Why do I need 18 guage wire for feeder's?

[rrinker]

I wouldn;t say #18 is overkill for feeders, nto with 3 and 4 foot lengths. But that's about as LARGE as I would say is practical for HO, and probably too big for N. #16 and #14 now, THAT'S overkill for a feeder. For bus runs, only if they are short.

Which brings up another point, related to the terminology. With DC and a central control panel (such as using Atlas electrical components), it is common to run the wire that connects to the track all the way back to the control panel. Technically that's a feeder, and they can be quite long, at which point a larger gauge wire is probably recommended - although it is perfectly ok to use smaller wire which is easier to solder to the track for the last foot. With DCC, we don;t usually wire this way, instead a heavy gauge (small number) 'bus' is run around approximately where the track runs. From this, smaller size wire makes short drops from this heavy bus to the track above.
As complexity is added - detection for signal systems, for example, additional elements are introduced - like the 'sub bus' Say a detected section has 5 sets of feeders - a sub bus would branch off the main bus, pass through the detector, and supply the feeders for that section. Since these sub buses would tend to be short, they dont require the same gauge wire as the main bus with which they connect - thus the information on Allan Gartner's site about #12 main bus and #14 sub bus. And even though it drops down a wire size, there will probably be less voltage drop in that sub bus than in the run out to it along the main bus - because it's short, and the load will be less - if your DCC system upports say 10 locos, the main bus will be carrying the full current for all 10 running locos, regardless of where on the layout they are. But the sub bus will only carry the current for the locos in that section - even if you triple-head, that's 3 locos' worth vs. 10.

--Randy

[Anonymous]

I work in pro audio in a large Dallas area church. Some of you guys are useing wire that I use on the job to pump 2000 Watts to subwoofers!! Our subs use 12 guage wire, other than that most every other speaker here is 14 or 16 guage. IMHO that is over kill for MRRing. Like some one else stated that for the current and resistance values of our trains maby 18 to 22 guage will work fine on most layouts. There are always exceptions.

[Don Gibson]

Thank's guy's

You re-affirmed my thinking. Choices of BUSS wire guage should be length related; where FEEDER'S are distribution oriented (more small vs. fewer large one's).

The Practicalities:
Most of today's RTR engine's pull 0.5 amps @12 volt's (some 0.25 amps) 3 engine's might pull 1.5 amps runninng on a single 3' piece of flextrack (2 B.B. Athearn engine's also pull 1.5a). Feeder's every 3 feet - or to each track piece - don't have to be very current oriented, wheras BUSS WIRES SHOULD be larger to offset voltage losses.

This also ignores any possibility of power transfer through potentially intermittant rail joiners - so I'm being conservative. I've also had rail's 'bow' on me, so I don't believe in soldering them, either..

[grayfox1119]

Is the bowing caused by all the rail joiners being soldered, or is it due to the sub base expanding with humidity? Untreated wood loves to expand with moisture. And of course shrinks when it gets very dry. If the rail joiners are on tight as they can be, and you do not solder the joiners to the rails, how can the rails expand any further if they are already butted tight up against the rail joiners?

[jfugate]

QUOTE: Originally posted by On30Shay

Just a question here, Joe.....how is this done using Kato Unitrack?

Hmm, haven't got the foggiest ... never worked with the stuff. Maybe someone on here who has worked with this track can post their insights?

[jfugate]

QUOTE: Originally posted by grayfox1119

Is the bowing caused by all the rail joiners being soldered, or is it due to the sub base expanding with humidity?

We're talking about temperature and humidity effects.

Example (which actually happened to me):
You lay some track in the winter, when it colder and wetter and you leave no gaps at the rail joiners, and or you solder the rail joiners. The track has no room to expand or contract with the benchwork and roadbed as the season changes. The hot dry summer comes and the rails pop right off the ties in that section of track.

Morale of the story:
You need to leave some rail joiners unsoldered and a small gap (say 1/32" to 1/16" or so) to allow the rails to expand and move around slightly as the roadbed and benchwork likewise expands and contracts through the year.

What I do:
I generally don't solder rail joiners, except in these cases:
- section of track is shorter than 1 foot ... I solder it to an adjoining track with feeders on one end
- point end of turnouts ... I solder the point end of turnouts to an adjoining track with a feeder, which follows the general rule of always feeding power to turnouts from the point end
- on curves I solder two sections of flex track together to get an extra long 6 foot section to make the curve flow smoother and have fewer kinks.

Otherwise, each 3 foot rail section gets its own 18 guage feeder, and I don't solder rail joiners.

[Anonymous]

wow. Something you don't know about.

QUOTE: Originally posted by jfugate

QUOTE: Originally posted by On30Shay

Just a question here, Joe.....how is this done using Kato Unitrack?

Hmm, haven't got the foggiest ... never worked with the stuff. Maybe someone on here who has worked with this track can post their insights?

[rrinker]

Don't have any Unitrack, but I don't see why you couldn't drill a small hole in the plastic base next to the rail and run the feeder up through there. Tin the wire first and you should be able to solder without melting the plastic base too bad. A couple of metal track gauges might help to act as a heat sink on either side of the area to be soldered.

--Randy

[rrinker]

I only solder joints for 2 sections of flex to form a curve, so there are no kinks at the joint. I'm not big on soldering turnouts, since if anything is going to fail as far as the track is concerned, it is far more likely to be a turnout with various moving parts than an ordinary piece of flex track. Since I don't solder the turnouts, I can slide the joiners back under the adjoining sections and lift out and repalce the turnout if I ever need to. If I was really paranoid, I'd solder feeders directly to the turnout, but so far I have had no issues having joiners with feeders at all 3 legs of the turnout (insulated frog turnouts - so they can be fed from the frog end with no problem - this won't work on non-insulated turnouts which require gaps at the frog end). I WAS going to power the frogs but not on the Atlas turnouts. The frog metal resists soldering, even though there is an eyelet for it. And the whole is too big for a 00-90 screw, and too small to tap for a 2-56. Maybe a 1-72, but I dont have any on hand and I don't have a tap for them.


--Randy

[grayfox1119]

I am very curious why you guys are not soldering rail connectors. Allowing for expansion is obvious, no issue here, but connectors can fail, I have had it happen, and not in an easy place to get at. So at this point, you need to solder the rail connector because you cannot lift the track without major surgery to replace the connector. Then to allow for expansion, you will need to cut the track and solder a feeder. A lot of work because the connector was never soldered. Granted, this may not happen very often..BUT, once is enough if in the wrong place !

[jfugate]

I don't have a problem with soldering rail joiners to turnouts. Solder will melt again later with an iron and can be wicked up with some solder wick. If necessary, I will cut the joiner with a motor tool.

For the rare times I've had to replace turnouts, it's not been too painful, and I sure appreciate the reliability soldering the point end feeders gives you when it comes to feeding power to the turnout.

[Pruitt]

Model Railroading has it's share of urban myths, and feeders every track section has to be one of the biggest.

If other aspects of railroading and layout construction were overbuilt the way some folks overbuild their electrical distribution networks, all rolling stock would have a graphite-epoxy composite spine to transmit the drawbar pull through the car axis, and the bodies would likewise be graphite-epoxy, this time with a honeycomb core for stiffness. All rotary components, including motors, trucks, gears, etc. would have micro-miniature roller or ball bearings, and compliant-surface thrust bearings on shaft ends. Track would be laid to micro-meter precision. There wouldn't be a scap of wood in benchwork - it would all be fabricated out of steel or aluminum, with each piece cutom-made for its location, milled out to the thousandth of an inch.

Feeders every ten to fifteen feet to compensate for voltage loss in the rails makes sense, so that's what I do. But feeders in every section of track (apparently to make up for the inability to make reliable rail-end connections using rail joiners) is like putting a motor in every single piece of rolling stock because you can't make couplers work right. It works, but it sure is the hard way to get there!

I'm not criticizing those who go the feeder-every-section route; quite the reverse. I admire your tenacity in completing what I see as a very tedious job. Go on out there and solder yourselves silly!

But in 35 years of model railroading, I have never had a properly installed rail joiner fail electrically. (And I've lived in quite a variety of climates, including the dry semi-arid high plains of Wyoming, the seasonal humidity of the northeast, the all-season humidity of Florida and Seattle, and the dry California coastal areas). So I spend the time I would otherwise waste soldering feeder after feeder after feeder after....(eeyargh!) on other mundane tasks such as locomotive tuning and maintenance, rolling stock upkeep, and so forth.

Just a different approach, I guess.

[tsasala]

I have to concur with Mark (mostly). Look at it this way, most computers are being manufactured with 2-5 mil trace widths (2/1000"-5/1000"). Although the power requirements of these signals is not as great as our MRRs, the signal integrity requirements are much more difficult to manage (harmonics into the 50-100GHZ range). The difference between 14 AWG and 22 AWG with respect to signal integrity of DCC signals is miniscule in my opinion. High power is to only reason to use a large gauge and I believe 20 or 18 will handle more than enough current to your rails. Try to jam 250 watts into a 5 mil trace and see what happens.

Like Joe, I solder feeders every three feet or so, and I also solder feeders to all turnouts. That eliminates the rail connectors as an issue. Although a properly installed connector may never fail, I rarely see (or install) a connector properly [:I].

-Tom

[AntonioFP45]

Even if it seems like overkill, I like the idea of using feeders on shorter sections. Maintenence is also a must.

At a club layout, I witnessed sound equipped locomotives hesitate or sputter over certain spots repeatedly. There were long length feeders througout the layout. Many of the solder points appeared oxidized. I don't know if the oxidation was part of the problem, but my overall view is that I'd rather be guilty of "over kill" and encounter few problems than do the bare minimum and have to fight headaches later on.

[grayfox1119]

I did a little research on the resistivity of copper wire, which we use for all our buss wiring. #12 gauge copper wire has a resistance in ohms of 1.62 ohms per 1000 feet.
So one way to look at this is, if you had a layout that used 500 feet for the hot run and 500 feet of course for the neutral return, you would have 1.62 ohms of wire resistance. If you had 5 amps being drawn on this circuit, the wire resistance would cause a voltage drop of 5A X 1.62 Ohms = 8.1Volts. Now you also have IR drops in rail connectors, barrier strip connectors, resistance of NiS track itself, etc.

Most of us do not have 500 feet of buss wire for the hot lead. So using the resistance per foot of .00162 ohms, and a layout that had 40 feet of #12 copper wire TOTAL ( Feed and return ), you would have 0.0648 ohms of buss wire resistance. At 5 amps of load current, the voltage drop due to buss wire resistance would be 0.324 Volts or 324 millivolts. For the same layout length, but using #18 for the main buss, the voltage drop would be 1.302 Volts. For #14 gauge buss wire, the voltage drop would be 0.516 Volts. That is a 4.3% drop in voltage from the 12Volt supply.

Here is the Copper Wire Resistance Spec, so that we can each figure out our IR drops according to what wire size we wi***o use, and what the load current will be:

Guage Ohms/1000 ft @77F Ohms/foot@77F
10 1.02 .00102
12 1.62 .00162
14 2.58 .00258
16 4.09 .00409
18 6.51 .00651
20 10.4 .0104
22 16.5 .0165
24 26.2 .0262
26 41.6 .0416

A more commonly used formula to figure voltage drop is : Voltage drop = K X 2 X Wire Length in feet X Current in Amperes
------------------------------------------------------------------
Wire area in circular mils

K (Specific Resistivity - circular mil/foot) = 11 for copper wire loaded to less than 50% of capacity. K= 12 for copper wire loaded to 50-100% of capacity.

If we assume greater than 50% current loading, say 3.5A, and a length of 40 feet total of #14 copper buss wire, then we have: 12 X 2 X 40 feet X 3.5 amps/ 4112 = 0.817Volts

GUAGE Circular Mils Area
12 6530
14 4110
16 2580
18 1620
20 1020
22 810
24 404
26 320

I hope you find this useful guys

[grayfox1119]

The numbers were well seperated on my reply, but somehow when it was posted, the columns ran together, I apologize for that guys, just look at the numbers by colums with the headers above them and you will be Ok.

[Pruitt]

Keep in mind - most DCC systems put 14+ volts on the outputs, so you can lose a little bit before your locos start to see any degredation due to line or track losses.

[jfugate]

I guess my motto is what someone on here shared recently in another wiring thread:

I'd rather overdo the wiring, than do the wiring over!

[rrinker]

QUOTE: Originally posted by grayfox1119

I am very curious why you guys are not soldering rail connectors. Allowing for expansion is obvious, no issue here, but connectors can fail, I have had it happen, and not in an easy place to get at. So at this point, you need to solder the rail connector because you cannot lift the track without major surgery to replace the connector. Then to allow for expansion, you will need to cut the track and solder a feeder. A lot of work because the connector was never soldered. Granted, this may not happen very often..BUT, once is enough if in the wrong place !

Simple - every piece of track has a feeder to it! So it doesn't really matter if the rail joiner conducts electricity or not. As for physically fail - I've been doing this a lot of years and I never saw that happen. Luckily our models don't beat up on the joints like the prototpye.

--Randy

[rrinker]

Those are the same numbers I used for my calculations for feeder drops. Sort of refutes the idea of needing #16 wire for a foot long feeder to a heavier bus.

Even for the bus wires - even #12 obviously can't run infinitely long without a noticeable loss. Pulling wire heavier than #12 can be a real pain, it doesn't bend nicely, even the stranded forms. What to do? Distribute the boosters, of course. If you are lucky eough to have a 100 foot long side of your layout, the thing to do would be to locate two boosters, each 25 feet from an end. Then the longest bus run you would have would be only 25', which #12 can handle easily enough.

<end>--25' of bus --<booster>--25' of bus --||--25' of bus --<booster>-- 25' of bus --<end>

--Randy

[Anonymous]

Mark,

Not to start a war here, but I personally have experienced correctly installed rail joiners failing electrically. On my on old layout I had to retrofit several feeders to track sections that mysteriously ran fine (for years) until painting and ballasting. The disparities in our experiences may have to do with weathering style. I paint the rails and ballast and dirt stuff in pretty close to the rails. I'm sure glue goes all over causing problems with rail joiners. Retrofitting feeders later is a lot more work than installing a few extra up front.

Currently, I don't feeder every section, but I do go with the every other flex piece rule both in soldering the rails and feeding. I also feed points and solder short sections of track as well.

Just another point of view.

[jfugate]

I too, have had rail joiners fail. I've noticed that in the last few years, my 14 year old layout has many rail joiners that no longer conduct (I sometimes test this out with my electrical meter).

I also am a big believer in ballasting and weathering the track like the prototype ...


(click to enlarge)

But this also means lots of moisture, paint, glue solution, and other gunk all around the track and rail joiners. By feeding each rail section, or soldering the rail joiners (in a few cases, mentioned earlier), I get totally reliable power fed to the track no matter how I weather or otherwise apply gunk to the track.

So I get realistic track that's also reliable -- with no worries!

[Pruitt]

Maybe those who have had joints fail only think you know how to correctly install rail joiners......[;)]

I'm not trying to start a fight. Maybe I've just been very lucky.

In any case, I figure if I get a problem section of track, I'll just add a feeder to that one.

I don't anticipate adding many feeders over the next twenty years.

In 2025, I'll tell you if I was wrong or not!

[Anonymous]

If you really had a long way to go, and didn't have a need/desire for a booster, or huge cable, you could run 2 #12s and get half the drop. There are many solutions to most problems, whatever works for you is a good one!

[grayfox1119]

Mark, I have a Christmas display ( Dept 56 Village ) I'm sure some of you guys know about these from your wives, that I set up for her every holiday season. And of course I have about 75 feet of sectional track on the layout with many curves. I do not solder the rail joiners for obvious reasons..I need to REMOVE this big display every January and store for 10 months. After having troubles with rail joiners after the 3rd year, I know check each joiner for FIT as they puch onto the rails. If it is nice and snug, things work nicely, if it goes on too loosely.....I can usually use needle nose pliers to snug the connector tighter before retrying. That always works. But it taught me a lesson, I will have feeders on every section of track, overkill or not, it is far better to be safe than sorry. And I have no problem soldering, so it is not an issue with me. If someone hates to solder, well.....that is another story.
I am still going to go with #12 solid wire for the buss under the table. I plan to use eyelets to hang the wire, no problem at all pulling/running the wire. I have at least 1000 feet of various lengths of house wiring, #12 gauge, so I have an ample supply. I also have about 500 feet of #14 wire, so that will be used for drops. There will be NO noticeable voltage drops on my layout which will be around the room 18' X 10.5'. I figure I will be busy enough with installing signals, tortoises, etc. I don't need other issues with dead tracks. Just my point of view.

[Pruitt]

I agree - for anything other than a permanent installation there's a chance you may "work" the rail joints. In that case, feeders on every unsoldered section are important.

[jfugate]

Interesting discussion ...

I guess for me, if there is a chance a rail joiner will fail, I don't like gambling. Enough people over the years report that joiners do fail that I don't believe in seeing if I can get lucky.

If I solder feeders to each rail section, track power is guaranteed. I don't have to hope I'm one of the lucky ones.

Guaranteed good performance from day one to the layout's end-of-life is worth the small amount of extra work adding a few more feeders takes.

It's just not that big a deal. This discussion is taking more time than it took me to add the extra feeders! [swg]

[grayfox1119]

But think of all the great info Joe, some people have printed out the whole thread to use as a reference document!! So after all this, I decided to not solder the rail connectors, and to use drops for all rail sections. Seeing I will be using 3 foot sections of flex track for most of the layout, this should be no big deal at all, and by leaving the rail connectors unsoldered, that should give me sufficient spacing for any track expansion. And, I will have a dehumidifier in the train room for days like today, and in winter, the furnace is just outside the train room, so it should be very constand temperature. I also will be applying water sealant to both sides of the plywood table top as an added insurance against swelling in summer and shrinkage in winter.

[Don Gibson]

QUOTE: Originally posted by grayfox1119

Is the bowing caused by all the rail joiners being soldered, or is it due to the sub base expanding with humidity?

Grayfox 1119 - (you asked) NEITHER.

I had expansion wipe out whole straightaway's in a temp and humidity controlled basement , using unsoldered rail joiner's!

Problem was (1)no gap's and (2) rails heating from running. Recommended (NMRA?) practice is leaving a biz card gap every flextrack joint (3 ft.?) - which I didn't do. There was no place for it to go.

My problem with soldering rail joiner's is metal rails expand with temperature, temperture increases from wheel's arc-ing*, and 'soldering' joint's negates safety gap's by mechanically linking rails together.

I know there are those that have never had expansion problem's. I hope the never do.
I'll bet some have never had a fire that burned down their home, too, but carry fire insurance.

* Run an engine lashup in total darkness, sometime.

[grayfox1119]

Hi Don, that was my whole point about rail connectors...they are tight when they are installed to begin with, so how can they expand? It has zero to do with soldering the joiners, because whether they are soldered or not, there is far too little room to expand even unsoldered. Soldering just takes that "lttle bit" of space away.
I agree with you, that is why I am NOT going to solder the rail connectors, except at the turnouts. And I WILL be leaving gaps for sure.