Jump to content

DCC 'ring main'


2e0dtoeric

Recommended Posts

Reading a rival mag about converting an existing layout to DCC from dc, The writer goes on to say that the DCC bus under the board should not be a ring main, but arranged as a 'star', or a spur with offshoots.
The converting bit is irrelevant to me,

as I am starting from scratch, - but -
can anyone say why a ring main wiring arrangement is not acceptable? He goes on to say that a full circuit of track won't work, either, it needs an insulated break somewhere, to stop it being a ring. Why?
If the

(for convenience) RED wire track is on the outside at the front of the circle, it is also on the outside at the back!
If his instruction is correct, then the basic circle and siding box of train-set cannot work!
He also goes on to say that the bus should

have a twist in it, every couple of feet. That I can understand, as it will reduce the amount of RF radiation coming from it. (I'm a radio ham as well) but I've not seen these instructions anywhere else.
I have nothing built yet, just a collection of components

awaiting a board! The controller will be the Elite, with the disk.
Comments?
Link to comment
Share on other sites

As I understand it, there is a theoretical possibility that a ring can cause data corruption but for this to happen the DCC bus ring or loop of track would have to absolutely huge. As for the twist in the cable, I thought it was supposed to prevent external

interference having an effect on the DCC bus but I suppose it could work both ways. Does he also suggest a twist in the track?
Link to comment
Share on other sites

Hi
I feel that some of the web sites advice being offered may not be of the best intended?

Have a read through the DCC page of this web site which may resolve many of the issues you have raised. http://www.brian-lambert.co.uk/DCC.htm
Another

good web site is the many pages of this one... http://www.dccconcepts.com/index_files/DCCadvice1.htm
As for ring, radial or star bus wiring, it really doesn't matter.
A ring wont provide any better power path than a correctly installed and sized radial

or star bus pair will. It will of course use more wire and therefore would become more expensive to initially instal. But that expense is proportional to the actual size of the layout.
Twisting of the bus pair of wires (often recommended to be at around

8 twists per mtr run of wire) on a small to medium sized layout wont probably have any effect. But on very large layouts where the bus wires have to run for long lengths the twisting may well help overcome both inductance and possible induced RF interference

too.
Link to comment
Share on other sites

Just a quick note to say that I have a 12ft x 6.6ft layout continous run. Double track. Yes there are breaks in the track because of Electrofrog pointwork. I have the bus as a ring main and no twisted wires. I have been using it for nearly 4 years and

I have had no problems whatsoever in that time in running on the Elite. There is probably over 40 locos sitting on the layout at any one time, again no problems. I have had no data issues whilst the locos are running. All run as they should do and told to

do.
Link to comment
Share on other sites

I have yet to button down the details of DCC signalling but I cannot see any reason for high frequencies to be used. (Hope I am right about this).

Ethernet cabling used for computer LANs comprises up to 4 pairs of wires all in the one cable. The

cables are often made up of each signal pair being a twisted pair to minimise crosstalk between the pairs. However, the cabling carries signals at highish frequency (say 100 Mbits or higher) where cross talk could be an issue. ALso the cabling runs were often

100's of metres as they ran across the floors of large buildings. I just cannot see how short runs on a model railway layout and at far lower signal frequencies could be subject to such crosstalk problems. I just think, as some others have hinted, that some

'internet advice' are taking theory from other cabling systems and thinking they also apply to signalling in DCC where the parameters are entirely different.

Using mains cable for the bus is possibly ok but bear in mind that mains at 440v p-p or 240

rms is 16X higher than 15v for DCC so, like power for power, DCC needs to carry 16x current amps. A good thick but simple speaker cable of say figure 8 type and capable of carrying say double amps of that of the psu with virtually no resistance for the entire

length should be fine (ignore the oxygen free nonsense that many believe in). I cannot see any reason to make it a ring so just a linear run from one end to the other. Then drop off to each track section as you see fit, these being much shorter could be thinner

(but not thin).

I would love to see justification from Lenz as to their comment of rings being bad. Do they believe the signal would run around and around introducing phase shifts in such short lengths? But why use a ring anyway.
Link to comment
Share on other sites


There isn't any! Except the DCC Bus

pair of wires doesn't have rail joiners in it.

But that said, many layouts, mine included, while they may well have a continuous run tracks the rails are not actually continuous DCC power wise. I have insulated rail joiners (IRJs) after every point

as all my points are all Electrofrog ones and I have several power districts too which means all the rails are insulated from the next areas rails by IRJs.

The issue being that while a ring of DCC bus wire can be used and shouldn't cause any harm, it

really serves no useful purpose over that of a correctly wire sized Radial or Star wired DCC Bus. It in fact a ring uses more wire!

The most important issue with any DCC bus wire is to ensure the wire used os of adequate wire size. You can never have

too larger wires nor can these do any harm, but can easily have problems introduced due to skimping and installing bus wires that are too small!

The other thing is that most UK model railways are considerably smaller than those found in the USA. Where

whole basements etc are filled with model train tracks. In the UK the average layout is more likely to be around 8 x 6ft up to around 14 x 8ft mainly due to UK homes not having basements! Yes, I know there are exceptions, but I am referring to the average

UK layout size. Which fall into the small to medium sizes.

Someone mentioned Xpressnet bus wires - These are not the same and shouldn't be confused with the DCC power bus wires.
Link to comment
Share on other sites

As I stated above ..... There isn't any!!!!

However, metal rail joiners (Fishplates) can let the DCC system down electrically and data wise as they are prone to becoming loose and then cause a high resistance. Hence a DCC bus pair of wires whether

installed as a ring, radial or star is the better option.

As was also stated previously too. Twisting a DCC bus pair of wires is normally only undertaken on very, very large layouts. Often outside the scope of many UK train set sized layouts.

Whether

you chose to instal a DCC bus pair, twist these wires or not is entirely your own choice. You don't have to do any of it. Its your railway and you can do exactly as you please.
Link to comment
Share on other sites

Hi

Having looked at the spec of DCC on NMRA site, I will stick my neck out and say I see no reason for twisting the pair of bus wires (unless we are next to a power station, or rather in it). And, even at an exhibition, where many DCC systems could

be next to each other, I still see no reason for twisting the pair of bus wires. DCC is an alternating signal of only 18 volts or so (depends upon scale, see spec). And other voltages around should only be up to mains (even 3 phase?), being 340 volts peak.

The

NMRA spec shows that square waves of 8.6 KHz constitutes a stream of 1 bits and 5 KHz or less constitutes a stream of 0 bits. So, the frquencies involved are really slow compared to most signalling used in obvious forms of communications, eg computer wired

networks being 100 MBits/sec, a frequency of over 10,000x faster. Radio is dramatically faster.

In networks, in multicore cable the wires are very close, touching except for the insulation, one can consider the effective capcitance between wires. This,

plus the much higher frequencies, may cause 'corruption' between the signal. In DCC, in the bus there is only the one core carrying the signal, the other wire is the 'return' which is nominally at earth potential. And the frequencies involved are really low.

So, the need for twisting the pair is really low. A would love to see a justification saying otherwise.

But, if it makes you feel better, buy a twisted pair. Much better still, but still of dubious advantage, is to buy shielded pair of wires. But, do

buy reasonably thick wire in all cases.

I agree entirely with previous posters that the primary difference between actual track and a pair of bus wires (at the frequencies used in DCC) is the high number of rail joiners that are introducing resistance

along the track. The bus wires, resonably thick (multi strand) copper wire gets around this problem. I see no point in making it into a ring.

Does this answer the advantage of a bus? I will soon be putting my money where my mouth and will use thickish

pair of wires but not using twisted pair or shielded.

I would love to put a 'scope' across the track or bus to actually see what is going on.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
  • Create New...