Is there some sort of rule of thumb for how many times to attach power to the tracks? Length, number of trains, lights on caboose, etc.? These bus lines seem overkill, but I assume there are good reasons.
There are good reasons, mainly to keep from having voltage drops across a layout, especially as layouts get bigger.
If you have a 4x8, the sure, a bus is probably overkill. Just like anything, it depends on size and scope of a layout.
Here is the big picture. DCC powers all the track all the time. As layouts get bigger, to provide sufficient power for all engines that may need to run and to make trouble shooting easier if there is a short, you divide the layout into power districts and have more than one DCC power booster. All track still gets a signal but boosters make sure that there is enough amperage to power all engines on a layout.
A single power booster may typically be 5 amps for systems like Digitrax or NCE. Depending on how much load an engine puts on the system, you can only run so many engines before overloading that 5 amp booster. So by breaking the layout into power districts, any district would hopefully have sufficient power for what ever engines are in it. Obvously it takes some planning on that front.
As for the bus, each district would have it's own bus and you provide feeders to the rail frequently from the bus. That ensures no dead zones and even power to the track. Otherwise you would be relying on the rail joiners to reliabley feed power to the rails from track section to track section.
I am not sure if there is a rule of thumb, but I would think for flex track, having a feeder to every other joint would be a good minimum. Every other joint would mean every piece of flex will have it's own feeder going to it.
Rather than solder rail joiners, I leave them unsoldered to allow the rail to breath, or expand and contract. I solder the feeder to the rail joiner instead.
Rio Grande. The Action Road - Focus 1977-1983
There is nothing about bus bars in your question.
One rule of thumb is a piece of wire or a soldered connection to every piece of rail. Because of expansion and contraction of the bench work, soldering every joiner is a bad idea. The other rule of thumb is a feeder every 6 feet. I would also suggest not soldering turnouts. It makes them easier to salvage if you start over.
When you ask about numbers of trains, lighted cars, multiple engine consisits and sound, you are really asking about how many amps do I need to run trains and how many boosters, not the quality/quantity of wiring. The introductory books seem to think that boosters are practically free and you should use as many as possible.
Eventually you are going to get sucked into the "power frogs or not" question.
A good site is http://wiringfordcc.com/ He gets too wound up about using auto bulbs for circuit protection. They do not provide adequate protection to prevent a decoder from burning up, so skip over that part.
Henry
COB Potomac & Northern
Shenandoah Valley
Not bus bars, but the power bus.
Yes, there are good reasons. See https://dccwiki.com/Power_Bus for more information.
As nickel silver has a lot more resistance than copper, the purpose is to reduce the impedance in the circuit as much as possible, to minimize voltage losses over distance. Inadequate wiring causes problems, notably it can interfere with the short circuit protection.
For example, C100 rail can have 10 times the impedance of an equal length of 14AWG copper wire.
https://dccwiki.com/Rail_Size has impedance values for various rail profiles.
This is why the 'quarter test' is a good gauge of the robustness of the power distribution for a DCC-operated layout. Placing a quarter or screw driver blade across both rails shorts the path of the electricity and should cause the system to detect the fault and to cut power to the rails in order to prevent damage to expensive things...................like decoders.
The idea is, if the quarter does not result in the system detecting the short, it has too much signal-to-noise, or rather noise-to-signal, and can't see the problem. This is almost always caused by poor voltage where the fault is. The convention is to simply alter, or improve, the voltage there, and that is often as simple as providing another pair of feeder wires.
But, all of this is near moot if the power has so long to go, or if the resistance is high further upstream. A larger wire bus helps to keep the 'pressure' of voltage high further down the rail system, and from there you can simply add feeders up to the rails with minimal voltage loss. This logic works until even the bus can't cope with the distance due to internal resistance, and at that point one must consider what is called a 'booster'. From there, if the rail system is so vast, you would run yet another bus, and from that bus more feeders.
selector The idea is, if the quarter does not result in the system detecting the short, it has too much signal-to-noise, or rather noise-to-signal, and can't see the problem. This is almost always caused by poor voltage where the fault is. The convention is to simply alter, or improve, the voltage there, and that is often as simple as providing another pair of feeder wires.
No. Signal to noise has nothing to do with it. As the signal is pure digital, it is either +14 or zero volts, (H0) so it cannot get lost in the noise.
Circuit protection works on rate of change. Too much impedance, and the time constant created by the inductance and resistance will interfere by impeding that change. So the booster will not see a sudden rush of current and shut down.
It doesn't care about the amount of current, it cares about a sudden change. Which sure beats a 3A current flowing through a short melting something because it needed to see 5A to trip.
The power bus is not overkill. For rule of thumb attach a feeder between the bus and each piece of track, dont depend on the track connector to carry the current between each piece of track. This also allows you to leave the track connectors unsoldered and allow for expansion / contraction movement.
Having said that, not everyone agrees and some have other schools of thought.
You won't go wrong with using 14 ga. for the bus and 20 ga. or so for the feeders, and keep the feeders short (12" or so). Make the bus insufficient and you will only have grief.
This is great information everyone. I'm starting with basically a 17' long loop with two passing sidings and two engines. Do I just run a 12 gauge two wire buss from end to end and tap into the track at intervals to start out? I plan on the 2w NCE starter package.
Thanks.
14AWG should be just fine for the power bus and 18-20AWG for your track feeders. How many feeders primarily depends on how many sections of track you will have over that 17' loop.
Will you be soldering to each section of track, or to the rail joiners? If the rail joiners then every other section of track should suffice. You need to allow for expansion and contraction of your track. Otherwise, your track could buckle on you.
FYI: The NCE Power Cab is a 2A system rather than 2W. And bus only has one "s".
Tom
https://tstage9.wixsite.com/nyc-modeling
Time...It marches on...without ever turning around to see if anyone is even keeping in step.
ObsessedDo I just run a 12 gauge two wire buss from end to end and tap into the track at intervals to start out?
If you just run a single bus straight down the center of the benchwork you will have to use longer feeders than is ideal. The feeders will have to be heavier, i.e. 16 ga or so. The problem with that size of wire is that it will show where it is soldered to the outsides of the rails. Some people care about that, others don't.
A better solution would be to have the bus follow the track around the loop. Then your feeders will be shorter so you can use 20 ga. or 22 ga., and there will be a lot less wire running back and forth across the bottom of the layout. Yes, there will be more bus wire, but you could end up with 30 or more pairs of feeder wires. Do you want them 6" long or 26"? Depending on how far you go with powered options like signals, frogs and lighting, you could end up with a lot of additional wire under the layout, so anything you can do to keep it simpler is good.
Before you go ahead and install the main bus, have a look at your track plan to identify any reverse loops. That is where the left side rail eventually comes in contact with the right side rail causing a short. Draw your track plan with both rails shown, and then take blue and red markers and start following the rails. If you come to a point where the blue meets the red that shows you that there will be a short, and the layout won't run with a short. If all you want is a straight oval there won't be any shorts, but if you put a track in that allows the train to reverse direction then you will have created a short.
Remember to colour code your wires!
By the way, if you want to make the title more clear, you can edit the title of the thread if you go back to your first post and click on the 'edit' button on the bottom right of the page. As somebody mentioned, the word 'bar' is a bit misleading. You can have 'bus bars' in your 'bus' but they are not the same thing.
Dave
I'm just a dude with a bad back having a lot of fun with model trains, and finally building a layout!
I second Henry's recommendation about consulting Allan Gartner's "Wiring for DCC" web site. There is also a publication "Basic DCC Wiring" brought to you by and for sale by Kalmbach, the folks that bring you MRR magazine.
Bear "It's all about having fun."
I relied on the Kalmbach "The DCC Guide" booklet that advises to stay within a 1/2 volt drop, from the DCC source to the locos. So bus and feeder wires should be used that take less than that drop if all the DCC current is going to one spot. That can occur when you have multi locos all in one place. Typical, recent HO locos run at maybe 1/2A - 3/4A at max load, so you may have all your 2A DCC output going to one spot on the layout where your 3-4 F7 A-B-B-A lashup is running. So the wiring to each spot should be able to handle the full DCC system (or booster if a smaller district) output within the maximum 1/2 volt drop desired.
You have a 2A system but might for some reason upgrade, so while at it, I suggest you provide bus wires that can handle 5A, but not jump past that. That will not require monster AWG bus wires.
A table in the DCC Guide indicates max "bus" length (at 5 amps) of 25' for 16AWG, 40' for 14AWG and 63' for 12AWG. Note that bus wires can be split by tee-ing off in a second direction, a "wye" configuration, or they can be run in a near circle (no need to connect the ends). What matters is the length from the DCC system (booster) output to any given spot. For instance, a 5A DCC system with a 40' straight bus running out from the booster can have a 14AWG bus per that guideline. If the the DCC booster is connected to the middle of a 40' bus, it would be ok to use 16AWG bus wire as the current would only travel a max 20' from the booster.
They recommend feeders of 20-22AWG but do not specify the length. Most folks target for 2' or less.
As an example, I have a modest 5' x 10' layout with a 5A NCE system that only needed 16AWG (I used stranded) bus, and I used (~2' max) 22AWG solid feeders, max 6' separation. I followed the advice of feeders to every "piece" of track, including turnout frogs. Many soldered rail joiners allowed me to have feeders each 6', with extra feeders to odd smaller pieces that have unsoldered joiners. Checked that all sections passed the "quarter test". Pretty consistent with all the suggestions above.
Taking the trouble to follow the guidelines is worth it and not much extra work.
Paul
Modeling HO with a transition era UP bent
betamax No. Signal to noise has nothing to do with it. As the signal is pure digital, it is either +14 or zero volts, (H0) so it cannot get lost in the noise.
HO scale DCC systems are ~14v RMS (approx 15V peaks / 30v peak-to-peak).
And yes, the signal can get degraded even though it's "digital".
"Rate" of the change doesn't really matter. If as you say, you're only pulling 3A through a 5A breaker, it'll never break, regardless of how slow or fast you start drawing current, until you pass that 5A threshold (and by enough that the fuse blows fast enough to protect things -- drawing 5.1A through a 5A breaker may not be enough to trip it quickly ...)
Note too that you have to consider fast-blow vs. time-delayed fuses. The latter takes longer (sometimes considerably so) to blow, so that you can deal with high-current draws at startup of inductive loads (e.g. an electric motor), that drop off considerably once "startup" is finished.
-Dan
Builder of Bowser steam! Railimages Site
Thanks hon30critter. That is a great explanation of why to run the bus along the length of the loop. I think I get it now. I just bought the Model Railroader, "DCC Projects and Applications" book, so I am learning a lot.
hon30critterIf you just run a single bus straight down the center of the benchwork you will have to use longer feeders than is ideal. The feeders will have to be heavier, i.e. 16 ga or so.
Sorry, but this is not necessarily correct. A 17' loop sounds to me like a tabletop layout which is usually less than 5' wide. If you look at this table from DCCWIKI, you can see that a bus down the middle, which would result in 24" - 30" feeders, is well within reach of 20 or even 22 awg.
What Dave suggests is best pratice for larger around the room layouts, and will certainly not hurt anything on a tabletop, but it is not necessary.
You could even run 16 awg from your booster to a pair of terminal blocks under the center of the table, then run all your feeders from there.
I have the right to remain silent. By posting here I have given up that right and accept that anything I say can and will be used as evidence to critique me.
You should wire the track with a power bus following the track. If you wish you can put the booster in the middle and follow the track in both directions.
At some point, usually the farthest, the bus will come to an end, and the track should be gapped so a loop isn't formed. (Same for the power bus.) Signals crashing into each other can be problematic under the right conditions.
NeO6874 "Rate" of the change doesn't really matter. If as you say, you're only pulling 3A through a 5A breaker, it'll never break, regardless of how slow or fast you start drawing current, until you pass that 5A threshold (and by enough that the fuse blows fast enough to protect things -- drawing 5.1A through a 5A breaker may not be enough to trip it quickly ...) Note too that you have to consider fast-blow vs. time-delayed fuses. The latter takes longer (sometimes considerably so) to blow, so that you can deal with high-current draws at startup of inductive loads (e.g. an electric motor), that drop off considerably once "startup" is finished.
carl425Sorry, but this is not necessarily correct. A 17' loop sounds to me like a tabletop layout which is usually less than 5' wide.
carl425:
I guess it depends on how you read "...17' loop...". I interpreted the 17' to be the length of the layout, which in hindsight might not be correct. If it is just a small oval then you are right.
Thanks Betamax for the tip about gapping at the end of the loop. I know I would have missed that, yet it makes so much sense.
Yum, good data hon3critter. Yes, it is relatively a bunch of 4x8s making a large tabletop layout. Boring as all get out, but appropriate for what the display is all about.
I have a 55 sf folded dogbone walkin-U layout with a mainline of about 50 ft. It has two power districts gapped with plastic rail joiners. I ran a 14 AWG stranded bus underneath the main and at the end, each bus feeds two sets of feeders (18 AWG single strand) just "before" the gap. Never had any problems.
There's a lot of half right info in here, but most of it isn't harmful, so it's ok.
A larger gauge bus allows for less voltage drop through the wires before the power/signal reach the locomotive. Enough voltage drop and the decoder can stop working (part of the power down the rails powers the logic in the decoder), or can prevent the decoder from delivering the required/desired power to the motor.
The quarter test allows for a rough check of voltage drop, if a short (quarter) across the rails is not sensed by the control station, then there was so much drop in the trip to the point of the short that the short circuit current wasn't required to drop all of the output voltage.
I will almost disagree with Betamax's comment on gapping the end of the loop, except that he did say "under certain circumstances". Those circumstances would be a very, very long loop. DCC signals are "slow" in electronic terms, and colliding with each other at the end of a loop isn't really an issue, generally. On the other hand, not gapping the loops means that there is twice the wire, or half the resistance, and therefore voltage drop through the bus. Unless I saw a problem, I don't think I'd break the bus. But, if everything has been done right, it really shouldn't matter, one way or the other.
Jeff But it's a dry heat!
Thanks bearman. You bring up the next question on my steep DCC learning curve ... "power districts". As an old model railroader, I had tons of blocks, but with individual operating engines, what is the point of having power districts?
Breaks up your layout into sections that a single booster can power.
"Power district" can be a confusing term. Some, like Dan, use it to describe the section of track powered by a single booster. Some call that a "booster district". Others refer to the section of track protected by a circuit breaker as a "power district". Some call that a breaker or protection district.
A layout is divided into booster districts when a single booster doesn't supply enough current to run the whole layout.
Booster districts are divided into breaker or protection districts so that a short in one area doesn't shut down the whole layout (or booster district). Breaker districts also have a side benefit of making a short easier to find.
Breaker districts, on layouts using signaling, are divided into detection blocks.
Even if one booster can power your entire layout (which is quite likely for most of us) power districts can allow for a short in one district not shutting the whole layout down until the short is cleared. That's a good thing!
I intend to put switches (as in on off) on the tracks in my roundhouse and engine terminal. That way I can have all my sound engines that are not in use, powered off and I can't accidentally switch to the wrong engine and send it into the pit.
BigDaddy I intend to put switches (as in on off) on the tracks in my roundhouse and engine terminal. That way I can have all my sound engines that are not in use, powered off and I can't accidentally switch to the wrong engine and send it into the pit.
I *PLAN* on doing one better -- the TT alignment sensor (whatever that happens to be) will also drive an input to the selected track (incl the lead at the other end) so, all the tracks will be off unless the turnout's aligned.
Will need to probably have some form of timer as well, so that it doesn't power-cycle everything as the bridge sweeps past.
carl425 "Power district" can be a confusing term. Some, like Dan, use it to describe the section of track powered by a single booster. [...]
"Power district" can be a confusing term. Some, like Dan, use it to describe the section of track powered by a single booster. [...]
Well, the way I had it explained was that a "Power District" is the booster's domain, which then (potentially) gets broken up into smaller sub-divisions by fuses / circuit breakers.
Guy who explained it to me came from the multi-cab-on-DC days, so perhaps his explanation was colored a bit (or a lot) by that experience.
NeO6874 BigDaddy I intend to put switches (as in on off) on the tracks in my roundhouse and engine terminal. That way I can have all my sound engines that are not in use, powered off and I can't accidentally switch to the wrong engine and send it into the pit. I *PLAN* on doing one better -- the TT alignment sensor (whatever that happens to be) will also drive an input to the selected track (incl the lead at the other end) so, all the tracks will be off unless the turnout's aligned. Will need to probably have some form of timer as well, so that it doesn't power-cycle everything as the bridge sweeps past.
If you're doing the control yourslef, just code it such that while the turntable is in motion, no tracks ever get turned on. No need for time delays.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.