Power Bus Wire

[Andymsa]

This is an extract from another forum 

 

"It's not recommended (by some people) to run the DCC bus main as a ring circuit. There can be problems with "ringing", basically the digital signal gets distorted (two signals arriving at the same place at very slightly different times) and the decoders are unable to decode the garbled DCC information."
 

[Crosland]

12 minutes ago, newbryford said:

 

But how many layouts actually have a continuous ring of two rails without insulating breaks?

 

It's not the rails that matter.

 

At some point a loco or stock will bridge the two ends of an open loop bus, creating a complete loop, i.e.:

 

bus end <-> dropper <-> rail <-> loco straddling isolation gap <-> rail <-> dropper <-> other end of bus

 

The bus is now a complete loop. The layout does not stop working, the sky does not fall in, ...

 

[newbryford]

3 minutes ago, Crosland said:

 

It's not the rails that matter.

 

At some point a loco or stock will bridge the two ends of an open loop bus, creating a complete loop, i.e.:

 

bus end <-> dropper <-> rail <-> loco straddling isolation gap <-> rail <-> dropper <-> other end of bus

 

The bus is now a complete loop. The layout does not stop working, the sky does not fall in, ...

 

 

Maybe.

But how many layouts will only have one gap to be bridged? Or unlucky enough to  have multiple "bridges" over multiple gaps.

And it's not a continuous loop for very long as the train passes over.

 

And I agree that the sky will not fall in.

 

 

[RFS]

21 minutes ago, Andymsa said:

This is an extract from another forum 

 

"It's not recommended (by some people) to run the DCC bus main as a ring circuit. There can be problems with "ringing", basically the digital signal gets distorted (two signals arriving at the same place at very slightly different times) and the decoders are unable to decode the garbled DCC information."
 

 

Whoever wrote that is talking complete nonsense. Electricity, like light, travels at 186k miles per second. On a model railway layout, with a bus measured in (usually) less than 100ft, the timing difference is likely to be so minuscule no DCC decoder could possibly be able to detect it. You would need a bus measured probably in km for any problem to arise.  

[Art Dent]

44 minutes ago, newbryford said:

 

But how many layouts actually have a continuous ring of two rails without insulating breaks?

 

... and how many layouts have a continuous ring with no breaks?

 

Thing is neither of us will know.  

 

Of course you could always wire the bus as a ring and if you encounter problems, split the ring and put bus terminators at each end of the 'C' shaped bus (and also put at least one isolating break in both of rails in the loop).

 

Art

[Art Dent]

28 minutes ago, RFS said:

 

Whoever wrote that is talking complete nonsense. Electricity, like light, travels at 186k miles per second. On a model railway layout, with a bus measured in (usually) less than 100ft, the timing difference is likely to be so minuscule no DCC decoder could possibly be able to detect it. You would need a bus measured probably in km for any problem to arise.  

 

Err ... actually no.

 

Electricty (or electrical current) to be more precise is a flow of negative charge carried by the electrons - they are not a form of electromagnetic radiation (which does travel at approx 186k miles/sec).

 

The drift velocity of electrons through the copper lattice is approx 200 m/sec IIRC.

 

EDIT: Actually, just doing a quick calcuation, for a current of 1A in a copper wire of cross-sectional area 1.5mm² , the drift velocity is about 0.05 mm/ sec.

 

The equation to use is  I = n A v Q  where I = 1, n = 8.5 × 10²⁸,  A = 0.5 × 10⁻⁶ and Q = 1.6 x10⁻¹⁹.  v is the drift velocity in m/s.

 

Either way it isn't anwhere close to c, the speed of light (= 300,000,000 m/sec).

 

See: Drift velocity in a copper wire  (scroll down to numerical example).

 

Art

Edited August 8, 2019 by Art Dent

[RFS]

1 hour ago, Art Dent said:

 

Err ... actually no.

 

Electricty (or electrical current) to be more precise is a flow of negative charge carried by the electrons - they are not a form of electromagnetic radiation (which does travel at approx 186k miles/sec).

 

The drift velocity of electrons through the copper lattice is approx 200 m/sec IIRC.

 

EDIT: Actually, just doing a quick calcuation, for a current of 1A in a copper wire of cross-sectional area 1.5mm² , the drift velocity is about 0.05 mm/ sec.

 

The equation to use is  I = n A v Q  where I = 1, n = 8.5 × 10²⁸,  A = 0.5 × 10⁻⁶ and Q = 1.6 x10⁻¹⁹.  v is the drift velocity in m/s.

 

Either way it isn't anwhere close to c, the speed of light (= 300,000,000 m/sec).

 

See: Drift velocity in a copper wire  (scroll down to numerical example).

 

Art

 

OK thanks  - point taken. Perhaps I was trying to be too simplistic! In reality, though, the speed at which a DCC signal travels around a DCC bus means that, if it's a ring, the timing difference is going to be too small to have any impact. 

 

Perhaps whoever believes that ringing on a DCC bus happens should provide some measurement data to demonstrate this, and that DCC decoders can be affected by it. 

 

My DCC layout is 36'x10' and the bus is a U-shape with the DCC command station at the bottom of the U. A while back I did actually turn it into a ring by connecting wires across the other end, and it made absolutely no difference whatsoever. After several months I removed the wires again.

[Art Dent]

That's OK RFS, no problem.  I agree with you that any timing issue on our layouts is likely to be extremely small (and like as not covered under the NMRA specifications so that it won't be an issue).

 

I tend to get a bee in my bonnet when I see something posted that isn't correct (I'm somewhere on the spectrum I guess).

 

Although the drift velocity is fairly low (well, compared to c anyway), the signal travels rather quickly - otherwise the transatlantic telephone cable would be useless across 4000 miles of the North Atlantic.  I've never seen a calculation as to how quickly the signal travels.

 

You can explain the effect like this - "Imagine a drainpipe packed end-to-end with billiard balls.  As soon as you push another ball in at one end of the pipe, a ball drops out of the other.  The ball you put in hasn't travelled the full length of the drainpipe - but the movement of the balls (effectively the signal) has travelled from one end to the other extremely quickly."

 

Anyhow, interesting as this discussion is, the OP's thread seems to have take something of a diversionary path (and I am partly to blame for this).

 

Time, methinks, to put this to bed or possibly coninue the discussion in a new thread.

 

Regards,

 

Art

Edited August 8, 2019 by Art Dent

[WIMorrison]

1 minute ago, Art Dent said:

...

Although the drift velocity is fairly low (well, compared to c anyway), the signal travels rather quickly - otherwise the transatlantic telephone cable would be useless across 4000 miles of the North Atlantic.  I've never seen a calculation as to how quickly the signal travels.

 

...

 

In this case at the Speed of Light - they are all fibre optic, and very high quality fibre at that

[Art Dent]

1 minute ago, WIMorrison said:

 

In this case at the Speed of Light - they are all fibre optic, and very high quality fibre at that

 

Yes, but I was thinking of the early copper wire transatlantic cables dating from 1955/6 - long before the advent of lasers and fibre-optic cables (actually the first transatlantic telegraph cable was laid (according to Wikipedia) in 1858.

 

Article: Transatlantic Communications Cable

 

Art

[Sol]

I initially had my bus not joined but then as I now use common return for DC circuits and one rail for signalling  purposes, all my independent power supplies use the common return & one rail, the main bus is now a ring  ( layout room size is 12 x 20ft). Track is actually powered via sub-bus  via subsections with the powered rail ( to distinguish from the common) and I have bus teminators on the sub-busses to absorb spikes that any short circuits may generate.

[Crosland]

14 hours ago, Andymsa said:

This is an extract from another forum 

 

"It's not recommended (by some people) to run the DCC bus main as a ring circuit. There can be problems with "ringing", basically the digital signal gets distorted (two signals arriving at the same place at very slightly different times) and the decoders are unable to decode the garbled DCC information."
 

 

It's utter rubbish. Most tales of magical improvement by cutting the ring also seem to involve upgraded wiring and/or adding filters. Guess which is the real, effective, treatment

 

 

 

14 hours ago, newbryford said:

 

Maybe.

But how many layouts will only have one gap to be bridged?

 

The point is it only needs ONE gap to be bridged, the gap between the two ends of the broken ring.

 

Something that doesn't seem to be mentioned is that a lot of UK layouts are small end-to end affairs. In that case it's fine to have a linear bus. No need to join the ends.

[Andymsa]

57 minutes ago, Crosland said:

 

It's utter rubbish. Most tales of magical improvement by cutting the ring also seem to involve upgraded wiring and/or adding filters. Guess which is the real, effective, treatment

 

 

 

 

The point is it only needs ONE gap to be bridged, the gap between the two ends of the broken ring.

 

Something that doesn't seem to be mentioned is that a lot of UK layouts are small end-to end affairs. In that case it's fine to have a linear bus. No need to join the ends.

 

Dont kill the messenger, I posted this from another forum just to show how hot a topic it is. But why create a ring bus apart from the extra cost and effort in putting it in. A well designed liner bus run should be just effective.

Edited August 9, 2019 by Andymsa

[WIMorrison]

15 hours ago, Art Dent said:

 

Yes, but I was thinking of the early copper wire transatlantic cables dating from 1955/6 - long before the advent of lasers and fibre-optic cables (actually the first transatlantic telegraph cable was laid (according to Wikipedia) in 1858.

 

Article: Transatlantic Communications Cable

 

Art

 

Do you recall the delay that we used to get on the line when chatting to people in the States, very noticeable - and with poor voice quality. There was a time when I used to drop the call and redial in the hope that I got a digital fibre circuit rather than an analogue copper line, and eventually they got to digital copper which was almost as good as fibre.

[Sol]

4 hours ago, Andymsa said:

 

Dont kill the messenger, I posted this from another forum just to show how hot a topic it is. But why create a ring bus apart from the extra cost and effort in putting it in. A well designed liner bus run should be just effective.

 A ring bus is perfect for a layout which runs around the walls & the operating area is inside as many layouts I operate on ( including my own)  are  that format - not just an end to end on a largr plank layout

[WIMorrison]

I guess what this thread has proved is that whilst everyone has their own view of what best practice should be there will always be someone else who has a different version of what that best practice is.

 

And in the end does it matter what anyone does as long as it works for them? 

 

After all this is only model railways that we are talking about here - which may be the most important thing in the world for some posters but ultimately it is still a model - and the beauty of this hobby is that everyone that has something working for them has the right solution for them  

 

 

[Art Dent]

1 hour ago, WIMorrison said:

 

Do you recall the delay that we used to get on the line when chatting to people in the States, very noticeable - and with poor voice quality. There was a time when I used to drop the call and redial in the hope that I got a digital fibre circuit rather than an analogue copper line, and eventually they got to digital copper which was almost as good as fibre.

 

No, I don't remember the delay.  All I remember is phoning relatives in Australia as a kid at Christmas (to say thank you for presents) and it costing over £1 per minute!

 

Folks these days don't really appreciate the (almost) instant and (virtually) free communications that we have now become used to - for example my son is currently in Florida and we have been chatting by text about the Hulk roller-coaster (which he said was bl**dy scary!).  My daughter who was recently in The Outer Banks, N.C. (and was lost) so we were following and guiding her progress towards the Post Office using Google Street View and saying "Now you should see a Texaco garage on your right hand side, turn next right by the WalMart store".

 

Totally agree with your comment " ... the beauty of this hobby is that everyone that has something working for them has the right solution for them ".  Some solutions are, however, more elegant than others 

 

Art

Edited August 9, 2019 by Art Dent

[Guest]

On 04/08/2019 at 16:06, Junctionmad said:

The reality is you don’t need an actual DCC bus feed for small layouts , just feed the track from the booster , there’s no such thing as “ digital “ wires 

Digital bus wires discuss here and digital wires are synonym as DCC/MM and mfx system.

If we discuss analog layout we write analog wires what you feel and like.

Of course you can use DCC bus feed at small layout.

Nothing is impossible.

You feed the track from a common connection too via booster.

[newbryford]

8 hours ago, Crosland said:

 

 

The point is it only needs ONE gap to be bridged, the gap between the two ends of the broken ring.

 

Something that doesn't seem to be mentioned is that a lot of UK layouts are small end-to end affairs. In that case it's fine to have a linear bus. No need to join the ends.

 

Nobody likes a broken ring.........................

.

.

.

.

.

.

.

.

.

.

.

.

 

 

[RobjUK]

On 08/08/2019 at 19:40, Art Dent said:

Err ... actually no.

 

Electricty (or electrical current) to be more precise is a flow of negative charge carried by the electrons - they are not a form of electromagnetic radiation (which does travel at approx 186k miles/sec).

 

Totally wrong.

 

The propagation delay of an electrical signal through a cable is proportional to the speed of light x the "velocity factor" of the cable.

It is nothing at all to do with electron drift velocity.

 

Velocity factor depends on the cable construction, type of insulation & capacitance - but is typically between 0.6 - 0.9 so the actual signal propagation speed range is something like 180,000,000 -  270,000,000 metres per second.

 

Reference:

https://en.wikipedia.org/wiki/Velocity_factor

 

 

Back to the main subject, I don't really see any difference in result if a feeder is a ring or split system. Either could hypothetically cause cable ringing or reflection effects but in reality if whatever you use works, don't worry about it.

 

[Guest]

7 hours ago, RobjUK said:

 

Back to the main subject, I don't really see any difference in result if a feeder is a ring or split system. Either could hypothetically cause cable ringing or reflection effects but in reality if whatever you use works, don't worry about it.

 

I wouldn´t say that.

Digital experts recommended you not use ring system when you use digital system.

Otherwise you risk data interference with binary codes.

 

 

[RobjUK]

2 hours ago, Anders63 said:

Digital experts recommended you not use ring system when you use digital system.

Otherwise you risk data interference with binary codes.

 

I am a "digital expert" - I've been designing & consulting in electronics for manufacture for over 40 years, including high speed digital systems & I've worked with everything from digital signal processing & radio data links through multiple hundreds of kilowatts power controls.

 

I am also extremely familiar with transmission line effects that can cause problems in some kinds of systems.

In transmission line terms, an open end with no termination or load is generally the worst case for causing signal reflection & distortion problems.

 

In strict transmission line terms, every joint, transition or tap in a wiring system causes signal reflections, unless it is very accurately designed and the cable (and connector) impedances properly matched.

 

No one considers such things with any other part of the power wiring interconnections for a DCC setup, yet then are no less relevant than the bus ends being linked as a ring.

 

 

Suitability of any particular wiring / connection scheme depends totally on the application.

For a multiple-interconnected system like a typical DCC & track setup with numerous droppers off the main bus and multiple cable / track stubs, you have no more - or less - chance of suffering signal distortion with a ring config than an open end.

 

The whole subject is a non-issue and just confusing or worrying people who do not understand the technical details.

 

 

DCC uses a modulated data system (not raw binary) and has error checking in the data demodulation routines, so occasional corrupt data frames do not cause problems.

 

It was & is specifically designed for electrically noisy systems.

 

[Crosland]

17 hours ago, RobjUK said:

Velocity factor depends on the cable construction, type of insulation & capacitance - but is typically between 0.6 - 0.9 so the actual signal propagation speed range is something like 180,000,000 -  270,000,000 metres per second.

 

One nanosecond per foot is a good approximation we used to use in my high speed board design days. High speed for the late 80s/early 90s

[Crosland]

10 hours ago, Anders63 said:

I wouldn´t say that.

Digital experts recommended you not use ring system when you use digital system.

Otherwise you risk data interference with binary codes.

 

A little knowledge is a dangerous thing

 

That is simply not true in the general case. For systems like CAN and RS-485 (both used for MR control busses) you certainly don't want a ring. For some other applications it really doesn't matter.

 

7 hours ago, RobjUK said:

In strict transmission line terms, every joint, transition or tap in a wiring system causes signal reflections, unless it is very accurately designed and the cable (and connector) impedances properly matched.

 

No one considers such things with any other part of the power wiring interconnections for a DCC setup, yet then are no less relevant than the bus ends being linked as a ring.

 

The problem with DCC is that the loads have a habit of moving around

[Guest]

16 hours ago, Crosland said:

 

A little knowledge is a dangerous thing

 

That is simply not true in the general case. For systems like CAN and RS-485 (both used for MR control busses) you certainly don't want a ring. For some other applications it really doesn't matter.

 

When somebody like Lenz and other digital product manufacturer they says don´t make a ring by connection digital wires back to the booster, they verified a reason of why.

This is to avoid digital interference.

Under all mine year with the digital system i have never ever used a ring system, just split system.