Club layout short driving me nuts

[Maletrain]

Yesterday, a short circuit somehow occurred on the club's N scale DCC layout while it was shut down and I was away for an hour or so for a meal.  Before I left, I had looked at the breaker panel (for other reasons on a different section of track) and saw no shorts on any breaker.  When I got back and turned the layout back on, a dead section of track led me to the breaker panel, where a short was evident on one breaker. 

That breaker serves about half of a dog-boned loop, with reversing sections at both ends.  The dog bone is folded and twice twisted such that this section has about 50-70 feet of double track going through 2 very curved tunnels.   We spent more than an hour looking for anything on the tracks that could create a short, and found nothing.  No locos on that section, and the few cars at 3 industries were all on the rails.  Suspected a dropped coupler pin, but could not find any.  The last train through that section, the week before, had departed it without any issues.

Disconnecting the track feed from the breaker allowed it to reset, and checking the resistance between rails and also on the disconnected track feed shows only about 1 ohm.  So, the breaker seems to be good and the short seems to be a pretty good electrical connection.

So, I then spent a few hours in vain under the layout, trying to sort through a maze of poor wiring techniques.  The bus wires for that section go to several distribution blocks where track feeders are attached.  But, often bunches of feeders are soldered together before reaching the distribution blocks, so disconnecting individual feeders would require cutting them. To make matters worse, the whole 50-70 feet of double mainline has no internal gaps in either rail of either track, so the rails are completely parallel electrical paths with respect to their bus wires, with myriad interconnections by track feeders that can't be individually disconnected one at a time.

While under the layout, I could also see that the phases on many of the rails had previously been reversed - lots of red feeders going to white bus wires, and vice versa.  Many places, red feeder wires are soldered to white feeder wires on their way to the rails.  I believe that means that the rail phase in the opposite directions for the double tracks were initially not suitable for crossovers, and that was changed sometime in the past when crossovers were added.

I ran out of time and helpers last night, before getting very far down the buses with my attempts at trouble shooting.  Now I am trying to get my thoughts together on how to tackle this problem when I get back from Chattanooga.

One thought that has come up is to buy a thermal camera and start looking for warm spots on the wiring.  But, with such a low resistance, I am wondering if the problem is a miswired feeder that is soldered in place.  If so, would the warm spot be that feeder, or somewhere else, maybe at a rail  joiner that has more resistance?

How that short could suddenly appear after many years of operation and no (known) recent disturbance really puzzles me.  There are many places where feed wires have bare conductors or just a bit of tape on only one of the 2 phases, but I could not locate any shorts from those.

I may have to cut all of the feeders and rewire the whole section, unless I can soon find something on the rails that is causing this short.

Any suggestions for something I may be missing?

[Rasputen]

Are there any turnouts in this section, and if so, how are they controlled?

[Sumner]

.... To make matters worse, the whole 50-70 feet of double mainline has no internal gaps in either rail of either track, so the rails are completely parallel electrical paths with respect to their bus wires.....

I'd cut the rails about half way through the section.  One section should now be the problem.  Might have to run a long jumper from the command station to there.  Could makeshift with an extension cord if you don't have other wiring.

If you still can't find it in that one section cut it in half.

I had something similar the other day with very low ohms but not a dead short.  Ended up with it being two wires with spade connectors on their ends hanging in the air that I had disconnected and the spade connectors just barely touching each other.  Another a month ago with a handlaid turnout and a piece of fuzz metal still in an insulation gab that had been cut for the frog.

Sumner

[davefoxx]

This is one of my biggest fears as a layout owner: the mystery short circuits.  After checking the obvious, like a derailment or an engine/lighted car sitting on a rail gap, then it comes down to searching for a rail gap that closed up, probably due to seasonal changes.  They’re practically invisible, which is why I jam styrene into the gap.  But a club-built layout where the quality of construction can vary makes it tough to solve these issues.  Good luck chasing your club layout’s short down.

DFF

[Maletrain]

Are there any turnouts in this section, and if so, how are they controlled?

Most of the turnouts on the layout are old Atlas code 80 with the plastic frogs and imbedded rails that need a bit of paint to prevent shorts with wide metal wheels.  Most of them are controlled by servos.  There is one DS64 that runs a few old Atlas snap switches and one relay, but I don't think there are any of those in this section of track.  I am sure there are no polarized frogs to deal with.

[pdx1955]

You may have a couple bare connections touching each other in the wiring or across a barrier strip. I've had breaker sections go bad too, but that doesn't see.to be the issue. Since it was working and now it isn't,  I suspect something shifted just enough.

[Maletrain]

You may have a couple bare connections touching each other in the wiring or across a barrier strip. I've had breaker sections go bad too, but that doesn't see.to be the issue. Since it was working and now it isn't,  I suspect something shifted just enough.

I have found that on other layouts in the past.  Somebody even disconnected 2 feeders and loosely taped them together, apparently forgetting that they were still connected to the rails that were still powered through their joiners to other rails with electrified feeders.   Apparently, somebody eventually  pulled on the wires enough to get them to make contact, and the N-Trak layout shorted that loop.

But, I have already gone through the wiring on this layout's breaker district enough to make sure that there are no bare wires touching.  I even found a bunch of feeders pulled out of a screw-in connector - but for another breaker district. 

I first was suspecting that somebody had shoved something under the layout, or grabbed something and pulled it out, catching a wire in the process.  But, after asking,  I don't think anybody was near the layout while I was gone.  Even the closest potential "bumper" was on the other end of the layout.

[Rasputen]

This may sound corny, but how good is your hearing?  I have very acute hearing and I can think of at least three occasions where I have been able to locate the root cause just by listening for the source of the high frequency buzz that occurs if something is touching.  You may have to bypass any circuit protection if it is always tripped.

When you say you are measuring one Ohm across the rails, is this a good quality meter you are using?  What does the meter read with just its two leads touching each other??  At that low of a resistance, it's always a good idea to check the meter by itself.

[robert3985]

Having gone though a two day search for a mystery short when I pulled out all of my DC wiring on my modules and replaced it with robust DCC wiring...I went through disconnecting EVERY SINGLE track feeder (held by 3M suitcase connectors to sub-buses) and never found it.  I then decided I should look at the top, and I found it in about 30 seconds.  It was a broken closure rail that I evidently had broken when I broke down the layout and set the modules on their back edge on a 6' folding table to work on the undersides easily.

I am going to guess that the short is caused by track problems, not under-the-layout wiring problems.  If it suddenly appeared, it is most likely to have been caused by expansion/contraction of either the rails or the benchwork, probably a gap that has closed.  OR, if your switch machines are constant pressure (I am not sure how servos work) like Circuitron Tortoises, it could be caused by turnout breakage. 

First thing I would do is isolate it as much as possible.  For me, if my layout develops a short, I can disconnect the connectors between modules and when the short goes away, I will have at least isolated it to a six foot section.  Maybe you could do something similar on your club layout.

Secondly, I would cycle each turnout, checking continuity every time you cycle a turnout, then returning the turnout to its original position before moving on to the next one.  This will tell you if it's turnout wiring where the problem is, or isn't.

Isolating the short's position is the secret to finding it, but...snipping feeders under the layout is a drastic step.

Another choice would be to turn your track power off, then turn off the room's lights (do this at night) and then turn the power back on.  You might be able to hear it spark when you do this, or even see it in the darkened room.

Good luck! (I am not envious!)

Cheerio!
Bob Gilmore

[peteski]

If the short is one Ohm or less, I would not expect much of a spark.  I would feed that entire block with 12V 3A DC power supply and see what starts smoking or melting (yes, just kidding).

I agree with Bob that the problem is most likely on the top of the layout (not the wiring).

[Maletrain]

Thanks all for the thoughts.

Regarding my hearing, I am losing the high frequencies, but still good for the lower frequencies. However, the club environment is noisy, so I would have to try listening to the layout on an off-day with nobody there and the air conditioning turned off.

I am considering turning off the track power (and everything else) and powering the disconnected bus with a low voltage DC source - probably a couple of D cells.  That might work for using a thermal camera.  Again, that would probably need to be done when nobody else is around.  And carefully so that I don't melt plastic ties.

Regarding the meter readings: I use a small, old, cheap, auto-ranging Radio Shack multimeter.  I love it.  Resistance reading when holding the probes together with my fingers can be as high as 0.2 ohm if I have not gone to the trouble of cleaning the probes, first.  The lowest reading between the rails that I am getting is about 0.8 ohm.  Moving around, I see between-the-rail values usually between 1 and 2 ohms, but have not been able to find a pattern that points to the problem.  (Yes, I get no-continuity readings when the probes are not touching anything, and reasonable measurements for values for resistors, given their +/- ratings.)

Putting a constant power source on the track bus might allow me to see a drop in the rail voltage at various locations that might point me to the problem,

And, I am remembering some more things, such as there is another part of that track section's bus that has to go under a part of the layout that I did not investigate, and the vacuum cleaner is stored under that part of the layout.  That might have been a source of disturbance while I was at dinner.  So, when I get back to it, I will next look for issues under there.  (That might get tough, because previous modifications to the layout resulted in two layers of plywood separated by some inches, and I am not remembering if the bus is even visible, now.  )

But, first thing is going to be to take another look around the top of the layout.  The few cars on sidings will be removed, not just moved, and I will look for dropped coupler pins under them.  I will also take the advice to try throwing the turnouts.  Although none of the turnouts should have been moved to start running the train that revealed this short, people often unintentionally do throw turnouts in this section when dealing with other turnouts on the same panel.

The shorted track section is really convoluted.  It is effectively double track that goes out on one peninsula, makes a loop back over itself with most of that loop in a tunnel, then goes around to the adjacent peninsula, through another curved tunnel, and then loops back, crossing over itself again and going out on the first pennisula at a higher elevation.  That last end is connected to a reversing section that joins the two tracks at the top of that first peninsula.  The first end mentioned is connected to a similar breaker district for more double track, with another reversing section at its far end.  With no internal gaps, the only rail gaps of interest are at the ends of the problem section.  But the electrical bus under the layout is hard to follow, so I am not so sure that I will even see a hot spot with a thermal camera - at least not before I have put enough power through it to make it hotter than I think is a good idea.

This is probably my last post until after Chattanooga.

[John]

The most recent MRH has a good article on troubleshooting layout electrical issues .. worth a read before chasing ghosts

https://online.fliphtml5.com/buups/uqym/index.html#p=49

https://online.fliphtml5.com/buups/xcef/index.html#p=47

[Maletrain]

FOUND IT!!!

The technique that finally worked was to disconnect the shorted circuit from the breaker and power it with 6 VDC through a light bulb to limit current to about 1 amp.  Then, using a volt meter to check the voltage difference between the rails, a small area of track with the lowest rail-to-rail voltage difference pointed to the best location to do an in-depth dive into the wiring under the layout. 

There, we found some drops gathered together with masking tape around one phase, but the drops with the other phase left bare.  The tape had been penetrated by one of the drop wires, and that allowed it to touch the other, untaped set of wires of the other phase.  The tape had been stapled to the wood, and the penetration was against the wood.  But ,the untaped phase bundle had been "conveniently" tucked into the pocket created by the tape and the wood.  Apparently, somebody pushing/pulling stuff stored under the layout eventually caused the drop wire to poke through the tape and make the short contact.

I made this sound easier than it really was.  The search for the lowest rail-to-rail voltage had some brain-fart moments in trying to deal with a long stretch of 2 parallel tracks that were fed from a common bus with drops at various locations along the length of the track section.  The technique works great when tested on a single track fed from one end with a single shorting wire part way down the track.  But, with the bus feeding to multiple points on each of two parallel tracks, feeding the short from both ends and potentially both tracks, we were dealing with what is more like a spider web of wiring, rather than a single short path between 2 simple wires.  In fact, there turned out to be 6 short paths, as viewed from the rails on top of the layout.

Still, this was the only way we could deal effectively with a long, very convoluted and ungapped section of track.  And, because the rail phases on one of the 2 parallel tracks had been switched after construction to allow crossovers between the tracks where they were parallel in a dog bone, there were plenty of red feeders connecting to white bus wires, and vice versa - mostly done in inhospitable locations.

I suspect this club layout should be a nominee to the Guinness Book of Records for the worst wiring of a model railroad.  So, I can highly recommend the DC-through-a-lightbulb tactic for narrowing down the location of a short circuit.

Edit: I should have also mentioned that I was careful to be sure that the section of track I powered with 6 volts of DC was actually effectively isolated from all adjacent track sections, and that the DCC circuitry was not powered with the DC.

I am wondering if that would be a problem for Frog Juicers if they are powered with low voltage DC.  This layout has none of those, but this recommendation might need to be restricted to layouts without them. 

If nobody here knows the answer, I will ask on the Tam Valley group - as soon as my thread there about my sick Octopus 3 runs its course.

[$2 Bill]

Nice that you found the problem.

Your method works. Or for the meter, can use AC/DC Amp Clamp that clamp on a wire etc. (Many Clamp Meters are AC only.) UNI-T UT210C & others small Amp Clamps are handy for tight spaces or small loads. Larger Clamps might work but many are not made to read very low amp draws.

Related Method is a "Short Tester" that been around for Decades uses something to "blink" power while a "meter" follows a wire without bothering/cutting insulation.
Example: SG Tool Aid 25100 Short Tester available @ many places under different brands.
⚠ depending just how made, the "Blinker" may draw too much power for many Train setups. Many made them for Real Vehicles that uses 12V Thermal Flasher "cans" that need 3A to 5A or more to work.

Find other tracing/testing methods too.
Even "Tone Tracers" use in Telecom/Network world may help. These units are often @ Flea Markets & other 2nd hand places with most looks like this...
[ Guests cannot view attachments ]
If getting used units...
check the battery doors or pay very little. 9V batteries may leak while rest look "new."
Many probes have 2 "Spare" Terminals somewhere on them. Those are to connect "Butt Phones"/"Buttsets" for Telecom Workers in noisy places.
Many Tone Boxes have "Bed of Nails" & other weird Clips. The "Beds" are to Piercing insulation on small wires. Weird Shape is often to work on 66 blocks that Telecom been using ~ 100 years.

[Maletrain]

I would not expect a "tone tracer" tone to stop at a short circuit.  I would expect the tone over the entire electrical circuit that is shorted.  That would not help me find a short in there "somewhere".

In the particular case that was driving me nuts, the location of the short was where the same section of track passes by 3 times in close proximity, at slightly different altitudes.  All three sets of 2 tracks each were powered from a single clustered set of drops, all 6 pairs of drops were soldered to 2 rings, one for each phase.  The location of the unintended connection was from those 2 rings being separated by just a piece of masking tape, which had been penetrated by one wire that stuck out at an odd angle and touched the other ring.

So, low very voltage between rails occurred in 3 places on the rail pairs, and what attracted our attention to that area was that those 3 places with low rail-to-rail voltage were in the same physical location.

Electrically, this track section is more like a very large spider web.  The shorted section is quite long, goes into a loop and comes out and goes over itself, then around to another peninsula, loops back over itself again and goes over to the first peninsula again.  The supporting electrical bus splits shortly after the circuit breaker to go out on each of the peninsulas, where it has drop wires in multiple locations, with those drops at each point gathered together to feed whatever parts of the track are nearby to a bus connection.  So there are many places where the drops make short distance connections between track locations that are far apart along the rails.

So there are many feed paths to any point on any rail, even when there is no short.  When the two buses are shorted together "somewhere", the possible paths to any point are even more numerous and hard to figure.  The short can be fed from both directions on any rail, with many connection points at various distances from the power feed point.

So, the only thing that really stands out is the very low voltage when the measurement is taken between the rails at a location very close to where the short is connecting the rails.  That is why this method works.  But, it only narrows down the search location for the unintended connection,

If I had tried to look for this short with a thermal imager, I doubt I would have found it.  The location was where there was a very limited sight distance, up between wood supports and tight against one.  And, it did not feel at all warm when we touched it.