"Feed every piece of rail" - what's practical?

[narrow gauge nuclear]

For DCC most of the responses are right on.  I have commented on this before in other such posts.

This from me an electronic engineer by profession and the model railroad engineer by hobby.  I run HOn3 and formerly HO but all might benefit regardless of gauge.

1.   I run 12 gauge solid copper buss feeders under the layout.

2. I tap in off the buss feeders in the middle of each 3 foot length of flex track and also in the center of any isolated shorter run on a spur or siding.  

3. I feed off the buss feeders directly to all switches as needed.

4. By Kirchoff's circuit laws, you can almost halve your resistance at most any point, electrically, by soldering rail joiners provided they don't work loose, if the above procedures are rigidly adhered to.

In the above given advice, the absolute greatest distance of N.S. rail that has to be traversed by the buss electricity to reach any engine is only 18" with virtually little or no effective reistance to the DCC power or signal.

This is why the largest practical buss wire MUST be used based on the actual current required by your worst and most power hungry engine's motor.  O and G gaugers might consider #10 solid wire.  I realize that most HO folks use 14 gauge buss wire.  But I still choose the overkill of #12.  (belt and suspenders)

Richard 

[Perry County train man]

I run DC on my HO rail way and wired every 3ft just in case I decide to upgrade.

[hornblower]

I try to add feeders for every three feet of track.  Thus, every length of flex track gets feeder wires.  Where I have multiple short lengths of track adding up to three feet or less, I will solder the rail joiners of those pieces of track and add feeders to that composite length of track.  Yes, its a lot of soldering but I only do it once for each length of track. 

[rrinker]

 If rail was round, the .083" of Code 83 would fall between #12 and #11 wire. Since the width is less than the height (usually) it's probably more like between #12 and #13. Over a 15-20' run, #14 wire is plenty sufficient for a bus run with no noticeable voltage drop, thus 15-20' of Code 83 rail is also plenty sufficient, allowing for less conductivity than copper but larger size than #14.

                  --Randy

 

[ATLANTIC CENTRAL]

A friend and fellow modeler, who is by chance an NMRA Master Modeler, has a 30 plus year old layout that is about 15 x 40. It was obviously originally built and wired for DC block control. It is now DCC.

He converted it to DCC by simply connecting his existing block feeders into groups, connecting the groups to ciricuit breakers and connceting the circuit breakers to a base and one booster. He has not added feeders "every three feet". Most of his rail joints are soldered, most of his track hand layed.

His average "block" length is likely 15-20 feet.

His layout runs flawlessly on DCC, just like it did on DC.

I have not done the math, but I suspect a piece of code 83 rail has at least the cross section of a #12 wire, maybe even twice that. So even with slighly increased resistance over copper wire, your rail is likely just as good a "bus" as your bus.

But what do I know, I'm just an electrical designer with a pickup, some guns, and little trains without brains.

Sheldon

[rrinker]

 I just haven't had that problem. Even the paint and glue has no effect, it gets int he open space but neither seems to expand as it dries which would have t happen for it to force the joiner open. 2 layouts now, and LOTS of running, if anythign's goign to loosen the rails up, it's a constant parade of trains. Once the rail is painted and balalsted, that pretty much glues the joiner in palce too, it woudl take a lot to make them move and get loose. Maybe because I use foam, so the base never moves with changes in temperature or humidty? Or the caulk used for the track and roadbed, which allows things to shift slightly without stretching the rails?

 All I know is it works, and works well. So well that I can get a new loco, go into the train room for the first time in a month, put it on the rails, and it runs. Perfectly. No doropouts, no blinking headlight. No track cleaning - ever.

                      --Randy

[richhotrain]

Do as I say, not as I do !

I solder the rail joiners on all curves.  I add feeder wires to rail joiners at the ends of all turnouts.  I add feeder wires to rail joiners every six feet of straight track.

So what is wrong with that?

Soldering feeder wires to rail joiners is fine assuming that the rail joiners make constant and full contact with the rails.  But rail joiners can loosen over time.  Ballasting glue can and does get between the rails and the rail joiners.  Don't ask me how I know.

On my next layout, I will solder all rail joiners on curves and those rail joiners will have feeder wires.  I will solder all rail joiners on every piece of straight flex track (i.e., every 3 feet) and those rail joiners will have feeder wires.  I will add feeder wires to every rail joiner at the end of every turnout, although I will not solder the rail joiners to the turnouts.

Rich

[fwright]

gregc

Even in his article in "Realistic Reliable Track", Andy Sperando suggested not to takes his words too literally.   But while I don't plan to use rail joiners,  I have no qualms about electrically joining adjacent rails with short pieces of 22g wire across the joints.

The question is how long a section of rail, or sections of rail with soldered wired connections
should be fed by a feeder wire? 

I have never used rail joiners with handlaid track.  But I soldered a 24 gauge magnet wire to each piece of rail before I spike it down.  I then tied these wires to a 16 gauge common rail bus under the layout, and to a block terminal strip for the DC block wire.  This was quite sufficient, but I only run single engines in a given block.

By using jumpers across the rail joint, you are using the rails as your bus wires.  Although nickel silver has a much higher resistance per foot than copper, the size of the rail cross section makes up for a lot of the higher resistance (assuming you are not using code 40 rail).

The voltage drop is going to depend on how much current you are drawing through the rails on a given set of feeders.  DCC doesn't like more than about a 5% voltage (0.7 volts) total drop; DC can easily handle 10% drops. 

The quarter test is considered a good test of whether or not you have too much resistance for DCC.  I would experiment with how many lengths of jumpered rail you can go before the quarter test fails.  FWIW, the quarter test using the overload on a DC power pack is equally valid.

my thoughts, your choices

Fred W   

[rrinker]

 On my previous layout, I only soldered flex track on the curves. This time I've takento solderign every other pair - so the two pieces in a curve are soldered, then the next piece is NOT sodlered to those, but the 4th piece is soldered to the third piece:

------||--------|-------||------

where || = soldered joint with feeders and | - plain joiners (with feeders attached)

Probably overkill, as with the soldered joint with feeders, the max distance from a feeder would be 3 feet with no joints. But, this way I should never ever have to go back and add more feeders. Using a foam base plus being in a mostly climate controlled room ( I tend to not set the heat as high nor set the air conditionaer as low as inthe main living space  bit it is an insualted and carpeted room), I shouldn't have any issus with expansion and contraction and it might be safe to sodler every joint, but with different materials and/or more variation in temperature and especially humidity I would advise caution. Also, I don't like the idea of ever soldering a turnout in palce - if any piece of track is going to need to be replaced, it's a turnout, if something breaks beyond the ability to repair in place. A soldered one will require a bit of work to get out, and then put back in place. Removal is easy - just cut through the rail, but what about reinstallign it, or installign a new one? Unsoldered, it's simply a matter of sliding the joiners back and lifting it out.

                --Randy

 

[selector]

As a general principle, I do favour ensuring each contiguous length of rail gets its own feeder.  However, if one can solder effectively, and joins another length of rail to a first one by soldering the feeder, and if one at the same time solders a feeder at that junction, you effectively feed only three feet in either direction...as Simon and Randy have suggested.  Said differently, the electrons need to go no more than 3' from the fed point.  In our hobby, that is most acceptable.

However, there are times when it requires one to take the time to feed one segment by itself, regardless of its length, also suggested.  Sometimes you must leave a gap, meaning the rail one one side of the gap will be dead unless you power it individually.

It only takes a few minutes drill the roadbed, bare some feeder wire ends, thread them, kink the one end against the rail web, wrap the other around a bared bus or sub, and you can do those things while the soldering iron/pencil heats safely nearby.  Get several such feeder pairs prepared in an area and solder them all up at the same time.  It works, and things move along.

Crandell

[ATLANTIC CENTRAL]

Here is an alternate view on the subject:

DC or DCC - solder all, or most all of your rail joints.

ESPECIALLY those on curves and in areas of short setctions of track.

Been doing this for 40 years, and my father did it for 15 years before me, solves most of these problems.

I run DC, and have track sections (blocks) as long as 30'-40' with one feeder - no voltage drop or power issues.

Sheldon

[mobilman44]

Hi!

I've had layouts for decades, but built the first DCC layout 3 years ago.   I did a lot of homework, and realized that electrical contact was a extra important for good DCC operation.   So, I put in feeders every 3 to 4 feet, and on all dead end tracks (no matter how short).  As the layout is 11x15 with a lower staging area, and "track intensive", it certainly was a lot of feeders to attach.  

And, to complicate matters, I put in 8 power districts as well - each with its own breaker.  

I won't minimize the effort, although doing it "mass production" style made it easier.  But the final result has been that the trains (and many sound systems) work beautifully, and I am happy with the results.  The extra effort paid off.

To make a point, here is something I realized this a few months ago..........   There has been a lot of talk on this forum over the years about feeders.   But never have I seen it written that someone was sorry they put in "too many" feeders.  

[rrinker]

 Worst POSSIBLE case on my layout is the same as Siomon's - a feeder soldered to the center of a pair of soldered together sections fo flex track. But EVERY one of my rail joiners, except for the insualted ones between power districts, is a feeder. I make all of my joiners feeders - don't buy the Atlas ones, or you'll go broke. That means for example, that a turnout has 3 pairs of feeders. I have no stalling or pwoer loss issues, and no problems even after painting and ballasting. In a situation where the insulated joint landed between pieces of flex with the only feeders beign at the opposite end of each piece, I cut The flex back and used two short filler sections. Each filelr is soldered to the flex with feeders, then there are insualted joiners between the fillers. Overkill? Perhaps, ebcause everythign ran fine with just one pair of feeders powered for testing, but with every joint powered I will never have a problem, and it's really no bother to do this - I make up dozens of feeder joiner pairs at the bench and keep them in a drawer in my tracklaying cabinet. The joienrs are all fresh, I have a small collection of extras used for test fitting whuich have loosened up through repeated use, but the ones with the feders attached are fresh and tight and don't go in untilt eh track is ready for final attachment. I liberally apply paint around the joiners so no shiny bits show, so if the paint is goign to wick into the joiner, it surely does do. The tigth conenction of a fresh joiner prevents loss of contact when doing this, and nothign changed after applying ballast and spraying it with alcohol and then alcohol/glue mix. Second layout using this technique, never a power problem.

                      --Randy

 

[simon1966]

I believe that on my layout the longest run would be a pair of feeders in the center of 3 pieces of 3' flex track that are soldered together with rail joiners.  Not saying that this is right or wrong, but never had a problem with power.  Every section of my track has either a solder connection to an adjacent piece of track or a set of feeders.

[MisterBeasley]

I feed every other piece of rail, at least.  If I'm using short pieces, I solder the rail joiners so they effectively become one long piece of rail.  You don't want to depend on loose rail joiners for power distribution, but soldered ones are pretty reliable.

Your track bus should always be the primary carrier for your track power.  Even with feeders on every piece of rail, most of the power should flow through the bus to the track further down the line.  This gives you the least voltage drop and the best performance.

[gregc]

Even in his article in "Realistic Reliable Track", Andy Sperando suggested not to takes his words too literally.   But while I don't plan to use rail joiners,  I have no qualms about electrically joining adjacent rails with short pieces of 22g wire across the joints.

The question is how long a section of rail, or sections of rail with soldered wired connections
should be fed by a feeder wire? 

thanks