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!

Feeders: how long without significant loss

1528 views
29 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    June 2007
  • 7,868 posts
Feeders: how long without significant loss
Posted by riogrande5761 on Tuesday, October 13, 2020 8:28 AM

How long can 20 AWG feeders be without significant loss?

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    September 2003
  • 13,485 posts
Posted by Overmod on Tuesday, October 13, 2020 8:44 AM

You'll need to determine what you consider 'significant loss' to be electrically.  I presume you have DCC, say about 14.5 nominal, and would use something like 5A as a reference load.

The resources I'm familiar with 'advise' that 20ga feeders (I presume in solid wire) be 'as short as possible' and not longer than ~12" from bus to soldered attachment point on the track, with the connections at both ends being considerably larger in area than the gauge of the wire.  That would be the "answer" I'd give, but it isn't the answer I think you're asking for.

In practice the run can be much longer than a foot or two.  Use this calculator to see where the voltage drop becomes too great for your purpose... and design your feeders to be shorter than that  Change the variables in the URL to what you want if the ones you want to use are different.

https://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=33.31&voltage=14.5&phase=dc&noofconductor=1&distance=10&distanceunit=feet&amperes=5&x=0&y=0

The size of your bus should remain heavy; I wouldn't use smaller than 14ga just on general principles. 

  • Member since
    June 2007
  • 7,868 posts
Posted by riogrande5761 on Tuesday, October 13, 2020 8:59 AM

Understood on the main bus.  That isn't the droid i was looking for.

For reference purposes, what is the distance a 20 AWG feeder can go with a 3A reference load and not experience a drop of more than 1/2 volt?

https://tonystrains.com/news/wire-sizes-to-use-in-dcc/

I did see this from Tony's Trains but it isn't clear what the yellow blocked area in the chart stands for.

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    September 2003
  • 13,485 posts
Posted by Overmod on Tuesday, October 13, 2020 9:05 AM

riogrande5761
For reference purposes, what is the distance a 20 AWG feeder can go with a 3A reference load and not experience a drop of more than 1/2 volt?

ASSuming 14 to 14.5V, 8'2".

Note that the calculator assumes a pair, not just a single wire.  If you are going to ground through something like a common rail, the resistance may be effectively higher.  In that case you'd likely have to measure the actual drop with a meter, then add this to the feeder resistance and do the calculation.

For reference, one inch = 0.08333... foot, with the terminal 3 repeating.  (If anyone knows how to put a bar over a number on the forums, let him speak...)

I'm presuming the yellow means 'marginal'.  It would presumably be red if 'not recommended at all'.  If you compare the number from the (presumably more precise) calculator, you can compare the yellow values with the indicated length to produce an 0.5V nominal drop.

  • Member since
    June 2007
  • 7,868 posts
Posted by riogrande5761 on Tuesday, October 13, 2020 10:49 AM

Thanks.  The application here is staging where staging tracks will only have one train on each one so I am providing sub-busses (so to speak), and each will be electrically isolated and can be disconnected from the main bus by way of a switch.  The staging tracks range from about 17 - 24 feet in length so it appears I can gang drops together to a larger bus as long as they are 8 feet or less to the buss.

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    January 2017
  • From: Southern Florida Gulf Coast
  • 9,618 posts
Posted by SeeYou190 on Tuesday, October 13, 2020 11:05 AM

Rio:

I use 22 AWG feeder wires soldered to the bottoms of my rails. I allow these to be 24 inches long to reach distribution blocks. I have noticed no difference in locomotive operation from one block into another with different length feeders.

I hope this helps.

-Kevin 

Wink Happily modeling my STRATTON & GILLETTE RAILROAD. A Class A line located in a personal fantasy world of semi-plausible nonsense on Tuesday, August 3rd, 1954.

  • Member since
    February 2002
  • From: Reading, PA
  • 28,926 posts
Posted by rrinker on Tuesday, October 13, 2020 12:19 PM

The yellow area is "not recommended" 

I would use something a little heavier to run from a toggle on a control panel out to the tracks, like #18. I think #18 is a bit large to solder to HO rail, so I would use #20 or #22 feeders - depending on how long each staging track is, you can run multiple feeders from the #18 coming from the control panel.

                                              --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
    June 2007
  • 7,868 posts
Posted by riogrande5761 on Tuesday, October 13, 2020 12:48 PM

The drops themselves are about 11 inches long of 20 AWG wire.  I wanted to use the thickest which was still workable.

Since this is staging, I'm making no effort to hide the drops near the track.

If what Overmod says is true, I can tie the drops together no more than 8' length using the 20 AWG and connect them to the sub-bus of larger wire which will go to a switch, which will tie-in to a 14 AWG main bus.

The actual lengths of the isolated staging tracks range from about 18 to 25 feet long approximately.

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 22,268 posts
Posted by selector on Tuesday, October 13, 2020 1:07 PM

I have used longish feeders of 22 gauge wire to feed an off-layout five track staging yard.  I had a bridge about six feet long to a shelf on a bookcase, on which was the five-track ladder, the longest about 6 feet.  That's a total of 12 feet of track on the one ladder, the others diminishing, all fed with one pair of 3' feeders at the proximal end to the layout.  A test confirmed that all ladders provided enough signal/voltage to trip the shorts circuitry.

  • Member since
    December 2008
  • From: In the heart of Georgia
  • 3,800 posts
Posted by Doughless on Tuesday, October 13, 2020 1:22 PM

I'm building my layout now.  All feeders are 22 AWG soldered to rail joiners prior to installing the joiners. 

When in the test faze, I simply placed one feeder pair on one end of my 18 linear feet of track where the interchange/yard is.  That part of the layout has  72 inch passing siding and several long spur tracks, all connected only with rail joiners...no solder.

That one feeder pair was probably 24 inches long, which is what it took to run from the temporary locaton of the command station to the end of the track.

I was able to run two DCC Sound locomotives with no problems anywhere on that 18 foot section.  I didn't try more because of amperage limitations of my NCE Power cab.  (It could probably run 3, but I have no need for it to do that so I did not try)

I have since installed feeders on all three legs of each turnout.  The longest stretch of track without feeders is planned at 10 feet.  Feeders themselves are no longer than 12 inches.  Every joint will be soldered, including turnouts.

I expect no problems.

- Douglas

  • Member since
    December 2008
  • From: In the heart of Georgia
  • 3,800 posts
Posted by Doughless on Tuesday, October 13, 2020 1:32 PM

nvm

- Douglas

  • Member since
    October 2006
  • From: Western, MA
  • 8,462 posts
Posted by richg1998 on Tuesday, October 13, 2020 2:00 PM

Some years ago our club used six to seven inches of #24 solid. We changed from DC to DCC. I was under the layout doing the switching with a #14 buss. We had the NCE Five amp system. No issues.

The DCC amp meter never showed more than four amps. Maybe ten HO sound locos.

Edit. We measure about 13.6 vac at different points. Verified by a Scope. Handlaid track.

Rich

If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.

  • Member since
    June 2020
  • 801 posts
Posted by Lastspikemike on Tuesday, October 13, 2020 2:36 PM

Put another way, how much voltage drop would make an operational difference?

It seems to me that gearing may be more limiting than voltage, practically speaking. 

We have a MRC 780 and a 760. Locomotives are more sprightly when powered by the "24v" maximum voltage 760 but I do not see them running any faster. Pulling power might be the difference? But just how often does anyone run a train at full speed or needing maximum power?

At what point does voltage drop with distance actually make a practical difference?

Alyth Yard

Canada

  • Member since
    February 2001
  • From: Wyoming, where men are men, and sheep are nervous!
  • 2,494 posts
Posted by Pruitt on Tuesday, October 13, 2020 11:12 PM

From a practicality standpoint, you can go with much longer feeders than the table on Tony's Train Exchange website shows. The voltage drops he's showing is if you're pulling the full amperage through that one set of feeders. You'll probably be doing that only if you have a dead short in that section, and even then the current load will be shared by however many sets of feeders are in the shorted section.

Allowing for an average loco pulling 1/2 amp (for simplicity), you'll get a 1/2 voltage drop at 14.5 volts by running the current through 49 feet of 20 ga wire. At 5A the voltage drop is about 5 volts.

Not that I recommend such long feeders. I use 12 ga stranded buss wire that feeds 22 ga feeders that typically aren't longer than about a foot.

  • Member since
    September 2003
  • 13,485 posts
Posted by Overmod on Wednesday, October 14, 2020 6:20 AM

Here's a page I consider a useful resource on practical wire sizing:

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

I invite others with experience to comment in detail on any part of what he says.

Note that DCC being actively used for control has a slight inherent voltage drop from 'nominal' due to the data modulation.  This was as I recall a reason for the ~2V nominal increase over legacy "12-volt DC".  I would tend to agree with the effect being relatively slight at low peak current (amp) draw, as I think recovery of the data signal from DCC is robust even down to large amounts of voltage sag.  The issue I'd see cropping up first with respect to small feeders is that transient voltage might drop below that at which some function on the decoder, perhaps sound, would drop out and reboot itself ... say at even a short sag below 7.2V.  Cure for this, as noted in other threads, is a suitable keepalive ... which when present in equipment ought to be recognized as potentially allowing longer 20ga feeders.  

 

  • Member since
    July 2009
  • From: somerset, nj
  • 3,288 posts
Posted by gregc on Wednesday, October 14, 2020 6:48 AM

i believe the 2V increase for DCC track voltage is to make up for the 2 diode drops in the decoder

presumably the reason for having a feeder on at least every other rail joiner and that rail joiner soldered to the rail is to guard against a faulty non-soldered rail joiner.    this implies that the all the current to a loco may have to flow thru a single feeder and rail instead of flowing from both ends of the rail

measurements don't differ much from standard measurements

while rail is much thicker compared to wire, code 100, 83 an d70 nickle-silver rails has an ~equivalent resistance of 24, 26 and 28g wire.    with a faulty rail joiner, code 70 rail looks like 3' of 28g wire.    there's a ~0.15V drop/ft across a pair of 28g conductors, ~0.45 across 3'.

while bus wire that carries current to more than one loco should be thicker, feeders need not be when compared to the rail itself.

i've used 4" pieces of bare copper 14g wire which are easier to hold in place against the rail while holding it from under the benchwork

 

 

greg - Philadelphia & Reading / Reading

  • Member since
    July 2006
  • From: west coast
  • 5,510 posts
Posted by rrebell on Wednesday, October 14, 2020 8:21 AM

A lot of you are not looking at real world numbers, sure if you have one set of feeders you can have voltage drops but we typicaly use many and the places with less are ussually smaller lengths of track so you don't have multiple engines drawing power. I mean you can look at a lot of tecnical losses in voltage etc, but for our uses they are ussually moot.

  • Member since
    February 2002
  • From: Reading, PA
  • 28,926 posts
Posted by rrinker on Wednesday, October 14, 2020 9:00 AM

 But that's why DCC track voltage starts a bit higher - 14-15V is typical for HO. Not 12V. So after the diode drop, you have 12V or better available to the motor drive. With adequate wiring, I don't see such a sag as even happening - I've seen some big layouts, few home layouts are bigger then Ken McCorry's, and his trains run around the entire layout just fine. Or club layout is fairly large, with #12 bus lines there are no problems with control or trains slowing anywhere other than those oft-mentioned fitter sections between some older modules, which have only rail joiners for power - a prime exampel of relying on unsoldered rail joiners to power a section of track is a bad idea.

                               --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
    June 2020
  • 801 posts
Posted by Lastspikemike on Wednesday, October 14, 2020 9:13 AM

I noticed the very small gauge wires inside the locomotives and concluded that voltage drop from feeder wires likely isn't much of  an issue practically. 

We used 16 gauge for the main return wires (white or red) and 18 gauge for the main power wires (black) because with common rail or bus wiring through block switches only the return wire is drawing full amperage. The block feed wires are all operating in parallel (and yes, we used 16 gauge black wire to feed the block switches from the power pack). Feeders become extensions of locomotive internal pick up and motor connection wires.  

Those decoder wires are very high gauge.

Highest resistance may turn out to be at the pickup points of the wheels on the rails.

Alyth Yard

Canada

  • Member since
    January 2017
  • From: Southern Florida Gulf Coast
  • 9,618 posts
Posted by SeeYou190 on Wednesday, October 14, 2020 9:17 AM

rrinker
A prime example of relying on unsoldered rail joiners to power a section of track is a bad idea.

I could not agree more. Over time unsoldered rail joints have become a source of problems for me on my previous layouts.

-Kevin

Wink Happily modeling my STRATTON & GILLETTE RAILROAD. A Class A line located in a personal fantasy world of semi-plausible nonsense on Tuesday, August 3rd, 1954.

  • Member since
    September 2003
  • 13,485 posts
Posted by Overmod on Wednesday, October 14, 2020 9:34 AM

The wire 'inside' locomotives is indeed short, and indeed fairly fine-gauge.  It is also extremely short by 'feeder' standards, and is right at the 'last mile' to the devices that are to be powered.  The purpose of adequate feeders is to limit the voltage drop up to the wheel contact, wipers, and small wire that necessarily (in most actual locomotives) constitute sources of loss.  Think of it the same as good steam jacketing and internal streamlining reducing temperature and throttling loss to steam-locomotive cylinders.  

High resistance at wheel contacts almost immediately invokes a risk of microarcing.  That is always and everywhere a bad thing.  Technically that may be worse with higher available voltage, which may need to be borne in mind when assessing really good high-amperage feeder systems...

The two most sensible approaches (and I suspect they can be combined on a particular layout) are feeders to every rail joiner with that joiner then soldered to both adjacent rail ends with a lower-melting solder, and feeders to the midpoint of each rail segment, whether or not that segment has rail joiners or just gaps at its ends.  As I understand the latter, even if you solder the rail joints you still want a feeder to the midpoint of each stick of rail, going to an adequate bus through a good soldered areal contact.  gregc and others are indicating that the nominal voltage drop through rails with good soldered joiner connection is relatively low, so there certainly will be cases where just having some long strings of soldered rail without feeders will be adequate, but it still seems rational to me to take the time to drop a feeder from every rail, as a one-time action that guarantees the 'best' reasonable electrical supply right from the time track is laid.  (See our good threads on track fixation and benchwork for advice on holding either soldered or unsoldered track effectively once it has been 'prepared'... and some of the notes on painting the rail and prepping the railhead after painting...)

  • Member since
    June 2007
  • 7,868 posts
Posted by riogrande5761 on Wednesday, October 14, 2020 12:56 PM

Some of my feeder runs were a little on the long side (over 8 feet) with 20 awg so I've replaced them with 18 awg.  The rest are about 6' or less 20 awg but will be tied in to the main bus with 18 awg via on/off switches.

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    February 2002
  • From: Reading, PA
  • 28,926 posts
Posted by rrinker on Wednesday, October 14, 2020 8:21 PM

 Last two layouts I built, the feeders were soldered to rail joiners - do not EVER buy the Atlas terminal joiners, for the price they charge for those, you can make a couple dozen pair, and since you can solder them at the workbench, there's no danger of melting ties. ANyway, I used terminal feeders at EVERY track joint. I only soldered every other one though, so it was flex track - soldered joiners - flex track - joiners - flex track - soldered joiners - flex track, etc. ANd All 3 legs of every turnout (except where I needed insulated joines to divide power districts) got wired joiners - none soldered.

 This was highly reliable, even after painting the rails. And I wasn't shy with the paint around joiners - I wanted to hide all the shiny.

 The traick is - I have a handful of joiners I use to test fit things - they are well used and pretty loose fit. However, all ther terminal joienrs I made up were fresh out of the pack joiners. Only installed after the track was fitted and cut to length. So they were very tight. Slid on once, and done. The soldered sections, I soldered at the bench.

 If any of the non-soldered ones failed, I never knew it, because I never had any power issues. There was effectively a soldered joint with power feeds every 3 feet, with an extra 'helper' one in between that wasn't soldered to the track. And the turnouts (Atlas) were extremely reliable with feeds at all 3 legs. Even without powering the frogs, nothing ever stalled.

 I assembly lined it. Cut lengths of wire from my roll of feeder wire. Strip the end to solder to the joiner of all pieces, and bent the stripped end over at 90 degrees. Then soldered all the prepped wire to new joiners. With a built up supply of a dozen or more pairs, I could just keep on laying track without having to go back and solder things. Can make a couple dozen sets in an hour of bench time.

                                              --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
    March 2016
  • 1,124 posts
Posted by PRR8259 on Wednesday, October 14, 2020 9:52 PM

Practically speaking if on your home layout you only run an engine or two at a time, and they aren't drawing much amperage at all, the voltage drop is much less, even in the provided calculator (I got it down to about 3%).  I have a decent mainline run at about 81 feet, and the maximum wire length is about 30' to any of the three total drops.  I use Kato track and none of the joints are soldered anywhere.  It's just a simple folded dogbone, and most of the time for plain dc operation it functions just fine.  Do I wish I'd put in another feeder or two?  Perhaps, but it works.  It even functions ok in dcc operation too.  I think it's because the NCE system tells me that even maxed out some of the (Genesis) engines are only actually drawing 0.2 amps.  At that kind of amperage, the voltage drop is a lesser concern.  I do have to keep my track clean. The issues occur if/when there are any dark, grimy deposits on a track section or particularly on the rail surface at a joint.  As long as my track is clean, things run very well, whether dc or dcc.

If I were running more units, maybe I'd see more of a difference, but a single unit or steamer will pull most of my trains.

  • Member since
    September 2003
  • 10,348 posts
Posted by mlehman on Friday, October 16, 2020 4:02 AM

My experience is that you should have no problem using 20 AWG on those staging tracks. I tend to use stranded 22 AWG (or did before Radio Shack closed up locally_ and have had no problems with it. I assume the grade is close to zero in staging, so little added load because of that. So long as you hook up eqch track to feed from a 18 AWG source, it will be fine.

I don't want to disparage anyone's formulas but they tend to be overkill for what we do as model railroaders. Such formulas are good for critical circuit design in order to precisely define what is needed so that circuits won't be degraded by voltage drop nor will excess costs for overly heavy comp[onents be borne, which can confound certain aspects of a design.

In contrast our locomotives operate at a wide range of values and only rarely at full throttle. There's almost always plenty of reserve power should it be needed. Yeah, some voltage drop may be present, but get out the hair splitter if you think it matters much.

Mike Lehman

Urbana, IL

  • Member since
    February 2002
  • From: Reading, PA
  • 28,926 posts
Posted by rrinker on Friday, October 16, 2020 8:22 AM

 The formula includes the factor of current draw - and where the charts come from and what people often do is look at their system capability, say 5 amps, and plug that in. But that's the MAXIMUM current - on a given stretch of track, the maximum current would be determined by the current draw of the locomotives and how many can fit in the area. That number is usually significantly less than the max capability of the booster. This means the voltage drop over the same length of the same size wire is less.

Where this matters more is in the case of a short cause by derailment or some other reason. Then, the fully current of the system, or the current setting of any circuit breaker (if your track section has a circuit breaker set to 3.5 amps between it and the booster, you're not going to get the 5 or 10 amps of the booster, you are going to get a max of 3.5 amps), flows through that wire. If the wiring introduces too much resistance, then the current may never reach the trip level of the breaker. A breaker set at 3.5 amps will allow 3 amps all day long without tripping. 3 amps at 15 volts is 45 watts, a significant amount of heat. This is why the quarter test is important - you set (do not press) a quarter, or similar size coin if outside the US, on the rails all over the layout. In each case the breaker should trip. If it does not, you have inadequate feeders to the area. If it does always trip the breaker, you cna be reasonably confident that any derailment or other issue will also cause the breaker to trip, instead of flowing considerable current through the 'short' which isn't 0 ohms.

 And the idea is to have multiple feeders to a given length of rail - so under normal circumstances, you may have 2 sets or more of feeders powering the track a loco is setting one, so your feeders may be 2x #22, instead of just that single length of #22. 

                                  --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
    June 2007
  • 7,868 posts
Posted by riogrande5761 on Friday, October 16, 2020 10:24 AM

mlehman
My experience is that you should have no problem using 20 AWG on those staging tracks. I tend to use stranded 22 AWG (or did before Radio Shack closed up locally_ and have had no problems with it. I assume the grade is close to zero in staging, so little added load because of that. So long as you hook up eqch track to feed from a 18 AWG source, it will be fine.

All the feeders are approx 11 inches long of 20 AWG from the track down to under the staging yard.  Those feeders on each staging track are tied together with longer 20 AWG wires of a distance of less than 8 feet to 18 AWG sub-bus.  Any distance longer than that is tied to the short 20 AWG drops with 18 AWG. 

Each staging track will be tied-in via connectors to an 18 AWG wire to a on-off switch which will tie them all to a 14 AWG bus.  The bus will go from a PSX1 breaker to the booster or command station.  

I just wanted to make sure what radiates out from the 18 AWG to the tracks using 20 AWG are not too long so there isn't a significant drop in voltage at 3 amps or less.   From what Overmod commented above, the distance of 20 AWG should be ok if it is less than 8 feet.

Rio Grande.  The Action Road  - Focus 1977-1983

  • Member since
    September 2003
  • 13,485 posts
Posted by Overmod on Friday, October 16, 2020 11:30 AM

riogrande5761
From what Overmod commented above, the distance of 20 AWG should be ok if it is less than 8 feet.

And that as noted is extremely conservative under most operating conditions.

  • Member since
    September 2010
  • 262 posts
Posted by Mister Mikado on Saturday, October 17, 2020 4:04 PM

Online voltage drop calculator.  I set it for copper, pick wire size, 12 volts, DC, single conductor.  I put in .25 amps for a modern loco, is that correct?  At that current, even 18 gauge can go far without a big drop.  There's lots of technical info lower down. -Rob

https://www.calculator.net/voltage-drop-calculator.html

  • Member since
    February 2002
  • From: Reading, PA
  • 28,926 posts
Posted by rrinker on Saturday, October 17, 2020 6:30 PM

 I do think that calculator is assuming a single wire - so for your complete circuit, you need to double the length entered to get the real value. But yes, at a mere 1/4 amp, the voltage drop of even #18 is neglible over a very long distance. If you are only ever running 1 loco on your layout, you can make such assumptions. But if you run more - and more importantly, if there is a short, you need to take into account the full current capacity of your DCC system, or the circuit breaker you have feeding that section of track. Too small a wire can add enough resistence that even a pair of pliers across the rails doesn't draw enough current to shut down the booster.

 If you are using a system that puts out say 2.5 amps maximum, it's pointless to do the calculations with 5 amps as the load - it will never get that high. Or using more than 5 amps if you have a 5 amp booster, etc.

                                         --Randy


Modeling the Reading Railroad in the 1950's

 

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

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!

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!