Every wire has inductance. The amount is proportional to the length and the diameter. As the length increases for a given diameter, its inductance increases. The opposite is true for the diameter: as it increases the inductance of a given length decreases.
To make it more interesting, two wires placed side by side have less inductance than a single wire, because their effective surface area is greater, which is the same as using a larger diameter wire.
Twisting them loosely together just helps to keep them close, the same effect can be had with parallel wires and cable ties.
Twisting them tightly has the undersired effect of creating a capacitor, the tighter the twist, the more you get. It isn't much, but it adds up over the length of the wire.
Why does this matter with DCC?
With the nature of the DCC signal, which is turning on and off constantly, the wire's inductance creates a reactive component. Inductors oppose any change in current. They store current in the magnetic flux around the wire, and release it when the current decreases. The nature of an inductor is that it's effect is frequency based, so excessive inductance can distort the signal.
Ask any physicist and he'll tell you the same thing.
No one has mentioned what I thought was the conventional guidance, that if your busses are more than 25' that twisting the wires and snubbers on the ends are recommended to reduce some kind of electrical problem. I'm not sure what the problem is, not being an EE, but there is definitely a recommendation by qualified individuals out there to this effect. Whatever the problem is, it gets worse with the length of the bus, and 25' is considered the point at which it becomes good practice to do this.
Probably resolving most of the actual concerns in this and in 'that other thread' re-animated from 2014:
https://www.allaboutcircuits.com/tools/twisted-pair-impedance-calculator/
Twisted-pair adds some conductor length and hence some marginal DC resistance... see threads on this concern. As DCC maintains peak DC track voltage at all times, the resistance is a less-significant consideration than it would be if using pure analog voltage control.
Likewise, the DCC modulation is imposed across the 'whole' of the wiring on most layouts, and 'crosstalk' between adjacent runs or bundles may not be significant at DCC effective frequency with DCC modulation structure.
Impedance is a transmission characteristic between end-to-end transmission and reception on the pair. Its importance in reception of DCC signal via a multiplicity of random drops (the feeder taps) is an interesting technical discussion, but vast everyday experience would indicate that its real-world relevance to layout builders is comparatively slight.
richhotrain Did you simply twist the bus wires and nothing else to make the problem go away?
we added capacitors to the detectors. our wires weren't twisted.
greg - Philadelphia & Reading / Reading
Greg, are you sure that untwisted wires were your problem? Did you simply twist the bus wires and nothing else to make the problem go away?
Rich
Alton Junction
this sounds like a good reason not to twist (but i've had this problem with untwisted wires. we added capacitors across the detector inputs)
from Wiring for DCC, under Simple Bus (without explain why)
Caution: If you use twisted pair wiring within a detection block, it may contribute to false occupancy events. It is best to keep the wiring untwisted and keep that wiring as short as possible at the same time. This is best achieved by placing the block detector near the block it is going to detect as apposed at some central wiring panel.
mobilman44 On my layout (2 level HO 15x11) the bus wires were kept separate for ease of connecting all the feeders.
On my layout (2 level HO 15x11) the bus wires were kept separate for ease of connecting all the feeders.
From "whatever"........ So much has been written about why to twist or not to twist, but has anyone experienced problems before or after installing the wiring?
ENJOY !
Mobilman44
Living in southeast Texas, formerly modeling the "postwar" Santa Fe and Illinois Central
crossfeed from what? the other DCC line?
On my layout (2 level HO 15x11) the bus wires were kept separate for ease of connecting all the feeders. DCC operation was fine for the 12 year life of the layout.
This "twist or don't" topic came up then, and likely a handful of other times without (IMO) firm resolution.
That said, I have a simple question - which I don't believe has been asked before.....
Has anyone every experienced any "crossfeed" problem from twisting or not twisting their wires?
gregc richhotrain I have been running multiple trains on a large DCC-powered layout for nearly 19 years with 14 gauge solid copper untwisted bus wires with absolutely no problems. it's unreasonable to suggest that because you've never had a problem that no one would ever have a problem.
richhotrain I have been running multiple trains on a large DCC-powered layout for nearly 19 years with 14 gauge solid copper untwisted bus wires with absolutely no problems.
it's unreasonable to suggest that because you've never had a problem that no one would ever have a problem.
gregc the track is part of the bus. It's not twisted
the track is part of the bus. It's not twisted
https://cs.trains.com/mrr/f/744/p/228673/3411813.aspx#3411813
I did go on to say in this thread that home-based layouts just don't seem to need any of that stuff that may apply to huge club layouts or other sophisticated electronic or electrical setups.
So, I think that it is reasonable to suggest that based upon my personal experience, it is unlikely that most home-based layouts would experience a problem with untwisted bus wires.
What I look for in threads about the topic of twisted bus wires is a summary of problems, if any, that do occur with untwisted bus wires. I emphasize "do occur". In other words, do any model railroaders actually experience problems on their home layouts that are caused by untwisted bus wires?
richhotrainI have been running multiple trains on a large DCC-powered layout for nearly 19 years with 14 gauge solid copper untwisted bus wires with absolutely no problems.
SeeYou190 wrench567 Being a big truck and heavy equipment mechanic for far too many years and being in the trench when electronic controlled engines and drive trains came out. Me too! Were you around when the Cummins PACE and ECI systems were first put out? I would love to go back to 1988 and warn all the old techs that said they were too complicated to hold onto their butts, because you have no idea where we are going with this! wrench567 Especially with a cam sensor. Now trucks have more position sensors than you can imagine, all with twisted wires. My first nightmare with a cam position sensor was on a CELECT+ M11. The cam thrust had worn to a point where when the cam was all the way forward the gap between the cam sensor and the tone wheel would excede 0.025" and the engine would lose good signal. It never induced a fault, and the only symptom was a miss under hard accelleration. wrench567 Every speed sensor had twisted pair wiring right up to the ECUs. That was why I was refering to J1587 and J1939 in my previous answers. I found more than one owner-induced problem due to improper wiring harness repairs. wrench567 I build my DCC modules with a twisted buss for insurance. Peace of mind. Weather it makes a difference, I don't know or care. Same here. On the DCC layouts I have been involved in building,we always used twisted 12 TFFN AWG wire for the main bus wires. If there evere were problems to troubleshoot, we would know it was not for lack of care during assembly. Being able to rule out possible causes prior to beginning troubleshooting is a real time saver, and good for peace of mind. -Kevin
wrench567 Being a big truck and heavy equipment mechanic for far too many years and being in the trench when electronic controlled engines and drive trains came out.
Me too! Were you around when the Cummins PACE and ECI systems were first put out? I would love to go back to 1988 and warn all the old techs that said they were too complicated to hold onto their butts, because you have no idea where we are going with this!
wrench567 Especially with a cam sensor. Now trucks have more position sensors than you can imagine, all with twisted wires.
My first nightmare with a cam position sensor was on a CELECT+ M11. The cam thrust had worn to a point where when the cam was all the way forward the gap between the cam sensor and the tone wheel would excede 0.025" and the engine would lose good signal. It never induced a fault, and the only symptom was a miss under hard accelleration.
wrench567 Every speed sensor had twisted pair wiring right up to the ECUs.
That was why I was refering to J1587 and J1939 in my previous answers. I found more than one owner-induced problem due to improper wiring harness repairs.
wrench567 I build my DCC modules with a twisted buss for insurance. Peace of mind. Weather it makes a difference, I don't know or care.
Same here. On the DCC layouts I have been involved in building,we always used twisted 12 TFFN AWG wire for the main bus wires.
If there evere were problems to troubleshoot, we would know it was not for lack of care during assembly. Being able to rule out possible causes prior to beginning troubleshooting is a real time saver, and good for peace of mind.
-Kevin
Kevin.
Hate to tell you. I started working on the big rigs when the government mandated the FMVS121. The very first ABS nightmare for the trucking industry. I got my Chauffers license in a 62 Brockway with an NTC350 in front of a 5&4. No power steering. No front brakes. Heat year round and 10.00/20 bias ply on split rims.
Ya, I been around awhile.
Pete.
What is the reason to twist dcc bus wires together?
Answer: There is no reason to twist DCC bus wires together.
I have been running multiple trains on a large DCC-powered layout for nearly 19 years with 14 gauge solid copper untwisted bus wires with absolutely no problems. None. Nada.
I also started out with no snubbers, later added snubbers when they were in vogue, and now don't use any snubbers. I had absolutely no problems with or without snubbers.
Home-based layouts just don't seem to need any of that stuff that may apply to huge club layouts or other sophisticated electronic or electrical setups.
wrench567Every speed sensor had twisted pair wiring right up to the ECUs.
Living the dream.
betamaxThe main reason to twist, or even keep the wires close together is to reduce the reactive component in the bus impedance. Full Stop.
so what is the effect of the "reactive component" for us model railroaders? why should we care?
what is the effect on performance? packet corruption, voltage drop, signal loss?
what is the quantitative difference between twisting and not twisting? how much better is twisting?
"Interference" is a red herring in DCC wiring.
To really impact any crosstalk or interference takes a lot more math and effort than the average modeller would be willing to undertake. The nature of the DCC signal is aready a robust transmission method, so interference is almost a non-starter. If that were true, every AC wire in your house would be twisted to prevent radiation or reception of interference.
The main reason to twist, or even keep the wires close together is to reduce the reactive component in the bus impedance. Full Stop.
In comparison to a network cable, whose twists, wraps and how they are orgranized is carefully engineered to result in the desiried electrical characteristics for that cable's application. Plus the addition of a shield to provide a path to ground for any induced currents.
Being a big truck and heavy equipment mechanic for far too many years and being in the trench when electronic controlled engines and drive trains came out. I can open up a four inch diameter bundle of wires and tell you every pair of wires going to an induction probe. Every speed sensor had twisted pair wiring right up to the ECUs. There were four just in the transmission of a front loader. Two or sometimes three on the engine. Electronic controlled ABS could have ten or more. Repair of twisted wire had to be done right or codes would pop up. Especially with a cam sensor. Now trucks have more position sensors than you can imagine, all with twisted wires.
I build my DCC modules with a twisted buss for insurance. Peace of mind. Weather it makes a difference, I don't know or care.
I know that in the Army, we used "twisted pair" wire on our field phones. I was told by our Battalion Signal officer that it eliminated cross talk between circuits. Not being a Signal Corps type, that's about all I knew about it.
Fifteen years ago when I started this layout I landed two leftover spools of 12G wire each with 300' on them. There was a bit of a different clientele on this forum back then and I was criticized on this forum for using such heavy wire even though I paid very little for it compared to what I could have paid for a lighter gauge. I pulled the two lengths of wire down my driveway a 150', stuck it in the drill, and had it twisted in seconds.
Now I am a bit of a when it comes to the electronics side of this hobby. While I can and have wired houses it was a little different than wiring a DCC layout. Being a very proactive type of person I started reading a lot and decided that twisting the wire was probably a good idea.
I have been having some minor issues with programming newer Loksound chips and spent the day reading about how signals get from point A to point B. I learned about the skin effect, something I have never heard of before. What caught my eye was how a signal was carried in stranded wire versus solid wire. There have been great debates on this forum when it comes to solid vs stranded wire and I have made up my mind about which is better for DCC.
Twisted buss wires, solid or stranded, number of feeders, and how far apart, all these seemingly little things on their own might never cause an issue but all together may result in less than perfect performance.
Proactive or reactive? Personally, I do everything to prevent problems that may occur down the road.
Brent
"All of the world's problems are the result of the difference between how we think and how the world works."
the bigger effect is the loops created by twisting allow interfernce to be cancelled out down the length of the wire
Impedance is a function of inductance. Reducing the inductance reduces the impedance because 3+4=5.
Bundling cables together will reduce the inductance, thanks to the mutual inductance.
betamaxTwisting the bus wires loosely (or even keeping them together using cable ties) reduces the impedance of the bus.
There is no possible way twisting wires will change the impedance in the conductor. Impedance is fixed (for the most part) based on the physical properties of the conductor.
The reason for twisting wires in any digital communication bus (I am most familiar with J1587 and J1939 (CanBus)) is to prevent a false signal from being induced into the bus.
This is caused by the magnetic field created as current flows. We could go through the right hand and left hand rules about field formation and collapse, but that is not necessary.
Anyway, twisting the wires lowers the possibility of INDUCTANCE, it does not change IMPEDANCE.
You might have just typed it wrong, but the distinction is important.
Twisting the bus wires loosely (or even keeping them together using cable ties) reduces the impedance of the bus. That is the only benefit to it.
Twisting the Power Bus
That is the only benefit from twisting the bus wires together. The downside is a tight twist will create a capacitor, whose leakage currents can confuse any detection coils in use.
If the wires are twisted, it reduces (or eliminates) induced false signals from one wire to another.
There are digital communication protocols like J1587 or J1939 that require the signal wires to be twisted, but I really do not think it is necessary in a model train DCC system.
But... it sure won't hurt anything.
ndbprr See title
See title
Helps reduce the slack/droop/dangle of the wires. Makes for a nice, neat bundle instead of a rats nest. Helps with determining which wires are which if you need to troubleshoot.
I'm not an electrician, but I do all my own layout and decoder wiring. I don't believe it's necessarily about DCC, but the wires are conductors and they move energy from one point to another.
Inspired by Addiction
See more on my YouTube Channel