Jump to content

Power Bus Wire


petejones
 Share

Recommended Posts

The wiring on a layout needs to meet two  main criteria 

 

1. It has to power the operational track sections without excessive voltage drop and without appreciable temperature rise in the cable , up to the maximum operational current limit , however it’s reasonsble to select a particular maximum operational current for a particular track  that is less and design the cable accordingly. But the cable  must always meet citeria 2 

 

2. It has to cope with the short circuit circuit current for the duration of the booster trip fraction time and is must have a low enough loop resistance to allow a track short circuit to reach the trip current. “ coins or no coins “ . It’s entirely within engineering principles to allow a small rise over ambient but most cables ( including small cables) will withstand very significant surge currents for a short time . It does not have to handle indefinite short circuit currents. 

 

hence the the issue of short circuit currents tends to be about poor resistance connections ( current carrying fish plates , bad dropper to track soldering etc ) rather then fundamental under capacity wires , whereas the issue of voltage drop is far more critical in any medium to large layout and tends to dictate wire gauge, whereas almost any wire size in a layout will pass the track short circuit test 

 

 

 

 

Edited by Junctionmad
  • Agree 1
Link to post
Share on other sites

8 hours ago, Junctionmad said:

The wiring on a layout needs to meet two  main criteria 

 

1. It has to power the operational track sections without excessive voltage drop and without appreciable temperature rise in the cable , up to the maximum operational current limit , however it’s reasonsble to select a particular maximum operational current for a particular track  that is less and design the cable accordingly. But the cable  must always meet citeria 2 

 

2. It has to cope with the short circuit circuit current for the duration of the booster trip fraction time and is must have a low enough loop resistance to allow a track short circuit to reach the trip current. “ coins or no coins “ . It’s entirely within engineering principles to allow a small rise over ambient but most cables ( including small cables) will withstand very significant surge currents for a short time . It does not have to handle indefinite short circuit currents. 

 

hence the the issue of short circuit currents tends to be about poor resistance connections ( current carrying fish plates , bad dropper to track soldering etc ) rather then fundamental under capacity wires , whereas the issue of voltage drop is far more critical in any medium to large layout and tends to dictate wire gauge, whereas almost any wire size in a layout will pass the track short circuit test 

You're right that both criteria are important.  But both depend on the resistance in the bus, droppers and track, and if point (2) is satisfied then point (1) almost always is as well (possible exceptions on very small layouts with high-powered command stations). 

 

As someone mentioned above, a 5 amp command station needs the resistance out and back to be less than 3 ohms for the cutout to work, and actually significantly less given that the "short" itself will not be zero resistance. On a quick search, 16/0.2 wire which might be typical for feeds on a DC layout has a resistance of around 3 ohms per 100 metres so would probably be inadequate on a large DCC layout.  2.5mm2 cable reduces this resistance by a factor of about five - though there is still the resistance of the droppers and track to think about. 

Link to post
Share on other sites

Out of interest, what is the correct way to measure resistance using a multi-meter?  I'm assuming that it is in parallel with the circuit (as per measuring voltage)?  That is, I need one probe at the output from the command station and the other on the rail head to measure the combined resistance of the bus and dropper between the command station and pick up.  I would then do the same between the other rail and other command station output and add these together to give me the combined resistance of the feed and return paths.

 

Do many people actually check the resistance of their bus wires?

Link to post
Share on other sites

9 minutes ago, Dungrange said:

Out of interest, what is the correct way to measure resistance using a multi-meter?  I'm assuming that it is in parallel with the circuit (as per measuring voltage)?  That is, I need one probe at the output from the command station and the other on the rail head to measure the combined resistance of the bus and dropper between the command station and pick up.  I would then do the same between the other rail and other command station output and add these together to give me the combined resistance of the feed and return paths.

 

Do many people actually check the resistance of their bus wires?

Yes, that would be the hard way to do it. You may need very long test leads for a large layout, which makes their resistance influence your testing.

An easier way is to remove the command station/booster & replace it with a wire connecting the 2 terminals (technically, this will also have its own resistance but if you keep it short, this will be very low). Then you can measure the resistance across the rails at each part of the layout

  • Like 1
  • Thanks 1
Link to post
Share on other sites

30 minutes ago, Pete the Elaner said:

Yes, that would be the hard way to do it. You may need very long test leads for a large layout, which makes their resistance influence your testing.

An easier way is to remove the command station/booster & replace it with a wire connecting the 2 terminals (technically, this will also have its own resistance but if you keep it short, this will be very low). Then you can measure the resistance across the rails at each part of the layout.

 

Thanks - that sounds a much better way of testing the resistance and gives the resistance of the return and feed including dropper wires in a single reading. 

Link to post
Share on other sites

As a practical example I used 22 ga single bare tinned copper (non -insulated) wire as the two circular loop DCC buses (+ and - separately)  around my old 10 M long by 4 M wide oval.

 

Advantage one was that as the wire had no insulation, it was thin enough to lay on the top surface of the baseboard and disguise under ballast and the various scenic scenic ground surfaces.

 

Advantage two was no under layout crawling to install.

 

Advantage three was that there was no insulation to strip off when soldering to connect the many feeders to the rails and the feeder length was never more than 50 cm.

 

Advantage four was that short feeders have less length related resistance to worry about, so I was able to use almost invisible 30 ga single bare tinned copper wire (again without any insulation) just on the surface and through the ballast to thread the feeders in between the sleeper to access and and solder to individual rails in even complex track formations.

 

I too used wire resistance tables to come up with my wire diameters before I used this type of construction. So I wasn't  expecting any nasty surprises and I didn't get any. The lack of insulation on the wiring definitely increased the wire current carrying limits . In practice, the short circuit breakers worked fine, and even locos with heavy trains running at full speed, or a crawl, showed no discernible speed reduction at the far ends of the layout.

 

Unfortunately that layout was taken down due to a home move and I don't have pictures as I never got around to making any worthwhile scenery. But I shall use the same construction on my next and hopefully longer lasting layout. 

 

Tim

 

 

 

 

 

 

Link to post
Share on other sites

On 30/07/2019 at 08:27, Edwin_m said:

You're right that both criteria are important.  But both depend on the resistance in the bus, droppers and track, and if point (2) is satisfied then point (1) almost always is as well (possible exceptions on very small layouts with high-powered command stations). 

 

As someone mentioned above, a 5 amp command station needs the resistance out and back to be less than 3 ohms for the cutout to work, and actually significantly less given that the "short" itself will not be zero resistance. On a quick search, 16/0.2 wire which might be typical for feeds on a DC layout has a resistance of around 3 ohms per 100 metres so would probably be inadequate on a large DCC layout.  2.5mm2 cable reduces this resistance by a factor of about five - though there is still the resistance of the droppers and track to think about. 

Not so 

 

16/0.2 has a typical resistivity of 36 ohms per kilometer or 0.3 ohms per100m ,  0.03 ohms per meter ( not 3 ohms per 100 meters ) https://static.rapidonline.com/pdf/01-0900_v2.pdf

 

hence criteria 1( short circuit ) will almost always be satisfied on most layouts even with smaller ( under spec’ed ) wire as most resistance will be in the track and joints and hence the wire diameter is of little concern .  The time limited surge rating of thin cable is quite large , which is why DCC decoder wiring manifestly survives the short time limited  current 

 

hence the need for criteria 2 , the overall voltage drop calculation , this is a steady state operational parameter, ie it’s not time limited, hence the need for a conservative design , ie no temp rise in the cable ( which sets the operational  cable limit ) , and a known controlled worse case voltage drop. 

 

Dave 

 

 

Edited by Junctionmad
  • Agree 1
Link to post
Share on other sites

4 minutes ago, Junctionmad said:

Not so 

 

16/0.2 has a typical resistivity of 36 ohms per kilometer or 0.3 ohms per100m ,  0.03 ohms per meter ( not 3 ohms per 100 meters ) https://static.rapidonline.com/pdf/01-0900_v2.pdf

 

You have an error in your maths.

3.6/km = 3.6/100m = 0.036/m

..so 3ohms / 100m is close enough for our needs.

 

But I agree that you will see a larger resistance through rails & across joins.

Link to post
Share on other sites

14 minutes ago, Pete the Elaner said:

You have an error in your maths.

3.6/km = 3.6/100m = 0.036/m

..so 3ohms / 100m is close enough for our needs.

 

But I agree that you will see a larger resistance through rails & across joins.

Sorry , yes re the 100m figure 

 

the point is that  100m of a wire run on a model railway ( out and back) to a single point is very unusual , whereas overall it may be reached ,  hence. It’s entirely possible to meet the time limited short circuit current requirement with wire that is marginal operationally. Ie the wire gauge is very rarely the issue for this  criteria  , so 16/0.2 could be fine in many applications . Hence the two criteria.  Because voltage drop is a somewhat “ length of string “ argument you can get wide interpretations of what’s acceptable . Industry tends to use 5% and 10% 

 

Having said that typical DCC buses are usually far too big in my experience on quite modest layouts , 2.5sqmm being way over the top in many applications 

Edited by Junctionmad
Link to post
Share on other sites

I agree that a 100 m wire run is unusually long for most of us.  The layout that I'm constructing will be about 15' by 8', which means if I were to run a single pair of bus wires around the perimeter of the layout such that the far end were to terminate inches from where it starts at the command station, it would be just 92' or 28 m in length (14 m feed and 14 m return).  Using the figure of 36.8 Ohms per kilometre, that means that the maximum resistance of a 28 m 16/0.2 DCC bus on my layout would be just 1.03 Ohms and mounting the command station centrally the resistance to each half would be just half that.  I've established that to shut down my command station, I need a trip current above 5.5 Amps, which means at a track voltage of 14 Volts, I need the resistance to be less than 2.5 Ohms.   It therefore seems that a 16/0.2 bus wire would be more than adequate for my purposes with regards criteria 1 (the short circuit scenario).  The problem is that I'm not sure what to add to the specification for the resistance of connections between wires (eg rail to track feed and track feed to DCC bus) and that is probably why some people potentially over-specify the DCC bus.   It's easier to 'beef up' the bus than minimise the resistance of the joins.  I've only bought 2.5 mm2 wire because that's what some others seem to use.

 

I guess the uncertainty lies with the more open ended question of what sort of voltage drop is acceptable and how does one calculate that?  If the total resistance of my wiring (bus, droppers, track and joins) to the far end of my layout is say 2 Ohms (ie it will meet the short circuit criteria) and the locomotive that is operating there draws 0.5 Amps under load, am I correct in assuming that the voltage drop at that location will be 1 Volt (ie 0.5 Amps x 2 Ohms), which is less than 10 percent of the nominal 14 Volt track output, in which case I can say that this is acceptable?  Obviously if I want a voltage drop of just five percent (0.7 Volts) then I either need a locomotive that draws a lower current (eg 0.35 Amps) or I need to reduce the resistance of the wiring (to 1.4 Ohms) or a combination of the two.  Increasing the bus specification would seem to be the easiest way to reduce that resistance, although shortening the droppers would obviously also help.

  • Agree 1
Link to post
Share on other sites

  • RMweb Gold

Calculating voltage drop to determine DCC bus size is only part of the equation. DCC current also carries the DCC signal superimposed on it. It's a square wave operating at approximately 7000 cycles which results on a DCC bit-stream of around 7k and this does not have any parity checking. Therefore it's most important that a clear DCC signal is received at all parts of the layout. For a garage layout 15'x8' then 16x0.2mm wire is going to be far too small for the main DCC bus. You cannot over-configure the wire size but you can under-configure. 2.5mm from T&E is very cheap and will meet all the needs of a layout of this size provided you have plenty of droppers.

 

If you want to research further, I would strongly recommend having a read through of the excellent documentation on the DCC Concepts website, with this file on DCC wiring requirements being a good starting point - https://www.dccconcepts.com/manual/dcc-advice-9-dcc-power-and-layout-wiring-part-2/

 

Obviously there are plenty of references to their products but the basic philosophy is very well described.

Link to post
Share on other sites

I think there is a tendency to over specify the bus cable for DCC . I can understand this because it errs on the safe side 

 

having said that for average layouts , up to 15 feet long centrally wired in a bus/star format, 2.5sqmm is quite an amount of overkill and good quality stranded cable of 2.5 cross section is expensive stuff 

 

I also see people sticking rigidly to “ bus “ wiring structures , which often result in longer overall runs then necessary 

 

by the way computing impedance for DCC frequencies is quite ridiculous , as the complex impedance present on typical tracks on a layout dwarf   the impedance issues in the feeder cables and makes that dcc concepts advice quite nonsense 

 

in professional circles solutions like this are resolved on a “ systems “ basis are not focusing on a single element ( often simply because it’s the easiest to understand ) 

Link to post
Share on other sites

22 hours ago, RFS said:

 It's a square wave operating at approximately 7000 cycles which results on a DCC bit-stream of around 7k and this does not have any parity checking.

 

It most certainly does have parity checking. The final byte in every packet is an XOR checksum of the preceding Address and Data Bytes - refer to NMRA S9.2.

  • Thanks 1
Link to post
Share on other sites

  • RMweb Gold
10 minutes ago, Gordon H said:

 

It most certainly does have parity checking. The final byte in every packet is an XOR checksum of the preceding Address and Data Bytes - refer to NMRA S9.2.

 

OK thanks - my misunderstanding. Still not a good idea to have wiring issues that lead to packets not being read due to bad parity. 

Link to post
Share on other sites

On 02/08/2019 at 13:35, Junctionmad said:

On a layout that small , you could just connect two wires ! 

It doesn´t matter what size of the layout.

It´s digital wires and there is only two.

The most important is to have a common connection which you can after divided up by feeding wires to the track every second metres, which are ideal for the digital layout and analog too.

Link to post
Share on other sites

7 hours ago, Anders63 said:

It doesn´t matter what size of the layout.

It´s digital wires and there is only two.

The most important is to have a common connection which you can after divided up by feeding wires to the track every second metres, which are ideal for the digital layout and analog too.

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 

Link to post
Share on other sites

On 02/08/2019 at 13:17, Pete the Elaner said:

You have an error in your maths.

3.6/km = 3.6/100m = 0.036/m

..so 3ohms / 100m is close enough for our needs.

 

But I agree that you will see a larger resistance through rails & across joins.

 

Sorry to be pedantic Pete,  but whilst the overall figure of 0.036 Ω/m is correct, the maths inbetween has a small error.

[I got upset seeing 3.6/km = 3.6/100 which is incorrect]

 

The Rapid site says the resistance is 36.8 ohms per kilometre.  Actually this rounds to 37 ohms per metre to 2 sig. figs.  However, let's use 36 as previously stated.

 

A kilometre is 1000 metres, so that is 3.6 ohms per 100 metres or 0.036 ohms per metre.

 

This gives the figure (as you said) of approx 3 ohms per 100 metres (actally it's closer to 4) or 0.55 ohms per 50ft (=15.24m) or just about 0.1 ohms per 10ft (=3.048m) or just about an ohm for a 100 foot run.

 

This gives us a nice figure (for those of us that still think in feet and inches) of roughly 0.1 ohms per 10 foot of wire.

 

For those that work in metres and centimetres use 0.4 ohms per 10m.

 

Remember these figures are for 16/0.2  0.50mm²  20AWG  equipment wire

 

Art

Edited by Art Dent
  • Like 1
Link to post
Share on other sites

Interesting stuff set out above. As one whom is to start a DCC layout this autumn of a size of 8.0m x 6.0m approximately, I was going to use 2.5mm² copper as the bus with 1.5mm² droppers via Wago connectors - but what hasn't been mentioned above (unless I missed it), can one complete the bus as a 'ring main' - or is that a no-no? To me, it would minimise any voltage drop, the bus having no end, as it were.

 

(2.5mm² and 1.5mm² cable (in as many colours you can shake a stick at) over here, is €21.90 and €13.90 per 100m roll respectively.)

 

Cheers,

 

Philip

Link to post
Share on other sites

For the sake of completeness (and to help answer the OP's question) ...

 

Solid-core mains wiring has the following resistance:

 

1.0mm² has a resistance of 18.1 mΩ per metre  (= 0.018 Ω/m)
    or 0.006 Ω per ft,  0.055 Ω per 10 ft and 0.28 Ω per 50 ft length

 

1.5mm² has a resistance of 12.1 mΩ per metre  (= 0.012 Ω/m)
    or 0.004 Ω per ft,  0.037 Ω per 10 ft and  0.18 Ω per 50 ft length

 

2.5mm² has a resistance of 7.4 mΩ per metre  (= 0.0074 Ω/m)
    or 0.002 Ω per ft,  0.023 Ω per 10 ft and  0.11 Ω per 50 ft length

 

So it looks like even humble lighting cable (= 1.0mm²) would suffice for DCC Bus Wiring.

 

Art

Link to post
Share on other sites

11 minutes ago, Philou said:

Interesting stuff set out above. As one whom is to start a DCC layout this autumn of a size of 8.0m x 6.0m approximately, I was going to use 2.5mm² copper as the bus with 1.5mm² droppers via Wago connectors - but what hasn't been mentioned above (unless I missed it), can one complete the bus as a 'ring main' - or is that a no-no? To me, it would minimise any voltage drop, the bus having no end, as it were.

 

(2.5mm² and 1.5mm² cable (in as many colours you can shake a stick at) over here, is €21.90 and €13.90 per 100m roll respectively.)

 

Cheers,

 

Philip

Esentially - yes you can wire it as a complete ring.

 

After all, if you have an oval of track, what do the rails do??

 

Art

Link to post
Share on other sites

  • RMweb Gold
2 hours ago, Philou said:

As one who is to start a DCC layout this autumn of a size of 8.0m x 6.0m approximately, I was going to use 2.5mm² copper as the bus with 1.5mm² droppers via Wago connectors.

 

Whilst 2.5mm copper wire is fine for the bus, you don't need wire as thick as 1.5mm for droppers. First of all the droppers are going to be very short anyway, but more importantly you will have difficulty soldering wire this thick to the rails.  7x0.2mm stranded wire is fine for short droppers (up to about 12-18 inches) with anything longer ideally being 16x0.2mm.

Link to post
Share on other sites

4 hours ago, Art Dent said:

Esentially - yes you can wire it as a complete ring.

 

After all, if you have an oval of track, what do the rails do??

 

Art

 

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

Edited by newbryford
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...