Bi-directional DC ELV Control Circuit via 3-core cable

[JohnW2]

Hi there,

This is pretty basic! However, although I don't recall having thought about this before, I think I need some help in getting my thinking straight about what Mr Kirchoff would have to say about something I'm contemplating.

I have a control circuit which sends 12V DC to ('very') distant relay coil. I now need to add a further similar circuit, going 'in the opposite direction' between the same two locations, but routing a new cable would be a major mission, involving, for a start, extensive lifting of floorboards on two floors.

However, I note that the existing circuit uses just 2 cores of 3-core cable (flex), so I was thinking that I could simply use the 'spare' core for the new circuit, with one of the cores being common to both circuits. The present and contemplated situations are illustrated in the diagrams below.

The thing about the proposed arrangement is that current in the 'common' conductor would be flowing in opposite directions for the two circuits. If (as is far from impossible) both relays were activated simultaneous and if they were drawing the same current, the net current in that common conductor (the 'return path' for both relay coils) would be zero. I am therefore trying to get my head around the question of whether it would actually work.

A similar question would arise if we were talking about AC, but it's easier to think about with DC (and, indeed, DC is what I've got).

I suppose it would be very simple for me to get the answer 'by experiment' but, in the meantime, can someone help me get my thinking straight? Thanks!

Kind Regards, John

 

[DetlefSchmitz]

I don't see an issue provided both supplies are floating.

 

[bernardgreen]

 

[JohnW2]

I don't see an issue provided both supplies are floating.

Nor did I when I first contemplated the arrangement, but then I started thinking (perhaps too deeply?) about what current would (or would not!) be flowing in the common conductor. Time to 'conduct the experiment' to satisfy myself that my thought processes were being silly, I think!

Kind Regards, John

 

[DetlefSchmitz]

I think you are thinking too deeply. If you labelled the green line 0V at each end and at intervals all along then it suddenly looks clearer IMO.

 

[JohnW2]

<a diagram>

Thanks, bernard, but I am guilty of having 'over-simplified' for the purpose of asking the question!

Despite what I drew, the actual situation is not that there is a 12V PSU, with a switch on it's output, at both ends. The reality is that what is actually switched (by other things) at both ends is at mains voltage, and that switched '230V' feeds a 12V PSU, the output of which is permanently connected to the interlinking cable. For that reason, your 'single PSU' approach obviously cannot be used.

The reason for this is that the 'mains' at the two ends are on different phases, and I don't want to unnecessarily bring one phase into the 'territory' of another phase.

Kind Regards, John

 

[JohnW2]

I think you are thinking too deeply. If you labelled the green line 0V at each end and at intervals all along then it suddenly looks clearer IMO.

Yes, that's what I originally did - I just thought of the entirity of the green line as being "0V", and then all then looks/feels fine. It's only when I start thinking about currents that the mental rot sets in!

However, although I don't doubt that the reason is that I am "thinking too deeply", I confess that I'm still having trouble getting my head around it! On the face of it, when looked at from the viewpoint of a relay coil (with both relays activated), it is being 'fed' by a conductor carrying it's operating current, but with the return path back to the PSU carrying zero current - as I implied, something that Mr Kirchoff would not like!

Kind Regards, John

 

[DetlefSchmitz]

Kirchoff said that the sum of current at any node must be zero. It is in your case. The hard bit is to visualise why the current travels in the correct direction. If your 0V line was a ground plane on a pcb you would not even be able to draw it in a confusing manner.

 

[JohnW2]

Kirchoff said that the sum of current at any node must be zero. It is in your case. The hard bit is to visualise why the current travels in the correct direction. If your 0V line was a ground plane on a pcb you would not even be able to draw it in a confusing manner.

That's very true - and is the very reason why I was initially perfectly comfortable thinking of 'everything green' as simply being "0V" (i.e. in effect thinking of it as a 'ground plane'). I'm going to think and scribble a bit more, and then will probably show my face again.

Kind Regards, John

 

[DetlefSchmitz]

OK.
But divide it into 2 questions:
1) Will it works? YES
2) What happens on the common bit of wire? Are there 2 different sets of electrons each going in opposite directions, or do no electrons at all flow in that piece of wire? Or...

 

[JohnW2]

OK. But divide it into 2 questions: 1) Will it works? YES ...

Indeed, and that's what I have assumed all along - it's when I started "thinking too deeply" that I got onto the "Why" (theoretically-speaking) question.

... 2) What happens on the common bit of wire? Are there 2 different sets of electrons each going in opposite directions, or do no electrons at all flow in that piece of wire? Or...

No electrons moving at all, I presume (assuming relay currents are identical). I think I've now convinced myself about the theory (see diagram below). When relay currents are zero the, per Kirchoff, the flow into/out-of either end of 'interconnecting conductor' is zero, hence presumably no reason why any electrons should have (net) movement.

Kind Regards, John

 

[Harry Bloomfield]

That will work absolutely fine, you could even use just one single pair using DC one way and AC the other - with a bit of circuit complexity. An issue you might have with DC is if the cable is extremely long, the relay may prove to be reluctant to release. As happened to me when installing a control system for a reservoir on Anglesey, the two were several miles of cable apart. Solution was to use AC down the cable and a rectifier at the DC relay coil.

 

[DetlefSchmitz]

@JohnW2: I'm glad you're happy. I don't think it's easy to think about because there are so many different levels of physics involved. From 'conventional' current, to what actually persuades an electron to move. Potential gradient at that 'physics' level is a rather obscure characteristic.

At a classical level, I think that accepting that all wire has a finite resistance helps somewhat. At least with that in mind, there is always an uphill and downhill.

 

[DetlefSchmitz]

when installing a control system for a reservoir on Anglesey, the two were several miles of cable apart

Thank goodness John's house is not quite that big!

 

[JohnW2]

That will work absolutely fine ...

Indeed. I would just have 'done it' if I hadn't started "thinking too deeply" about the theoretical aspects

... you could even use just one single pair using DC one way and AC the other - with a bit of circuit complexity....

True, but if one were contemplating "a bit of circuit complexity", one could take that one step further and have umpteen control circuits working over a single pair!

.... An issue you might have with DC is if the cable is extremely long, the relay may prove to be reluctant to release. As happened to me when installing a control system for a reservoir on Anglesey, the two were several miles of cable apart. Solution was to use AC down the cable and a rectifier at the DC relay coil.

No, the cable is not 'extremely long' - probably 25-30 metres, if that. In any event, the 'one way around' system has been working (with DC) for very many years, without ever having got 'stuck (as far as I am aware) - perhaps helped by the fact that the nature of what it's doing is such that the relay will never be energised for very long period of time.

Mind you, on reflection, I'm not sure why the cable length should have a bearing on that potential problem - do you know what mechanism is/was responsible for what you describe?

Kind Regards, John