Bus Wire Configuration

[91rioja]

I am wondering if there is any reason to keep the bus wire separated when running around the layout? I am not really versed on DCC just yet, but the "digital" part of this leads me to ask the question. Is there a true digital signal moving through the wires? If there is, then it seems to me that unshielded wire would not do well in close proximity to each other due to the possibility of crosstalk in the wires. Or is it OK to just run them through the same pathway all the way around?

Thoughts or help?

 

[fsm1000]

I do know that if you coil wire it can and will cause RF interference [radio frequency] which will affect any wireless things you use that are near it. Other then that, if the wire is insulated it should be fine.
Cross talk is more a problem with lower voltages and very high voltages on thin wires. For what we use in model RR it should not be a problem. However I am not an electrical engineer. Former adiophile yes, engineer no. LOL
Maybe someone else can give more precise info on it.

 

[MasonJar]

Another reason to keep them separate is to avoid confusion and possibly hardwiring a short into your electrical!

Andrew

 

[Iron Horse]

Hi....I don't know the answer myself but this is worth reading !

http://www.gatewaynmra.org/dcc/powerbus.htm

 

[Bill54]

From everything I've read shielded cable is recommended but not essential. You should not run the power bus parallel with the control bus because interferrence is possible. If the two must cross try to do it at a 90 degree angle.

I have set up a DCC system on a small 5' x 10' layout and had 4 UTP outlets connected that ran around the parmiter of the table without interferrence. I kept the power bus separated from the control bus as suggested.

Bill

 

[91rioja]

Bill54:

Did you use flat 4 pair, or Cat 5 or better twiste pair?

 

[NCMRailroad]

I think that using the thicker 10 AWG single strand wire would conduct alot better and have alot less resistance on the DCC Booster. Less the resistance the longer the life! Work= Force x Distance.. the Distance - we can not do away with, but if we can get away from resistance by increasing the gauge of wire we use on our bus line the less Force the DCC booster has to work. ....Just a thought!

 

[Gary S.]

My understanding is that DCC actually operates at a fairly low "speed" of data transfer compared to modern network systems. In other words, it carries a lot less data per second than the cat 5 data systems used in computer networks and such. Because of this, the DCC wiring is not as critical when it comes to cross talk and interference. The lower the data transfer rate, the less susceptibility to problems.

I am using Digitrax but I haven't used it that much yet because my layout isn't complete. I do have the wiring under my 70 feet of shelves "finished". For the track feeder bus, I am using #12, for the feeds to the track off the bus I am using #18, attached to the track about every 4 to 5 feet. The #12 feeder bus wire is located at the back of the shelves. The 6 conductor flat cable used for the loco-net connections is run near the front of the shelfs. This keeps them seperated from 18 to 24 inches. So far, it has all worked great with no problems.

Although I feel the DCC systems are fairly immune to interference, it still makes sense to keep the data-bus away from the feeder bus or any other power conductors.

My 2 cents, but take it with a grain of salt because I am pretty new to this.

One other thing... I was working on the layout one night while listening to a baseball game on a portable radio I had set on the layout shelfing in close proximity to the loco-net bus. When I turned on the Digitrax system, the radio emitted a continuous high-pitched squeal, at first I thought the Digitrax system was about to explode! Then it dawned on me what was happening.

 

[fastlane]

I think Digitrax recomends keeping bus lines under 30' to avoid interference.

 

[NCMRailroad]

mine is only going to be 15 and I am running NCE., not sure what they specify for BUS Lengths.

 

[60103]

With regard to bus wire separation nd twisting. wasn't the system designed originally to be carried on the rails? I would expect that wires could be as close together as N gauge track - 9mm or 1/3" - and I hope no-one ever expected them to be twisted.

 

[Gary S.]

I think Digitrax recomends keeping bus lines under 30' to avoid interference.

what I did was put the DB150 at the midpoint of the length of the shelves. The loconet "A" connection runs to the UP5s and UR90s to the left side, the "B" connection goes to the UP5s and UR90s to the right side. That way, length isn't an issue, at least not so far.

One thing that is confusing though... in the Digitrax Super Empire Builder manual, page 23, 4.3 Other Track Wiring Considerations #4, it states "to minimize the possibility of radio interference, twist all conductors." But they recommend using 6 conductor flat cable forthe loconet bus.... that isn't twisted.

Do any of you feel it is necessary to twist the track feeder bus wires?

 

[Gary S.]

With regard to bus wire separation nd twisting. wasn't the system designed originally to be carried on the rails? I would expect that wires could be as close together as N gauge track - 9mm or 1/3" - and I hope no-one ever expected them to be twisted.

Good point. Obviously, the track itself is not twisted and is certainly carrying the power and the data signal to the decoders. So it seems the track feeder bus is pretty much the same and doesn't need to be twisted.

 

[papasmurf37]

If you're talking about physical separation along buss wiring run under layout and not total length[another issue; per DCC system manufacturer's recommendations]; was in modular club and we used 12 ga. solid buss wires: one black; one white. We had 6" separation, with holes drilled horizontally thru cross-bracing. This was done only to make our 20ga. vertical feeders attachment a lot easier to do. TTFN.....Old Tom in NH

 

[Gary S.]

Does anyone know the "bits per second" data rate of the signal on the track that goes to the decoders? What about the speed of the Digitrax Loco-net bus?

I did a superficial search on the net and saw several references to 9600 for the signal to the decoders, mostly on the Lenz website. If it is 9600, that is pretty low compared to current network technology. But low is good in this case, because it is more immune to interference. And obviously the track and feeder bus system isn't anywhere close to meeting Cat 5 requirements for cabling.

As for the data bus, I saw some references to 19,600, again a relatively low data rate.

 

[baldwinjl]

Does anyone know the "bits per second" data rate of the signal on the track that goes to the decoders? What about the speed of the Digitrax Loco-net bus?

I did a superficial search on the net and saw several references to 9600 for the signal to the decoders, mostly on the Lenz website. If it is 9600, that is pretty low compared to current network technology. But low is good in this case, because it is more immune to interference. And obviously the track and feeder bus system isn't anywhere close to meeting Cat 5 requirements for cabling.

As for the data bus, I saw some references to 19,600, again a relatively low data rate.

Track data: There isn't an easy answer. A nominal "1" is 116 microseconds, a "0" will be between 200 and 12000 microseconds. This allows for stretching one side or the other of the "zero" bit to allow for operation of a DC loco. I expect (but it is not required) that when a DC loco is not being run the zero is about 200 microseconds. I you assume that there are an equal number of ones and zeros, then the data rate averages 158 microseconds, or about 6330 Hz. Here is the NRMA standard, if you want to look at the gory details: http://www.nmra.org/standards/DCC/standards_rps/S-91-2004-07.pdf

From this http://www.digitrax.com/ftp/loconetpersonaledition.pdf it looks like Digitrax's Loconet is running at 16.66 kbps. In this area every manufacturer's sytem can be different, there are no NMRA or any other standards. So there isn't any one anser there. There are no standards because the only requirements neccessary are between the booster and the decoder, how the system commumicates internally is not impertant. Tha't why items like the throttles are not interchangeable.

To get to the main point, the signal on the track is slow, and in electronic terms, huge. That makes it fairly immune to noise (a good thing with a little moter spinning right next to the decoder), and reliable without much concern for termination and length. The biggest concern is really just voltage drop over distance. The control bus is most likely faster, and more sensitive to noise, but also usually uses more specialized cables and connectors.

Jeff

 

[Gary S.]

Thanks for the info. As you mentioned, the track and power feeder bus signals have to be fairly immune to interference i.e. an electric motor running less than an inch from the track.

 

[NCMRailroad]

POLL QUESTION: When running a BUS wire under one's layout, what type of wire would be prefered - SINGLE or STRANDED?
NCMRailroad.

 

[LongIslandTom]

According to this: http://www.wiringfordcc.com/trakwire.htm

The difference in voltage drop between stranded and solid 14AWG wire isn't all that big... Either one should be fine I think.

I've done bus wiring using both solid and stranded wire... I wired up my modular layout sections with solid wire, but in retrospect I think I should have gone with stranded because it's flexible and less fragile.

 

[60103]

For electrical purposes, I think the same gauge of stranded or solid wire should be equivalent. My only significant criterion would be that wire that gets flexed should be stranded. I think stranded wire of a given gauge is bigger -- I saw a wire stripper where the holes were labeled sort of N solid, N+2 stranded.