Forgive me if this has been addressed. I looked, but didn’t see EXACTLY what I was looking for. Would someone please explain why I need extra power feeders on a DCC layout? My thinking was that if it’s connected, it’s connected. Is there any rule of thumb as to how many feeders versus the layout size? I know some refer to a “bus”. Is this the same thing? I’ve read a lot of DCC material, but evidently, I’m too stupid to figure it out unless I’m led by the hand.
I’ll try my best to help you out. The reason for feeders is that it gives the layout a more even power distribution. If you connect the wires to one side of the layout the far side would have a power loss. By putting in feeders in you make sure every track section is getting equal power. With DCC when you loss power the decoder has a problem recieving the information from the system.
Feeders are ussually placed every three feet on a layout. I don’t know why three feet was picked but that seems to be the standard. The bus wire is different than the feeeders. The bus wire is run the whole length of the layout and it is what is connected to the power pack and then to the feeders.
Hope this helps.
Andrew
1 Like
Andrew is correct in that it does give you better power distribution throughout your layout. I use a 12 gauge buss with 22 gauge feeders. Some like to use a larger feeder, but I haven’t experienced any problems with 22. I placed mine every 6-9 feet apart where practical and I like to put in extra ones on spurs and yards.
One thing is for certain:it will help eliminate power drops from bad track connections, turnouts, and etc. Added note: Always solder track joints. This will prevent corrosion and keep paint and foreign material from interupting power at the connection.
REX
You are attempting to avoid power (voltage) drops. In model railroading these occur through mechanical connections, chiefly rail joiners on the layout - there is a small loss at each one. When the joint corrodes, the resistance at the joint increases and the loss is greater. To avoid the mechanical connection, you can solder your rail joints, supply feeders to every section of track, or a combination. Since flex track and rail come in 3 ft sections, you could solder a feeder to each one which is where the 3 ft comes from. Or you can solder some joints and leave others to allow for expansion and contraction, you would then solder feeders to each section of soldered rail. Wire size is important because voltage drop in wire is a function of smallness of the wire and distance. Using large wire such as #12 for the bus pretty much eliminates the problem for model railroads. Smaller feeders from the bus to the rail work because the distance is short. The rail itself has a large enough cross section that it won’t cause voltage drop.
In theory you could solder everything, the rail, jumpers around turnouts, etc and only connect the power to the track at one place. In practice you need to at least leave expansion joints and thus use extra feeders.
Enjoy
Paul
Dropping feeders is good for DCC or DC layouts. Sooner or later those rail joiners will become loose or corrode - and you will have voltage drop problems. Soldering all of your rail joiners may sound like the simple answer, but you then have issues with melting the ties, and lack of rail expansion. I have seen layouts with everything soldered, and they have no problem - I have seen more layouts with soldered rail joiners and lots of expansion problems. I drill a hole by the rail and solder a pre-tinned #22 wire to the side of the rail( hardly every melt a tie). The feeders drop about 6" and are tied into a #18 ‘bus’ for the electrical block. This is very reliable and is easy to trouble shoot if I have a problem. That said, I have soldered rails together on some short sections of track or if the rail joiners are ar the start of a curve.
Our club soldered every other rail joiner on the 3’ flex track(alternating sections). This reduced the number of feeder drops. 16 major power districts, and only ‘1’ wiring error.
I am taking the morning off Wednesday and play trains down at the club!
Jim Bernier
either way you probably will need a feeder or two but if your track is in a stable environment (say a basement, or a room that has good artificial ventilation) soldering your rails together should not cause you any problems and it will decrease the amount of feeders you actually need.
Someone on the forum said every piece of track needs to be soldered to something, another piece of track or a feeder. Someone else said not to solder your switches. After screwing up a switch I decided that was very good advice. Now I solder a feeder or another track (with a feeder) to anything that is not a switch.
I am running DC now, eventually I will get a DCC system. I can tell a difference in the speed of my trains on some of the track sec ions that are not yet attached to the buss. I would worry about the digital signal along with the voltage drop. It is relatively cheap and easy to do and there plenty of people that know a lot more about MR than I doe recommend it. So I do it.
Ok, This all makes good sense, but is the bus a loop with a feeder from the power pack and then multiple feeds to the track, or does it have two ends and it’s own voltage drop issues hence the need for a large gauge wire? J.R.
I’m using Kato Unitrack. Not only diea Kato say it does not need soldering, I don’t think it could be done without some damage. I will add some feeders, though.
GMT, the buss is large so that it is capable of handling the total amount of current used on the layout without much power loss. The amount of loss on a 12, 16, or 18 gauge wire is very very little considering the lengths used on a layout.
The buss, by itself is not a loop. It is two seperated wires that are connected to your controller. Then each feeder wire connects one buss wire to one rail of the track. The electrical loop you may be thinking about is then made through the motor (decoder) of the loco.
FYI: Make sure you maintain proper polarity of your feeder wires throughout the layout. i.e., same buss wire to same rail.
REX
I’m still in the planning stage so I have track down in a temporary fashion, no soldering joiners yet. Also no bus or feeders… yet. I’ve been running a couple of diesels with no problem. I bought an NW 2-6-6-4 BLI engine that is a good bit heavier and running it over the same tracks brought on problems with stopping and I strongly suspect that the heavier loco is pressing the track and breaking the connections with the rail joiners, thus the power loss. If I’m correct in this it is the best argument I can think of for adequate soldering and drop feeders.
Jarrell
Jarrell
Yes, it is a possibility that the one loco is breaking the connections as it runs or is more sensitive to power fluctuations. Soldering should correct this (also, check for clean wheels). Many of the methods used for track installations are there to eliminate potential problems more so than as a necessity for electrical connection. One area of your layout may do just fine by not soldering joints or installing feeders. But, what if later…more locos in the same area, dirt and grime, corrosion, paint, glue, something dropped on track at joint, and etc…?
It is always a good idea to do anything and everything to insure that your track is as near ideal as possible. Nothing will ruin your fun more than track problems. I even have adopted methods that I don’t completely agree with. Why take the chance? [;)][:D]
REX
I have to admit that I’m very skeptical of this “feeder to every piece of track / solder all joints” philosophy. I’ve run buss wires below the track, but I only drop feeders every 15 feet or so, and I only solder rail joints on curves, to get a smooth curve through the joint. I don’t have any power problems yet. With some nearly 20-foot long tangents, I’m more concerned with expansion / contraction of rails, so I leave a slight gap ever other track section or so.
Adding a feeder to every single piece of track seems like using a sledgehammer to swat a fly. Then again, maybe the flies will overwhelm me…
Mark,
There’s good engineering and there is what will do. Feeders every 3 feet is over engineered (IMHO), but it is a good rule of thumb. With DC, it’s not so much of an issue, but for DCC, you have to be concerned about signal quality more than quantity (e.g, amplitude) like DC. With more feeders current will flow easier and you will (generally) get less noise and better response from your decoders. If every piece of track has a feeder, you don’t need to solder the rails together (IMO). I like the mantra - everything should be soldered to something.
-Tom
Guys, here’s the problem.
Nickel silver track, as an alloy, is a very poor electical conductor. For example, here’s the voltage drop per foot of code 70 nickel silver track, compared to 12 guage copper wire:
Code 70 nickel silver track: 0.605 volts per foot at 4 amps.
12 guage copper wire: 0.015 volts per foot at 4 amps.
Let me put this in perspective. The voltage drop in 10 feet of code 70 nickel silver track will be 6 volts, while the voltage drop in 10 feet of 12 guage copper wire will be 0.1 volts. Even at 3 feet the voltage drop in code 70 nickel silver track at 4 amps will be 1.8 volts! Meanwhile the voltage drop in three feet of copper wire at 4 amps will be less than 1/20th of a volt.
The voltage drop will be less for lower amperage values and larger code track, but still you see the point. Nickel silver track is a very poor electrical conductor.
So running copper wire feeds to every single 3 foot track section will solve any voltage drop issues and keep your trains running smoothly throughout the layout.
NOTE: Notice just soldering rail joiners will not solve this problem because it’s the nickel silver track itself that’s the poor conductor, not the rail joints (although they can be a problem too).
Joe:
I think you ment .0605 volts per foot at 4 amps.
I’m using code 83 this 18 ga feeders every 6 to 7 feet with a 16 ga bus and have no problems at all. My layout is 12’ X 15’ with 5 blocks
Teffy:
These figures are from Allan G.'s web site and were measured with a sensitive LCD meter. Allan’s figure for Code 70 nickel silver rail is 0.605 volts drop per foot, but that’s at 4 amps remember. The drop for 1 amp is 0.151 volt per foot. See: http://www.wiringfordcc.com/trakwire.htm …
The figure for code 83 NS track from Allan’s site is: 0.340 volts per foot at 4 amps, but since most loco lashups won’t draw 4 amps, if we check Allan’s table for 1 amp (which is more typical for an HO loco lashup) we find the voltage drop per foot is: 0.085 volts – since we’re now talking just a single train and not trying to get all the booster’s power reserve to the other end of the layout.
Since you have feeders every 7 feet, then that means the farthest any loco will be from a feeder will be 3.5 feet. So voltage drop will be: 0.085 x 3.5 = 0.2975 volts. If your power supply is running 14 volts output, you will still have 13.7 volts on the track at the lowest spot, which will be fine – you’re still getting well over your 12 volts.
But the farther you go between feeders, the worse it gets and the greater chance you run of having more trains drawing more amps through the same section of rail, which only increases the problem. Trying to push all those amps through poor conducting nickel silver rail gets harder and harder.
For 15 feet between feeders on code 83 NS track with one loco lashup drawing 1 amp, we have:
0.085 x 7.5 = 0.6375 volts lost
Now if you try to run QSI sound locos in your train, say a lashup of 4 locos, we’re talking about 3 amps on code 83 NS track, Allan says you have 0.255 voltage drop per foot, which gives:
0.255 x 7.5 = 1.9125 volts lost - or nearly 2 volts drop.
So the more power hungry locos you put between feeders, the more noticeable the problem will become. If your power supply is running at near 12 volts, 15 feet between feeders with sound equipped locos on the layout will re
You’re piling worst cases on top of each other, I think, Joe, at least for DCC.
Your calculations make sense, but how often will you run into the situation you mention? The first worst case - you’re pulling a full three amps on that 4-unit lashup. When running at protypical speeds, most locos aren’t running at full throttle. At lower speeds, you’re pulling less amps, and the motors will be getting fed substantially less than 12 volts.
The next worst-case being presented is that everything is 7.5 feet between feeders. In the worst situation for your scenario, the center of the lashup is 7.5 feet from a feeder, but the end trucks of the end locos is only somewhere around six feet from the nearest feeder, depending on how big the units in the lashup are, of course. So only one unit at a time would see the full two volt drop, with the others seeing incrementally more voltage (this may actually make it worse, I suppose, depending on the dynamics of the units). In this example, the end units would suffer only a 1.5 volt drop.
Using an NCE system (the one I have), if voltage drops two volts, that means the track between feeders is still getting 12.25 volts - plenty. I think most of the DCC systems provide something near 14 volts, per NMRA standards, don’t they?
One thing I don’t know (and would like to, if anyone is of a mind to explain) - does the speed table in the decoder function as a percentage of voltage available, or does it provide a specific voltage at each step? If the latter case, then unless you’re trying to feed more voltage to the motor than is available in the track (ignoring losses in the decoder), you won’t see any noticeable drop in speed as the track voltage varies between feeders.
Is my analysis flawed in some way? If so, please point it out, because I’m missing it.
This is a really enjoyable discussion!
Mark:
My understanding of how decoders work is they don’t do a lot of jacking up or voltage dropping, because that all takes heavier circuit components than you see on the average light weight decoder. Minor variations in track voltage are fine for the decoder, but anything notable, like .5 volt or more, will be felt by the loco.
Your analysis is basically correct and yes, I am postulating to a certain degree a worst case situation. But reliable operation means to me I give those electrons every chance they can to get the job done, so planning based on best case is, IMO, just asking to have operational reliability issues.
As to the current load for a loco lashup, until you’ve run a lashup of 3-4 QSI sound locos, you haven’t seen what modern loco current draw can be – 3 amps is not out of line for a lashup of 3-4 QSI sound equipped HO locos. And the real problems with nickel silver track become more apparent as you try to push more current through it.
The trend in modern locos with all the goodies is lots of current. Again, planning for the future (worst case, if you will), rather than seeing how little I can get by with (best case). Since I don’t like wiring, I want to do it once and then rely on it to be rock solid from then on no matter what I chose to put on the layout.
If there’s one place where it doesn’t hurt to over-engineer your layout, it’s in the wiring. At least that’s the philosophy I’ve operated with, and any place I’ve cut corners in the wiring, I live to regret it later.
Your points are well taken, Joe.
Since my layout is set in the 1930s to early 1940s, at this point it’s all steam (pluse a few Doodlebugs). At most I may doublehead a couple of BLI steamers, so my voltage drop away from the feeders should be proportionately smaller than with a large diesel lashup. I probably will run an occasional out-of-era train pulled by 1960s diesels (after all, this is for fun), and maybe I’ll be crawling under the layout swearing at myself while adding extra feeders.
I’ve no doubt that a layout built with 3/4" plywood subroadbed and feeders every track section is pretty much bulletproof if the workmanship is even moderately good, but having spent my entire career in aerospace, where such overengineering has severe penalties in terms of weight increases, cost overruns and range or payload decreases, I tend to design and build my layout more like I do an aircraft - engineered closer to the “line.” Time will tell if I’ve come TOO close.
So exactly how does a decoder control loco speed? Does it do it by varying the width and/or frequency of full-voltage pulses to the motor, or through some other means? I can probably find out at the NMRA website, but if anyone has a quick and dirty (but accurate) explanation, I’d rather not wade through all the electron-pusher verbiage to find out.