Intermittent Reverse Loop Problem

[Ed Whichard]

Have recently laid track and wired a new DCC layout.  I have an intermittent problem with reverse loops.  Layout is divided into 4 sub-districts each connected to a single Digitrax PM42 power manager.  Each sub-district is set up for Auto-reversing.  Layout is powered by a Digitrax power supply and DCS200 Command Station Booster.

 The problem is that when a loco crosses the separation point between two adjacent sub-districts, one of three things occur.  Sometimes loco crosses without any problem.  Other times loco will stop, command station will indicate a short, PM42 corrects the short and loco continues on.  Lastly, the loco will stop, command indicates a short, PM42 tries to correct problem but fails, command station shorts again, and process repeats until I physically move the loco across the break.

 The problem can occur at any of the four separations.  It makes no difference which direction loco is travelling when crossing a division point.  I am currently operating five locos, a BLI Paragon 2, two Walthers Proto 2000, an Atlas Gold series, and a 20 year old Kato converted to DCC with a Digitrax decoder.  Problem has occured with each of them.  I am also using two Digitrax throttles, a DT402 and a UT4.  Problem can occur with each throttle.

 I have tried the various Trip Current Settings and Short Circuit Sensitivity Settings as indicated in the PM42 manual but cannot correct the problem.  I know that intermittent electrical problems can be difficult to pinpoint but this one is driving me up the wall.

Any Suggestions?

 Ed in PA

[TomDiehl]

First, let me ask the obvious question: Why do you have an auto-reverser on each subdistrict? These should only be used when you have a reverse loop or wye.

[MisterBeasley]

I think Tom's question leads directly to your answer.  You've got two reversers, each trying to do the same thing at the same time - flip polarity as the engine crosses the boundary.  When both of them flip, the problem is still there.  You only need a reverser where you have a reverse loop or wye.

A wise modeler once told me not to "cascade" auto-reversers.  Don't have one connected to another.  I took his advice, and have had no problems.

 

[rrinker]

 That right there is probably the problem, the two adjacent districts each try to handle the reverse and just fight each other.

 Also, Digitrax no longer recommends using the same PM42 as both a breaker and autoreverser, instead feed the reverse loop with an AR1, feed the AR1 from a PM42 section for short protection.

                                                       --Randy

 

[jbinkley60]

If the command station is shorting and not the PM42 then you need to increase the delay on the DCS200 to allow the PM42 to operate.  Change OpSw18 to closed.  This will extend the DCS200 shutdown from 1/8th to 1/2 second.  Also try shorting each section (i.e. quarter on the track test) and ensure that the PM42 shuts the track off and the DCS200 continues to run.  I agree with the others that multiple autoreversers adds complexity and isn't necessarily more reliable.

 

[TomDiehl]

So Ed, you've got us wondering, what did you find?

[Ed Whichard]

First, some more background.  I am returning to model railroading after a 20 plus year absence.  DCC was, maybe still is, completely new to me.  My layout is basically an "L" shaped dogbone placed on 25 foot and 15 foot long shelves.  The long sides of the "bone" and each end loop are electrically isolated from each other.  Several turnouts and passing tracks connect the two long sides of the dogbone so that "reverse loops" are created.

A "leap in logic" lead me to connect each of the four sections to a single Digitrax PM-42 power manager with each section set up as "auto reversing."  My theory was the "electronics" in the power manager would detect and correct reverse loop situation.  That led to the "intermittent problem first described.

 Yesterday I changed one section, one of the end loops, from auto-reversing to just a circuit breaker configuration.  Part of the problem disappeared, at least on that section.  I think what I need to do next is rewire the power manager so that only one of the four sections is auto-reversing but am not sure which section that should be.

Thanks for all the comments.

 Ed

[simon1966]

I would switch 3 of the zones back to breaker mode and have one of the long dog bone sides set to auto reverse.  Based on your description I think that would work much better for you. 

[rrinker]

 Keep in mind a section of a PM42 set as auto-reverse does NOT also function as a breaker, so while the polarity would reverse going in and out of the loop, if a short occurs, such as a train derails in the loop, the whole layout is going to shut down because it will trip the breaker on the booster. Digitrax no longer recommend cacading section ont he same PM42 - ie, set section 1 up as a breaker, feeding section 2 set up as auto-reverse. What I would say is best practice for this would be to set up all 4 sections of the PM42 as breakers and get an AR1 to connect to the loop. The AR1 will handle the polarity issue, and if a short occurs then the PM42 will trip for that section.

                                                 --Randy

 

[Ed Whichard]

Thanks for the suggestion.  Based on my experiment with one of the sub-sections yesterday, I had pretty much came to same conclusion.  I'm definitely learning aout DCC the hard way, a lot of trial and error.  Have done considerable reading and research but sometimes find it difficult to apply generalized insturctions to specific situations.  I'm keeping good notes.  After I've everything possible wrong at least once, I should know how to do it right.  Maybe I'll write a book.

[simon1966]

Ed strictly speaking this is really not a DCC issue as the reverse loop problem would have manifested itself if this were a DC layout

[Phoebe Vet]

Ed:

If you post a diagram of your layout we can easily show you where to wire your auto reverser(s).