DCC Bus wiring - to twist or not?

[C855B]

I'd be willing if I had the parts on hand. PM me if you do and can drop them in an envelope. Been a long time since working with DC power at the component level.

What I can tell you without actually doing it is it will round the corners of the waveform and possibly attenuate that little 1V bobble.

[peteski]

PM sent.
I realize what effect it will have on the waveform, but I would still like to see  it on the scope (to show how severe the rounding off is).

[C855B]

Here you go, @peteski . With the snubber, and then without to verify:



As you can see, a slight softening in the rise; did nothing about the +/–1V bobble.

EDIT: I take that back. It cleaned up that spur on signal rise.

[peteski]

Thanks Mike!
 Not much difference.  Smooths out the ringing at the top of the raising edge. The flat parts of the waveform are also slightly bowed.  I guess to really see any meaningful effect it would have to be on some DCC bus with problems (not clean like in your example).

Any chance to show the DCC signal difference between twisted and untwisted bus?  I suspect there will not be much difference.

[C855B]

... I suspect there will not be much difference.

As do I. Twisting is primarily to defend against inductive noise. Given the low frequency of DCC, unless you live close to an AM radio station or have some big motors nearby there's not going to be much to see. Also, such problems are usually transient, and hard-to-impossible to catch and prove. I think I mentioned that my reasoning is from long-ago work in RF systems, especially power supplies in 2-way radio installations, both mobile and base. Loosey-goosey with power leads and control interconnects was frequently a source of trouble, only "proven" by going back in and doing it the right way.

Besides, I like how twisted wire bundles handle. Less like spaghetti.

[Jbub]

As do I. Twisting is primarily to defend against inductive noise. Given the low frequency of DCC, unless you live close to an AM radio station or have some big motors nearby there's not going to be much to see. Also, such problems are usually transient, and hard-to-impossible to catch and prove. I think I mentioned that my reasoning is from long-ago work in RF systems, especially power supplies in 2-way radio installations, both mobile and base. Loosey-goosey with power leads and control interconnects was frequently a source of trouble, only "proven" by going back in and doing it the right way.

Besides, I like how twisted wire bundles handle. Less like spaghetti.

To me that would be a bigger benefit to twisted pair. Having everything organized as a set would help keep things in order and if there is ever a problem, easier to trace.

[Bill H]

Oh my, just when I thought this was "settled" - I was reading my DCB 240 manual (I know that RTFM is strange in itself) and apparently Digitrax recommends twisting. AFAIK, that is the first recommendation I have seen from Digitrax.

Page 19 "4. To minimize the possibility of radio interference, twist all conductors"

[nickelplate759]

It just occurred to me that the recommendation to twist might be more to avoid causing RFI , not preserving DCC signal integrity.  I don't know enough about it to know if that's reasonable.

[peteski]

This whole RFI interference on the DCC track bus seems silly (and we have bought it up few times in this thread already).  DCC signal is relative low frequency high-power square wave.  We are talking many Amps here. RFI interference is very low power, high frequency radio waves.  It is kind of like saying that a sound from a transistor radio (remember those?) would interfere with the sound from the giant speaker system at a live rock concert.  Really!

Well, unless you layout resides right next to a 400kW AM radio transmitting tower and the DCC bus is very long.  But in that situation, you in that location (as a human radio wave receiver) would also likely be affected by those very strong RF waves. 

[robert3985]

Although I really enjoyed the conversation here, I went through this 13 years ago when I tore out my rat's nest of DC wiring and did a lot of research into what the best "high speed/low drag" DCC wiring was for a medium to large sized N-scale portable layout.

Although I don't agree with many of the standards that Ntrak has/had, they are undeniably the one modular group that used to have the largest assembly of modules I had ever seen.  This led me to believe that probably if they could operate reliably and regularly with DCC, then many of their DCC standards might apply to me and my go-to-the-show 30' X 15' sectional layout.

After pondering the alleged pros and cons of twisting the power buses, I decided since Ntrak doesn't have a problem not twisting their power buses, I wouldn't either in my approximately 90' run of double-tracked mainline.

What I elected to do to minimize voltage drop and minimize signal loss was to go with premium speaker wire...high strand count, pure copper, black/red zip, 12AWG...because I planned on eventually having a much larger 30' X 30' sectional layout...so why not use 12AWG?

Later, I found a pretty good treatise written by a credentialed Electrical Engineer, that 14AWG was the very "best" for preserving signal strength...had something to do with waveforms and diameter of 12AWG wire.  Oh well...

I also preferred to have as few solder joints as possible in the layout, using Genuine 3M IDC's (Insulation Displacement Connectors) everywhere I could, getting away from any terminal strips that had screws that might work loose while the layout was disassembled and jouncing around in the back of my Suburban or in the trailer.

I also went with color coded crimped Anderson Power Poles between layout sections.

Lastly, although this was the first thing I did after removing all of my DC track wiring, I put a 22AWG 6" long solid copper feeder on to each and every piece of rail on my layout, soldering them onto the bottom of the railfoot with silver bearing solder that's 5 times stronger than tin/lead solder so they would have much less of a chance to pull off.

I connected the 22AWG 6" long feeders to 14AWG sub-buses each less than 3' long that flowed under the track subroadbed, then attached to the 12AWG red/black zip speaker wire main power buses which were attached to my L-girders running the length of each 6' long section/module.

Photo (1) - Finished new DCC wiring on one of my 6' long layout sections:


In the 13 years since I re-wired my layout using these specification, I've attended as many as 3 shows a year (on good years) transporting the layout over 100 miles for two of the three setups, and I've never had ANY signal loss, low voltage, or RF interference problems...with 100% reliability with not a single feeder coming off of the track or coming loose from any sub-power-bus.

At shows, I run as many as four trains per mainline, with a yard switcher for a total of 9 sound-equipped DCC trains...with a DC train sometimes thrown in doing zero-stretching operation on address 00 and additional sound equipped engines idling in the yard.

Since I chose top-quality components and shopped for good deals online (and got them by buying in bulk) and used "best practices" for feeder placement, the effort and expense has been worth it in every way, providing easy setup and takedown, total reliability, and so far, total lack of transportation damage.

Things will get more complicated when I finally decide to signal the layout, but that's not gonna happen until the layout gets more trackage to accommodate at least two more center sidings, which will involve a larger layout room too.

Cheerio!
Bob Gilmore

[wvgca]

to keep every body happy, just twist HALF of the bus line .... no, no, the OTHER half ,,, lol

[Bill H]

Bob:
I also noticed, if I read your color codes, that your bus wires, black and red, are very very close to each other, which effectively is the same result as twisting. So perhaps the question should be keeping the bus wires next to each other in some manner or another, or keeping them far apart or...

Kind regards,
Bill

[robert3985]

Bob:
I also noticed, if I read your color codes, that your bus wires, black and red, are very very close to each other, which effectively is the same result as twisting. So perhaps the question should be keeping the bus wires next to each other in some manner or another, or keeping them far apart or...

Kind regards,
Bill

Bill, I did not know that.  Yup, they're about as close together as they can get and not short since they're speaker zip cord.  Ain't I a lucky fella!    Pure, dumb luck....

Cheerio!
Bob Gilmore

[Bill H]

Bob:
Like you, I use ME code 40. I have been using 22 gauge feeders to the bus wire, but much more frequent than most. Because I think it is logical to assume that code 40 rail has less capacity than larger code rails and all my locos have sound - and what the heck it can't hurt.  So every rail has a feeder, regardless of length, and if the section is longer than a piece of rail, it gets more than one feeder.

What has been your experience with feeders and how frequently you drop a feeder to your bus?

Kind regards,
Bill
 

[robert3985]

Bob:
Like you, I use ME code 40. I have been using 22 gauge feeders to the bus wire, but much more frequent than most. Because I think it is logical to assume that code 40 rail has less capacity than larger code rails and all my locos have sound - and what the heck it can't hurt.  So every rail has a feeder, regardless of length, and if the section is longer than a piece of rail, it gets more than one feeder.

What has been your experience with feeders and how frequently you drop a feeder to your bus?

Kind regards,
Bill

Bill:

I also wire a 22AWG solid copper 6" long feeder to each and every piece of rail that should be powered on my layout...no matter how short.  Maximum rail length is 3', so I attempt to solder my feeders very near the middle of each piece of rail.

My layout is modular/sectional, with the maximum section length being 6' (with at least one exception) which mean that gaps for expansion/contraction are naturally every 6' at the module ends...the rest of the gaps are electrical gaps.

I also shorten my rail joiners so that they are the length of the space between ties, and are used only for mechanical strength, not for carrying electricity or DCC packets.

I use 14AWG stranded pure copper sub-power-buses no longer than 3'...most being less than 2'...that run directly under my layout's subroadbed, and attach my 22AWG track feeders to these using Genuine 3M IDC's (Insulation Displacement Connectors or "Suitcase Connectors").

My 14AWG sub-power-buses are connected to my 12AWG main-power-buses also with Genuine 3M IDC's.

When I re-wired my layout for DCC, since it is sectional, I was able to set each module/section on a 6' banquet table, C-clamping it to the edge and removing the folding legs...which greatly assisted me in getting the new wiring finished without having to once crawl around under the layout.

When I removed the old DC wiring, the first thing I did was install the new 6" track feeders.  Unfortunately, much of my trackwork was already painted, ballasted and weathered...and well glued down to my cork roadbed...so I bought a couple of longer "aircraft" length drills that were approximately the same diameter as my insulated 22AWG track feeders so I could put a small kink in the feeder wire, and it would make an interference fit in the hole and not fall through, allowing me to position it exactly under the rails, then solder it to the base of my Rail Craft Code55 flex and to my hand-laid Code40 PCB yard and branchline trackage.  The extra length also let me drill holes without the chuck of my drill messing up the railheads on my track.

As I stated in a previous post on this thread, I used 96/4 Tin/Silver Silver bearing solder, which is 5 times stronger than common electrical Tin/Lead solders...so I could minimize or eliminate any feeders pulling off the bottom of the rails, which is good, since after I re-paint, re-ballast and re-weather where they're at, they are INVISIBLE.

Photo (1) - Here's a photo of when I was cutting and inserting feeders into the newly drilled feeder holes in my Echo Yard Section.  Although this looks like lot of feeders, it's only half since I haven't inserted the black feeders yet or my green frog feeders!


Photo (2) - Since I mentioned them, here's a photo of my shortened rail joiners after soldering & painting, which I use only for rail alignment after rewiring.  Although they still conduct electricity, their electrical properties are superfluous:


If anybody's interested in how to make these shortened rail joiners, ask and I'll do a separate posting about how to make them and use them.  I think I've done it before here, but it's been a long time ago.

Cheerio!
Bob Gilmore