To Twist or Not To Twist


D

DALDEI

Member
Hello !

The layout I'm building is "E" shape about 10'x20' ... Its "sectional" (but not
modular). (8 sections 6-8 feet each) Under the core benchwork supports I plan
on running the main DCC power for each power district, branching off from the
middle where I'll have a single controller (Zephyr) and a few power districts
and AR's. I was reading up on layout wiring and decided to go for 12 gauge
(was going for 10 but changed my mind). I bought 50' of "romex" 12 gauge inside
home wiring (3 conductor).

I am planning on hooking this up to terminal strips under the benchwork for
every section. Then hook up each section to the main bus. In each
(detachable) section then I'll have its internal sub-bus wiring in 14 gauge.
My goal is that I should be able to remove a section and power it individually
for testing or benchwork, leaving the main whole-layout wiring permanent. Each
length of main buss will be at most about 20' long from the zypher.

Now my question ! I've been reading that for DCC main buss wiring should be
twisted (3-5 turns/foot). I realize DCC doesnt "require" this but that its
"suggested". I'm wondering if in a layout of my size its important enough or
not. I'd prefer to keep the wire in the romex point-to-point and only cut it
open where I connect to terminal strips (about every 8 feet). But then it
wouldnt be twisted, although I suppose I could twist the whole romex cable but
not sure how well that would work.

Is it worth cutting open the romex, pulling out interier cable and twisting it
or will I be OK just keeping it in the romex between terminal strips ?

Suggestions welcome !

=David
 
You'll be fine without the twist. Our club has hundreds of feet of track and the wire is straight.
 
Last edited by a moderator:
12 gage? isn't that what they use for high powered tool (20A) cords? bit overkill, no? :)

in any case, i definitely would want to have twist in my bus wire. easy and cheap way to lessen any potential interference of digital signal.
 
Many of us have long runs of track in small scales where the rails, each mimicking the bus under them, run for many tens of feet only a few millimeters apart, and they don't exactly wind around each other, do they. ;)

Crandell
 
DALDEI said:
..... WTF do I know ! this is the first time I've done anything like this and it seems there are more opinions then people :)
Click to expand...

:D

At least the same # of opinions as people...... :)

My 02c - Quality installation is more important than any nominal twist in the wire. Your Romex will be fine. (IMHO)

Cheers,
Ian
 
I don't post much here but this is one topic that amuses me.

I always enjoy a good laugh when the subject of twisting verses now twisting comes up. The concept of parallel wires causing or picking up interference is the general reason for twisting wire. Now how do you twist the track. Here are two pieces of metal fairly close together depending upon scale that will alway run in parallel. If there was a real problem with interference then surely the rails would be just as likely a point of entry as the wiring bus that feeds those rails.

Just something to think about the next time this discussion appears.

Al
 
Just to add gas to the fire ... Rails are further apart then wires. Even N guage.
EM radation drops off quickly with distance (r^3 rougly) so the inductive coupling of wires say 1mm from each other should in theory be vastly higher then rails 1cm from each other. 1000x higher.

But still at the frequency range and distance that DCC operates I have a hard time imagining its really an issue.
Audiophiles dont use twisted pairs to hook up speakers ...

but on the other other hand, it helps telephone signals going miles ... so somewhere between 30 feet and 10 miles theres probably a potential for issues...

Speaking of audiophiles ... speaker wires and connections are one of the biggest 'snake oil' scams of all times. I've actually seen ads for wires that are so special they "learn" the kind of music you like and auto-adjust over time. Sorta like a car engine being burnt-in. For the priviledge of these magic wires you pay about $50/foot plus extra for the magic connectors that anyone with the kings new clothes can "hear" the difference. If you can't hear the difference your obviously not enough of an audio snob. :)
(while funny, thats about how the ad went ...)
 
I have over 40 years of electronics experience so let me add:
The DCC output is a huge square wave, which in itself generates all kinds of harmonics (essentially noise) plus when you put a few motors on the track (locos), you get a whole bunch more noise with a fairly high current level. You should see it on an oscilloscope!

While twisting the wires may help suppress a little bit of noise, and may reduce interference to AM radios, it probably won't do much for your trains. Ferrite beads would probably do the same thing.

I'm sure someone could get on here and start talking about common mode rejection and all that, but our time is better spent on the layout.

On a practical not, twisting is prettier, but then it's harder to strip and tap in.

Chuck
 
i don't really feel strong about the subj. in my case it would be extra effort to untwist as i use a pair from cat5 cable for my runs (gage 22 BTW). while the effect is perhaps negligible in small or medium layout, recommendation from DCC manufacturers engineers is there. it makes sence to me and costs nothing to follow. don't want to twist? don't then.
 
I think it was already said but having heavy enough wire to handle the current for longer distance is probably more important than twisted or untwisted. For example, if you ran 100' of #22, which would be a HUGE layout, you'll have 1.6 ohm on each wire, or 3.2 ohms total which means you'll lose a little bit of drive voltage if you run several locos at once. On a small layout #22 will work fine, and it's a lot cheaper.

Twisting certainly won't hurt anything unless it's your arm.;)
 
DALDEI said:
Just to add gas to the fire ... Rails are further apart then wires...
Click to expand...

How would you know how far apart the bus wires are in comparison to the rails in any gauge, let alone N? Each wiring application is different, whereas a reasonably reliable constant will be the gauge of any one scale.

For example, if I place my two bus wires parallel, non-wound, but separated by 24", I should have near zero interference for many yards. If you place them within the same distance as the N scale rails above them that you are powering, the rails and bus are identical except for differences in efficiency in transmission, and each should have close to the same problems.

Crandell
 
Selector said:
...
For example, if I place my two bus wires parallel, non-wound, but separated by 24", I should have near zero interference for many yards.
Click to expand...

with cables that far apart what you not going to have is crosstalk i. interference is something external and will get "picked up" by whatever wire of the pair. perhaps of not sufficient magnitude to completely distort DCC signal but still.

again, i don't feel strong about the subj. engineers at [brand here] probably are knowledgeable about this topic and probably had some reason to recommend twisting (if only to cover their behind). it makes sence to me, i'm twisting (or more precisely - not untwisting), but i don't insist.
 
Anton, I am not an engineer or an electrical technician, so you'll hopefully excuse my improper terminology and limited understanding. My point, though, is that one's arrangement of bus wires is going to have a much higher statistical variance than will the separation of the rails in any one length of N scale tracks. So, if the rails, themselves, don't appear to have an impact on the DCC operation of locomotives over an extended length of contiguously powered rails, then the varied arrangements of bus wires powering them is hardly going to be any worse. In cases where a person places both wires in the same 3/16" hole drilled through wooden frame members, sure, it might be a good idea, and perhaps even highly recommended, to wind the wires around each other. In my case, my 14 gauge bus wires are typically separated by about 2-3", much more than my HO rails are just inches above.

Crandell
 
You must log in or register to reply here.



Back
Top