance from the power supply should not create a voltage drop.

You're mis-remembering.

Copper wire is a very good conductor, but it still has some resistance so there will be a voltage drop.  More wire length, more current, and smaller wire size will make it worse.  However, if you size the wire properly, it's usually negligible and can be ignored.  Here's some good info on the topic:

http://www.wiringfordcc.com/trakwire.htm 

This brings up another question, though:  Have you done the "quarter test" at various spots around the layout?  If not, you should.  If there is excessive resistance in your wiring, the loco entering/leaving the reverse section may act more as an overload, than the dead short that's required to flip the polarity.

  • Member since
    January 2008
  • 520 posts
Posted by mreagant on Tuesday, May 15, 2012 10:24 PM

All you guys have been very helpful as I try to sort this out.  Many thanks.

I was only able to work on this matter for about 15 minutes today and, of course, the AR worked at both ends--about half the time!   I'll dive back in tomorrow, but I need to ask another question:

The voltage drop at the "far" end of the loop is close to 2v---15.6v down to 13.8v on the other side of the insulated gap.  Is that enough to be causing the AR to not switch polarity and hold it?  I can easily run a feeder from the buss to the other (low voltage) side of the main.  I may do that tomorrow. 

Any thoughts?

Mike

  • Member since
    December 2001
  • 1,932 posts
Posted by Stevert on Tuesday, May 15, 2012 11:00 PM

mreagant
The voltage drop at the "far" end of the loop is close to 2v---15.6v down to 13.8v on the other side of the insulated gap.  Is that enough to be causing the AR to not switch polarity and hold it?  I can easily run a feeder from the buss to the other (low voltage) side of the main.

  That's a pretty substantial voltage drop, enough that it could certainly be the cause of your AR issues. 

  Have you tried the "quarter test" at various spots around the layout yet?  It's quick, easy, and can tell you a lot about the overall health of your bus/feeder system.  Lacking that, tell us about your bus and feeders - gauge and length for both the bus and for the feeders, what is the distance (track length) between your feeders, etc. 

Your comment about running a feeder "to the other (low voltage) side of the main" sure makes it sound as if you have marginal wiring to the track.  That could cause the voltage drop you're seeing, and that voltage drop in turn could cause the AR issues.

Oh, and it'll work at either end about half the time because the reverse loop's phase will match the main line's phase on either one end or the other.  And when the AR does flip the phase, well, now the other end becomes the one that "doesn't work".   That's likely to continue until the underlying issue gets resolved.

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 22,643 posts
Posted by selector on Wednesday, May 16, 2012 10:21 AM

Stevert

  That's a pretty substantial voltage drop, enough that it could certainly be the cause of your AR issues... 

Agreed, and not only substantial but probably critical.   I forget where, but I read fairly recently that DCC doesn't like voltage differentials greater than a volt.  Some will do okay, while others will give you grief.  I don't tolerate voltage drops on my layouts greater than a full volt, and so far none have come close.

Crandell 

  • Member since
    March 2008
  • 387 posts
Posted by ba&prr on Wednesday, May 16, 2012 3:49 PM

The problem is in the way you wired it up. The wires from the AR ARE NOT connected to the bus wires. There should be 4 wires coming out of the AR. 2 go to the track on 1 side of the gap. The other 2 go on the other side: IE 2 red on one side 2 yellow on the other side. Go to Digitrax web site and look up their AR1. Down load the manual for it and follow the wiring diagram. This should solve the problem if your AR isn't fried.   Joe

  • Member since
    February 2008
  • 7,397 posts
Posted by maxman on Wednesday, May 16, 2012 4:19 PM

ba&prr

The problem is in the way you wired it up. The wires from the AR ARE NOT connected to the bus wires. There should be 4 wires coming out of the AR. 2 go to the track on 1 side of the gap. The other 2 go on the other side: IE 2 red on one side 2 yellow on the other side. Go to Digitrax web site and look up their AR1. Down load the manual for it and follow the wiring diagram. This should solve the problem if your AR isn't fried.   Joe

I believe that he said that his AR is an MRC AD520, not a Digitrax.

  • Member since
    January 2008
  • 520 posts
Posted by mreagant on Wednesday, May 16, 2012 4:21 PM

I'm about to go recheck the voltage readings to confirm what I've already posted.  If earlier readings were correct showing an almost 2v drop, I'll ad that additional set of feeders and see what happens.  As to the AR being hooked up incorrectly, I posted a seperate inquery several days ago and was assured by several responders that attaching the wires to the main buss rather than the track outside the reversing section would work just fine so long as I don't cross polarity with the wires connected inside the reversing section.

Regarding the "quarter test."  What exactly will that prove--that a short at a given spot on the  layout will trip the breaker and shut everything down?  I already know that by experience, so what am I going to gain? 

Mike

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, May 16, 2012 5:28 PM

 The deal with the quarter test though is that you do not press down. Just the weight of the quarter. This is typically al ittle different then say if a loco derails and the weight of it is pressing down ont he rails. You can easily have spots on the layotu where if you put a screwdriver across the rails, it trips the breaker, but if you set a quarter on the track it doesn't. Just as Ken about inadequate feeders and the breaker not tripping...

              --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2006
  • 3,312 posts
Posted by locoi1sa on Wednesday, May 16, 2012 6:03 PM

Mike.

 How many feeders from the AR to the reversing section? How long is the section?  It sound to me that there is some resistance or bad connections (same thing) between the AR and the reverse section.

  The importance of a quarter test is to test the wiring and connections to the command station and boosters. Never Never underestimate the importance of this test. Melted rail and ties, toasted decoders, and the chance of a fire should never be taken lightly. If the safety systems do not trip instantly then trouble you will have. It does not cost anything to do and takes just a few moments in time. A friend of mine was working on his layout and had a loco pulling a track cleaning car on the main. He went to make a phone call and when he returned found that it had derailed on a switch. The booster did not trip and sent that 5 amp power through the 28 gauge feeders. The insulation was melted off and the wire got cherry red hot. The switch points had turned blue and melted into the ties.  If he had been a few minutes longer we are sure a fire was next. Go ahead and take a chance. Its not my house that will burn down. Too many times modelers will rely on rail joiners and sparse feeders to transmit all that amperage.

         Pete

 I pray every day I break even, Cause I can really use the money!

 I started with nothing and still have most of it left!

  • Member since
    December 2001
  • 1,932 posts
Posted by Stevert on Wednesday, May 16, 2012 8:33 PM

mreagant
Regarding the "quarter test."  What exactly will that prove--that a short at a given spot on the  layout will trip the breaker and shut everything down?  I already know that by experience, so what am I going to gain? 

Mike

Mike,

  The quarter test isn't so much to see if the breaker will trip at a given spot, but more to insure that the breaker does trip at every spot.

  Auto-reversers, especially the relay-based designs, are like circuit breakers except that instead of cutting the power, the relays are wired to flip the phase.  And like a circuit breaker, a high level of resistance (ie, a load) on the output side looks "normal" and will prevent them from flipping.

  An insufficiently sized bus, or too few feeders, will increase the resistance that the AR/circuit breaker "sees", with the result being that it won't trip/reverse.

  A nearly 2v drop from one side of the layout to the other indicates a fair amount of resistance somewhere.  If you look at the Wiring for DCC Web site I posted the link to earlier, you'll find that nickle-silver track has a relatively high level of resistance, as does copper wire sized too small for the amount of current it's expected to carry over a given distance. 

  That's why I suspect your underlying problem is too small of a wire size for your bus, and/or too few feeders.  The quarter test every few feet around your layout (or some info on the length/size of your bus and length/size/spacing of your feeders, which was also requested earlier) will go a long way towards identifying the problem, or at least eliminating one possible, and at this point I'd have to say highly probable cause.

  • Member since
    January 2008
  • 520 posts
Posted by mreagant on Wednesday, May 16, 2012 9:30 PM

After going back through the post I have pretty much figured out a plan of action.  I do have too few feeders in the longer section.  Since most of it is a 'hidden staging yard' that is not currently used, I had not paid much attention to it.  That section of the layout has only two feeders for 20 liner feet of track, much of which is 4 or 5 track parallel in the yard.  I'm going to go ahead and run a feed from the bus to the end of that yard section at a point just outside the reversing section.  It may be a day or so before I can get that done, but I will post results.

You all have been a lot of help.

Mike

  • Member since
    January 2007
  • From: Eastern Shore Virginia
  • 3,290 posts
Posted by gandydancer19 on Thursday, May 17, 2012 10:21 AM

 

I don't know if this has been real clear before, but you can set up a separate bus for just the reversing loop.  This reversing loop bus should be connected to the output of the AR and nowhere else.   The track in the reversing loop should be connected to the reversing loop bus by feeders just like you would do on the main section of layout.   The reversing loop bus should use the same heavy gauge of wire that your main layout bus uses.  (AWG12 or 14)

20 feet of track in a reversing loop is long enough that you should be using a bus for your reversing loop wiring.

Elmer.

The above is my opinion, from an active and experienced Model Railroader in N scale and HO since 1961.

(Modeling Freelance, Eastern US, HO scale, in 1962, with NCE DCC for locomotive control and a stand alone LocoNet for block detection and signals.) http://waynes-trains.com/ at home, and N scale at the Club.

  • Member since
    January 2008
  • 520 posts
Posted by mreagant on Thursday, May 17, 2012 6:37 PM

Drum Roll, Please!  It worked!  Feeder off the main bus to the rails just outside the reversing section indicated only a very small voltage drop --0.05v-- over the gap.  This without any other new feeders to the long section through the 'hidden' staging yard. 

Multiple test runs through the reversing section went flawlessly.

I could not have gotten through this Gordian Knot without the help/ideas from all you guys, so a great big Thank You to each of you.

Now to look at some of the suggestion you offered to make the layout even more bullet-proof.  I'm sure I'll be back soon with more mysteries to solve.

Mike

  • Member since
    December 2004
  • From: Bedford, MA, USA
  • 19,814 posts
Posted by MisterBeasley on Thursday, May 17, 2012 7:51 PM

Thank you, MIke, for working through this with us.  We've all learned something tonight.  This is another thing to add to our bag of 'things to try" when auto-reversers don' behave.  At the start, this was a puzzle, but it shows how a few questions, a few answers and a few experiments can narrow down a problem and solve it.

I've been watching this forum for years, and, as far as I know, this is a solution that's never come up before.  Well done.

It takes an iron man to play with a toy iron horse. 

  • Member since
    March 2008
  • 387 posts
Posted by ba&prr on Friday, May 18, 2012 4:02 PM

I have an MRC auto reverser on my layout. I wires it per the instruction sheet. It worked with the loco entering one way but not the other way. I went to the Digitrax web site to look at the manual for their AR1 and wired the MRC one up using the AR1 manual and it worked. .Joe

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!