Common Return Explained (I hope)

[AndyID]

The dreaded common return has raised its head yet again. Rather than just harping on with "do this" and "don't do that" I thought it might be worth a bit of explanation about why it actually works. I understand why it can seem confusing because it some respects it is a bit counter-intuitive.

 

Anyway, here goes. This is starting at the most fundamental level. If it turns out to be helpful please add to it

 

Figures 1 and 2 show batteries driving motors. No mystery there. In figure 3 the negatives of both batteries are tied together via a common connection. I doubt if anyone would have a problem with that. In figure 4 one of the batteries has been reversed to make its motor rotate in the opposite direction. Still not a problem really because there is no apparent reason for any current to flow through the common connection.

 

 

 

Then we come to figure 5. The common connection is now the current path for both motors. In other words it is a "common return". Obviously it's much more complicated than figures 3 and 4 because the currents are flowing in opposite directions in the common return so they must interfere with each other.

 

And I think that''s where most of the confusion arises. Electrically there isn't the slightest bit of difference between figure 3, figure 4 and figure 5. The electrons sort it out all by themselves and we don't have to worry about it.

 

I've deliberately not mentioned anything about types of controllers etc, etc. They have little to do with the fundamentals of common return.

 

 

[Sitham Yard]

6 hours ago, AndyID said:

I've deliberately not mentioned anything about types of controllers etc, etc. They have little to do with the fundamentals of common return.

 

 

Am I correct in saying that you must use separate a transformer for each controller?

Andrew 

[cctransuk]

28 minutes ago, Sitham Yard said:

Am I correct in saying that you must use separate a transformer for each controller?

Andrew 

 

You must use separate transformer WINDINGS; a twin output transformer with separate windings is perfectly fine.

 

CJI.

[BoD]

Is the only reason for having common return to save on wiring or is there something else?

[kevinlms]

30 minutes ago, BoD said:

Is the only reason for having common return to save on wiring or is there something else?

You can save by using single pole switches, or use the 2nd pole for something else, such as indicator lights on a control panel.

[AndyID]

5 hours ago, BoD said:

Is the only reason for having common return to save on wiring or is there something else?

 

Not only does it save on wiring it significantly improves reliability by reducing the number of connections and switch contacts. It also makes fault-finding much simpler.

 

That said I'm not trying to sell common return. I'm just trying to demystify it.

[AndyID]

9 hours ago, Sitham Yard said:

Am I correct in saying that you must use separate a transformer for each controller?

Andrew 

 

The power supplies are represented by batteries in the diagrams. That's because it's obvious that batteries can "float" and be either positive or negative relative to some reference.

 

With controllers they have to be able to float too and one way to achieve that is by supplying them from independent secondary windings (as CJI said) on one or more transformers.

 

There are ways to do it that don't require floating anything but as they are not compatible with commercially available controllers that would be best discussed in a different thread.

 

[Crosland]

Switch mode supplies that do not have an earth connection (e.g. a two pin mains connector or a 13A plug with plastic earth pin) can also be treated as floating, but DO NOT simply use a plastic earth pin with a PSU that is designed to be earthed.

[AndyID]

8 minutes ago, Crosland said:

Switch mode supplies that do not have an earth connection (e.g. a two pin mains connector or a 13A plug with plastic earth pin) can also be treated as floating, but DO NOT simply use a plastic earth pin with a PSU that is designed to be earthed.

 

Yes, NEVER defeat the earth connection on equipment that is supposed to be earthed. Doing so creates a serious safety hazard.

 

Just a bit of info; Modern power "bricks" and "wall warts" might not look as if they include transformers but they actually do. The transformers are much smaller than the  traditional 50 Hz variety because they are operating at much higher frequencies.

 

 

[dasatcopthorne]

So.

 

I have two switch mode power supplies. Each has an Earth tag in the connector block.

Can I use these two in a common return configuration please?

 

Dave

[Grovenor]

Yes, if you are using for accessories with the 0V (-ve) of each power supply commoned together.

NOT for powering DC train controllers where the 12V (+ve) may get joined to the 0V (-ve) of the other via the reversing switches and the track.

Switch mode power supplies with an earth connection will have the DC output 0V or -ve connected to the earth.

[AndyID]

5 hours ago, dasatcopthorne said:

So.

 

I have two switch mode power supplies. Each has an Earth tag in the connector block.

Can I use these two in a common return configuration please?

 

Dave

 

Please post a pic of the PSUs and some more information on their specifications.

 

Thanks!

[dasatcopthorne]

I have two of these Switching Power supplies.

 

The connections show are top to bottom: mains in, mains in, earth, 0vdc out, +12vdc out.

The earth tag tests as connected to the metal body and NOT the inputs or outputs when the unit is not live.

 

I have the 0vdc of one connected to the +12vdc of the other.

this gives +12, 0v, -12v and this is used to operated Tortoise point motors and nothing else.

 

The two Earths are currently connected to a Kettle type mains lead having a 3amp fuse.

 

All these and a chassis type transformer are housed in a plastic project box.

 

Should I retain or remove the earth connections to these two please?

 

Dave.

 

 

 

 

Edited March 14, 2023 by dasatcopthorne

[DCB]

When you think of it Common return is the oldest form of wiring as the old 3 rail systems used it

 

1 hour ago, dasatcopthorne said:

I have two of these Switching Power supplies.

 

Should I retain or remove the earth connections please?

 

Dave.

 

. I really hope you have some overload protection on those outputs

[dasatcopthorne]

59 minutes ago, DCB said:

When you think of it Common return is the oldest form of wiring as the old 3 rail systems used it

 

. I really hope you have some overload protection on those outputs

Good thinking. It's what concerned me the most when deciding to use them.

 

They have their own internal.

 

Dave.

Edited March 14, 2023 by dasatcopthorne

[AndyID]

4 hours ago, dasatcopthorne said:

 

Should I retain or remove the earth connections to these two please?

 

 

DO NOT REMOVE THEM!

 

Sounds like the outputs are floating and they are suitable for common return. But they output a lot of current.

 

Are you planning to use them to power locomotives?

 

What type of controller will you use?

[dasatcopthorne]

25 minutes ago, AndyID said:

 

DO NOT REMOVE THEM!

 

Sounds like the outputs are floating and they are suitable for common return. But they output a lot of current.

 

Are you planning to use them to power locomotives?

 

What type of controller will you use?

 

As I said AndyID. They are not used for anything else (and will never be).

 

The layout uses DCC.

 

Cheers

 

Dave.

 

 

[5BarVT]

21 hours ago, dasatcopthorne said:

Should I retain or remove the earth connections to these two please?

As Andy said, don’t remove the earths as they are a legal requirement with a metal cased supply.

You have done the ultimate ‘can I use them as common return’ test by connecting 0V on one to 12V on the other and they have passed.  ‘Not passed’ would have resulted in a big bang with lots of smoke, so although it is proof that they work in common return, it is not a methodology I would want to use without knowing that it would work first!

Paul.

[roythebus1]

I've used common return wiring for well over 50 years on most layouts I've built, including the MRC's New Annington exhibition layout. that one threw up an interesting problem which I've recalled elsewhere, but will repeat here.

 

the MRC's HQ "Keen House" is a late 1950s industrial type building built on 3 floors with 3-phase supply. Originally one phase supplied lower hall, one the upper hall and the other the library and caretaker's flat. This was causing an imbalance on the 3-phase meters as the upper hall use more than the others. The place was re-configured in the late 1980s to even out the supply problems and it was around that time that plugs with RCD or ELCBs became available. We used them on every extension lead from the high level 13 amp ring mains down each side of the hall for added safety.

 

Sometimes every RCD in the hall would trip, nobody could work out why. Here the plot thickens. the club layout had a ring main 12 volt supply for point motors, signals etc with an exposed thick copper wire for common return. that worked ok until we put overhead line equipment in. The catenary wasn't live nor were the masts. One evening I was soldering something and suddenly a bit of catenary melted! Quite odd as it was electrically isolated from everything so we thought. At the same time all the RCDs tripped out. We found one of the mounting screws had contacted the common return under the board. But a thorough check and test found no obvious fault.

 

We decided to test all the high level sockets in the lower hall. This was done with the power still on initially and this is where the fun began. Yours truly got a "belt" of mains from one of the metal socket boxes! All the RCDs tripped out. A close inspection found that whoever had installed the mains in 1959 had trapped one if the live wires between the socket plate and the box, causing an occasional "live to earth" fault. The circuit was made via my soldering iron, the trackwork and common return which caused the OHLE to melt in places! Luckily the RCDs prevented me or anyone else getting seriously injured. 

 

Apart from that, the common return does save a lot of wiring. I don't "do" DCC, far too complicated and anything computerised never works as it should do. I want to see trains run, not see operators fiddling with a computer keyboard all day. Switches and relays are far easier to fault-find. 

 

 

[kevinlms]

Perhaps I've misunderstood you, but what has common return have to do with the building earth?

[DCB]

I gave up on common return some years ago.   I had this random fault.  The outer main line would short and shut down every now and again, really annoying, usually with a train in the up main platform loop, and  when I changed the points for the Up back platform.  It took ages but I found it, it was live frog points, and a same side short usually from a loco with tender pickups.  I had to fit a point tie bar operated isolating switch and an insulated fish plate to cure it.  I should have insulated the lower rail on the platform loop not the upper, but the upper was the common return.

[kevinlms]

3 hours ago, DCB said:

I gave up on common return some years ago.   I had this random fault.  The outer main line would short and shut down every now and again, really annoying, usually with a train in the up main platform loop, and  when I changed the points for the Up back platform.  It took ages but I found it, it was live frog points, and a same side short usually from a loco with tender pickups.  I had to fit a point tie bar operated isolating switch and an insulated fish plate to cure it.  I should have insulated the lower rail on the platform loop not the upper, but the upper was the common return.

So the fault wasn't anything to do with Common Return, but the fact you parked a train over a gap. Gaps should be moved if necessary to prevent such shorts, especially if that is where you park trains.

Any gap with opposing polarities, is going to cause trouble if a train is parked over it.

I'm surprised you didn't blame the problem on the tender pick ups!

[AndyID]

22 hours ago, DCB said:

I should have insulated the lower rail on the platform loop not the upper, but the upper was the common return.

 

 

You lost me completely :)

 

You "never" put an IRJ in the common return rail. That rail should always be either the inner rail or  the outer rail on the entire layout and never both on the same layout.

 

The only exceptions I know of are when you have to add an IRJ where the common-return rail meets a switched frog and when you have to isolate both rails in a reversing loop.

 

You said it yourself. It's just like Dublo 3-rail. The difference is you designate one rail on the layout as equivalent to the two outside common-return rails on a HD 3-rail layout and the other rail is equivalent to the HD center rail, and it's the only rail that ever gets isolated.

[Junctionmad]

On 17/03/2023 at 18:57, roythebus1 said:

 

Apart from that, the common return does save a lot of wiring. I don't "do" DCC, far too complicated and anything computerised never works as it should do. I want to see trains run, not see operators fiddling with a computer keyboard all day. Switches and relays are far easier to fault-find. 

 

I'm sorry this luddism. DCC works extremely well when installed and configured by people "who know what they are doing TM", like all things some basic knowledge mst be acquired.

Edited March 20, 2023 by Junctionmad

[AndyID]

42 minutes ago, Junctionmad said:

I'm sorry this luddism. DCC works extremely well when installed and configured by people "who know what they are doing TM", like all things some basic knowledge mst be acquired.

 

There's a lot more to it than that but please try to keep this thread on topic.

 

There are plenty of other threads that discuss the pros and cons of DCC v DC. This one is supposed to be about one particular aspect of DC control that some find a bit confusing.