Dumb question about bus wiring lenght

[gavino200]

This question is so dumb I don't even know how to search it. I've read and watched a lot about bus wiring at this stage and I don't have an answer. I'm sure it's obvious beyond belief to those with an technical or engineering background, but not to me. So, here it goes.

 

Does it matter where you connect your power input to a bus? How does that affect what you consider to be the length of the wire?

 

Let's say I have a layout that's 50 feet long with a single loop of track. I'm going to put my command station right in the center, 25 feet from both ends. Do I consider this to be a 25 foot long bus? Or a 50 foot long bus?

 

I'm assuming that it's the longest single span of wire that's important in choosing what gauge to use. So I'd consider this hypothetical bus to be a 25 foot bus when choosing a wire gauge. Would I be right?

[katoftw]

40 foot bus with up to 5 feet branches.

[gavino200]

1 minute ago, katoftw said:

40 foot bus with up to 5 feet branches.

 

What you talkin'bout Willis?

 

(Can you expand on that?)

[katoftw]

I would make your main bus 40-45 foot in length. And all the smaller branching wires 2-5 feet off that bus.

[gavino200]

Ah, I see. No, the question was even more basic than that. It's about wire gauge. Given the 40 foot but you describe. If you hook the booster connections up to it right in the middle - at the 20 foot mark, you consider that a 20 foot bus for purposes of deciding wire gauge. It's the longest span of wire from where the power wires are connected, right? So 20 foot in each direction. You use the appropriate gauge for a 20 foot bus. Right?

[nicolahu]

Yes and no, yes what matters is the longest cable you install from the source. Because it is the longest way electricity has to travel. But also no, if your cable is 20 foot long the distance the power has to travel is at least 40 foot. One way to the track and back to the source.

So if you wanna calculate the resistance for choosing the right wire gauge, take the longest distance of cable and double it.

Hope that helps a bit xD

[Dani]

Hello,

 

I'm not electrician, and for sure there are expert ways to calculate the wire gauge.  I just share what it works for me (doesn't matter if it is right or wrong). We are using 1,5mm2 (AWG15) wire for the main bus (DCC bus, 5Vcc bus and 12Vcc bus), and 0,5mm2 (AWG20) to connect rails, lights or whatever to the main bus.

 

It works well with just 3 modules connected (2 or 3 meters long), or more that 50 meters long. We had no issues, and wires got never hot.

[gavino200]

4 hours ago, nicolahu said:

Yes and no, yes what matters is the longest cable you install from the source. Because it is the longest way electricity has to travel.

 

That's what I thought. But in matters where I don't have expertise, I make it my practice to doubt my assumptions.

 

4 hours ago, nicolahu said:

But also no, if your cable is 20 foot long the distance the power has to travel is at least 40 foot. One way to the track and back to the source.

So if you wanna calculate the resistance for choosing the right wire gauge, take the longest distance of cable and double it.

 

Interesting. I hadn't considered that. I'm going by the chart here on the DCC wiki. It just refers to "Bus length". There's no discussion of how to measure bus length. 

 

So I guess now there are two questions:

1. Is it the longest single span if you connect the power anywhere except the end of the bus

2. Do you consider both wires "There and back again"?

Edited April 10, 2022 by gavino200

[gavino200]

2 hours ago, Dani said:

Hello,

 

I'm not electrician, and for sure there are expert ways to calculate the wire gauge.  I just share what it works for me (doesn't matter if it is right or wrong). We are using 1,5mm2 (AWG15) wire for the main bus (DCC bus, 5Vcc bus and 12Vcc bus), and 0,5mm2 (AWG20) to connect rails, lights or whatever to the main bus.

 

It works well with just 3 modules connected (2 or 3 meters long), or more that 50 meters long. We had no issues, and wires got never hot.

Thanks. For me it's a choice between 14 or 16 AWG for my longest bus. I doubt there'd be much difference. I bet I'd be fine with 16 for everything and 20 for feeders. But now I'm curious.

[cteno4]

Gavin,

 

I would talk to folks that know their ee and do dcc a lot about buss gauge needs, I’ve seen a lot of lore (not sure how much based on good data or ee) that you want to over gauge your buss wire on dcc to get the best signals for dcc. I’ve not, so far, been looking to go dcc so scan dcc stuff to keep up somewhat but not carefully study and remember it all totally. From what I recall it was about dcc signals doing better on the main buss being larger gauge and to stay away from solid core wires. I think the over sizing of the cable helps from when you have too many amps in a smaller gauge wire it can create eddie currents that can interfere with the dcc signals some. It’s foggy but check.

 

the buss has to be beefier on dcc as it can be drawing more amps with many trains running at once, whereas the feeders are only really ever drawing for a single train at once ever really if placed every 3-6’ apart (path of least resistance). With regular dc trains you usually only have one train per circuit so you don’t need much beefier busses, but extra gauge can help overcome voltage losses a bit in some situations. 
 

you can determine your max amps for just shoving current through wire with standard awg tables. Realize the more cores your stranded wire has the lower the amps you can safely pump through them. All this also depends on if wires are bundles, insulation material and what is cooling permitted in the space. My rule of thumb for regular hobby low voltage stuff is to run at half the awg charts to just be safe and no worries.

 

https://www.engineeringtoolbox.com/wire-gauges-d_419.html

 

jeff

[Martijn Meerts]

For my bus on the yard and helix, I use common multi-strand electrical wire, so no solid core. I have no clue what gauge those are over here, but they're not overly beefy, but definitely beefier than what's usually sold as model railway wire. For the helix I also have it in a full loop, for the yard it's just a simple length of wire.

 

I have read that doing it in a loop is better, but I can't remember why exactly. I've also read that a bus wire can cause issues with occupancy detection if those wires are close to the bus wires, something with the bus wires generating a magnetic field, but so far I've not had any issues with that on either my current "layout", or any of the previous ones I had that had bus wires.

 

I'm sure there's some truth to a lot of the myths about wiring, but I've not had any of the issues mentioned in those myths, or at least, not to a point where things were getting unreliable.

 

You might want to have a look at https://www.wiringfordcc.com/. The site looks a bit dated and can be a bit annoying to navigate, but the info should all still be applicable.

 

[gavino200]

20 hours ago, cteno4 said:

Gavin,

 

I would talk to folks that know their ee and do dcc a lot about buss gauge needs, I’ve seen a lot of lore (not sure how much based on good data or ee) that you want to over gauge your buss wire on dcc to get the best signals for dcc. I’ve not, so far, been looking to go dcc so scan dcc stuff to keep up somewhat but not carefully study and remember it all totally. From what I recall it was about dcc signals doing better on the main buss being larger gauge and to stay away from solid core wires. I think the over sizing of the cable helps from when you have too many amps in a smaller gauge wire it can create eddie currents that can interfere with the dcc signals some. It’s foggy but check.

 

Yes. There's contradicting advice about solid vs. stranded. I'm using stranded because it's easy to manage. They say to twist it 3 times per foot. to avoid inductance currents. I learned from an interesting video that the purpose of this is to keep the wires as close together as possible, and that's what makes the difference.

 

I definitely don't want to undergauge, as my layout is a tad long though not at all deep. Pretty much every DCC wiring site has AVG suggestions for different gauges, and lengths. But no one so far answers the "How to measure bus lenght" question. But the search goes on. 

 

 

20 hours ago, cteno4 said:

 

the buss has to be beefier on dcc as it can be drawing more amps with many trains running at once, whereas the feeders are only really ever drawing for a single train at once ever really if placed every 3-6’ apart (path of least resistance). With regular dc trains you usually only have one train per circuit so you don’t need much beefier busses, but extra gauge can help overcome voltage losses a bit in some situations.

 

Yes, for sure. I'm not in a super hurry as I'm one step away from laying wire. I'd like to know how much beef is wasted beef. Less beef is easier to work with.

 

20 hours ago, cteno4 said:

you can determine your max amps for just shoving current through wire with standard awg tables. Realize the more cores your stranded wire has the lower the amps you can safely pump through them. All this also depends on if wires are bundles, insulation material and what is cooling permitted in the space. My rule of thumb for regular hobby low voltage stuff is to run at half the awg charts to just be safe and no worries.

 

https://www.engineeringtoolbox.com/wire-gauges-d_419.html

 

 

Choose a wire that can handle twice the amperage of what you assume to be your max current?

[gavino200]

14 hours ago, Martijn Meerts said:

For my bus on the yard and helix, I use common multi-strand electrical wire, so no solid core. I have no clue what gauge those are over here, but they're not overly beefy, but definitely beefier than what's usually sold as model railway wire. For the helix I also have it in a full loop, for the yard it's just a simple length of wire.

 

I have read that doing it in a loop is better, but I can't remember why exactly. I've also read that a bus wire can cause issues with occupancy detection if those wires are close to the bus wires, something with the bus wires generating a magnetic field, but so far I've not had any issues with that on either my current "layout", or any of the previous ones I had that had bus wires.

 

That's interesting, and new to me. I know that the NMRA recommends some special device or current to link the ends of a linear but. It has a capacitor in it. I'll post it when I find it. So I'm sure that linearity is important.

 

 

14 hours ago, Martijn Meerts said:

 

You might want to have a look at https://www.wiringfordcc.com/. The site looks a bit dated and can be a bit annoying to navigate, but the info should all still be applicable.

 

 

Thanks. I will. I'll parse this and also the NMRA guidelines. Probably won't get to it until the weekend. But I'll post what I find.

[gavino200]

So I guess now there are three questions.

 

 

1. Is it the longest single span if you connect the power anywhere except the end of the bus

2. Do you consider both wires "There and back again"?

3. Ideally should a bus be a full loop. And if not what's that device that the NMRA recommends you add to the end?

 

@Kiha66, any light to shed on this?

[Madsing]

Sorry, no answer, but I have question for you: Where do you plan to put the feedback modules (DR5088 if I remember correctly)? I really struggled about that when I started with block detection. The feedback modules can be scattered over the layout, each of them close to the blocks they monitor. In this case the main DCC bus from the command station will be long (the distance from the command station to the last feedback module), but the wires from the feedback modules to the tracks will be shorter. Some people choose to have all feedback modules next to each others, next to the command station. In this case the main DCC bus is very short, but you need longer wires to reach each block from each feedback module. This is the solution I have chosen, so my DCC wiring looks like a star, not a bus, with all wires starting from a central point and going to their track/block.

Do you see what I mean? I am asking because most of what I see online doesn’t apply to block detection. They just assume that we need a bus to distribute the same DCC signal to all blocks.

Your layout is large, so I’d think that having the detection modules scattered all over the layout will be the best solution, but it’s probably worth thinking about that.

[gavino200]

5 minutes ago, Madsing said:

Sorry, no answer, but I have question for you: Where do you plan to put the feedback modules (DR5088 if I remember correctly)? I really struggled about that when I started with block detection. The feedback modules can be scattered over the layout, each of them close to the blocks they monitor. In this case the main DCC bus from the command station will be long (the distance from the command station to the last feedback module), but the wires from the feedback modules to the tracks will be shorter. Some people choose to have all feedback modules next to each others, next to the command station. In this case the main DCC bus is very short, but you need longer wires to reach each block from each feedback module. This is the solution I have chosen, so my DCC wiring looks like a star, not a bus, with all wires starting from a central point and going to their track/block.

Do you see what I mean? I am asking because most of what I see online doesn’t apply to block detection. They just assume that we need a bus to distribute the same DCC signal to all blocks.

Your layout is large, so I’d think that having the detection modules scattered all over the layout will be the best solution, but it’s probably worth thinking about that.

 

Thanks Marc. Yes, I've thought about this and have a plan. But to be honest, I've made assumptions and haven't verified my assumptions. Below is a page from the DR5088 manual.

 

I've been considering the wires from the DR5000/Command station to the Dr5088s to be equivalent to the bus

I'm considering the block detection wires and the common wires from the track to the 5088s to be equivalent to feeders.

Then I'm just choosing wire gauge based on standard recommendations for bus wire and feeder wire.

I haven't verified the assumption that these were equivalent. Actually, until you asked, I wasn't conscious of these assumptions.

 

On my practice layout I placed all the electronic boxes together and let the detection wires fan out like a star.

It wouldn't make sense on the new layout as it's quite long. It makes more sense to make a long bus and place a series of DR5088s along the layout at convenient intervals to keep the "feeder" wires as short as possible. Each DR5088 would have it's own "star" spreading out to the tracks around them. It would be like "stars on a string".

 

In fact, strictly speaking I don't need to create a separate bus for the commuter and tram lines, unless I run out of power. Mainly I think that making a second bus for these would make troubleshooting the wiring easier if there were a problem.

 

I'll try to find out if block detection creates any particular considerations for bus design. Fortunately, I have time to do this as I tend to research the next project while executing the current project. Hooking up the wires is still one project away. 🙂

 

I'll also need a LocoNet bus, but I think that's relatively simple. Just LocoNet wire and hubs.

 

 

 

 

[Kiha66]

1. Is it the longest single span if you connect the power anywhere except the end of the bus

For the purpose of wire gauging, I would consider the longest wired distance from a power supply.  That will be your greatest power drop so that is the main concern.

 

2. Do you consider both wires "There and back again"?

Generally you only consider power to the locomotive.  If there is enough power to run trains on on DCC, return signals should make it just fine.

 

3. Ideally should a bus be a full loop. And if not what's that device that the NMRA recommends you add to the end?

The bus does not need to be a full loop, unless it's coinvent with your layout.  Generally I'd run a single bus under the layout from end to end and out any peninsula.  There shouldn't be a need for any special termination on a DCC bus, unless your specific base station or application recommends otherwise.  LCC Canbus networks do suggest a special termination, but I don't believe NMRA specifically recommends anything for DCC.

 

These are just my thoughts from my experience, I do look forward to seeing other's input.

[gavino200]

Thanks Kiha! You rock!! 🙂

[nicolahu]

4 hours ago, Kiha66 said:

 2. Do you consider both wires "There and back again"?

Generally you only consider power to the locomotive.  If there is enough power to run trains on on DCC, return signals should make it just fine.

Back again 😉

Maybe i can explain it a bit better why you should have both ways in mind.

So, this is a simple schematic. On the left we have our power supply, when we follow the upper line we can see Rw. This should demonstrate the resistance of the wire to the track, for this example all lines are super efficient and don't have a resistance. Then followed by the Rl, the load aka Locomotive and back again by the the resistance for the wire back to the power supply.

Then we have 4 arrows, three labelled with U and one with an I. U is voltage and I is Ampere, but why are there three U arrows?

Well this is a series connection, so the voltage is different over every resistance while the Ampere is the same over the whole. For parallel it is right the opposite.

Our power supply can deliver 12V (Volt) with a maximum of 3A (Ampere). So we imagine our load is a yard where we need our 3A and the wires are the bus lines to the yard. 

We have a 50 foot long cable to the yard with an AWG16, so the resistance to the yard is 0.2 Ohm or 0.4 for both ways.

Now we wanna calculate what is our voltage drop for that we can take this formula: U=R*I

0.2 Ohm * 3A = 0.6V

So we lose 0.6V on the way to the yard and a another 0.6V on the way back, that means if we would measure the voltage in the yard we would have 10.8V instead of the 12V.

Now remember, if our yard is empty there is virtually no load. So we would measure our 12V or very close to it, because our voltagemeter only draws an very low load of maybe 0.01A

This calculation is of course a worst case scenario, because in real life those 3A will be more distributed over several cables or feeder along the bus cable. But if you choose the right cable for the worst case scenario, you should be fine.

 

@gavino200https://dccwiki.com/Wire_Sizes_and_Spacing

this chart already has both ways in mind as he also states in the explenation above:

"Remember, an equal length of wire is needed to complete the circuit, doubling the resistance of the circuit."

 

In the end there are more factors involved but these exceed my knowledge xD

I think dccwiki has a good chart overall and should serve fine.

If any questions arise i will try to explain them if i can.