Can I automatically switch a fridge between 2 sockets?

[RF Lighting]

Ah I see. I'm thinking a manual changeover switch and a power failure buzzer might be easier

 

[JohnW2]

Ah I see. I'm thinking a manual changeover switch and a power failure buzzer might be easier

Much easier, yes (with readily available components) - but in my case (and I presume also the OP's, if his house is unattended/unoccupied for significant periods, since he's talking about a fridge/freezer), I need it to be able to work when no-one is around to do it manually.

As I said earlier in the thread, the infuriating thing is that it ought to be so easy - just per your diagram (give or take the earth switching!). The reason it isn't, if one wants decent safety margins, is that (even though I can't believe it's that rare a requirement), there just don't seem to be (m)any 'readily available' and 'heavy duty' off-the-shelf changeover relays/contactors - by which I mean ones with a long travel and hence a very significant distance between the fixed contacts (manual c/o switches often have 'inches', not a millimetre or four, between the fixed contacts). One can fabricate such a contactor by mechanically interlocking two together - but that's about the nearest I've found (apart from a couple of possibly iffy things from China!).

Kind Regards, John

 

[RF Lighting]

My first thought was to re-jig the contactors from a star delta starter, but cost is going to be an issue there.

 

[JohnW2]

My first thought was to re-jig the contactors from a star delta starter, but cost is going to be an issue there.

Yes, there are various 'ad hoc' possibilities like that - but I'd really prefer to stick with at least vaguely 'off-the-shelf' components (which didn't need re-jigging) - not the least for the benefit of anyone who may inherit the installation, or care of the installation, at some point in the future. ... and, yet again, I remain surprised by how difficult it seems to find any such off-the-shelf solutions.

Kind Regards, John

 

[SimonH2]

My word, you lot are making it complicated !

Most off the shelf contactors have a mechanical interlock option - just clip a contactor on each side of the interlock unit and job done. If you don't care which supply the load runs from, then simply connect each coil to the supply that unit switches via a NC contact on the other contactor (or the interlock unit). If one supply is present, then that contactor will pull in and supply the load. If both supplies are present, then whichever was present first will continue to supply the load.

The addition of a single relay (timer if you want a delay before switchover happens) will create a priority where the system will switch to a specific supply when both are present.

Eg, interlock for Telemechanique :
http://docs-europe.electrocomponents.com/webdocs/10e0/0900766b810e02e7.pdf
For Allen Bradley :
http://uk.rs-online.com/web/p/interlock-switches/2381137/
Or ABB :
http://uk.rs-online.com/web/p/contactor-accessories/4560304/


For using a single relay, then there are types where clearance is guaranteed - ie it's impossible for any NO contact to close while any NC is closed, or vice-versa. I believe these are referred to as "guided contact". With a standard relay, it's possible for (say) a NO contact to weld and hold together when the relay drops out - which means you end up with one NO contact and one NC contact closed. So you end up (in this situation) with a live from one supply and a neutral back through the other.


As to the earth, surely switching the earth is a really bad idea - prohibited by regs in the UK ? Would this be one of those situations where going TT and providing your own earth would be better ?

 

[JohnW2]

My word, you lot are making it complicated ! ... Most off the shelf contactors have a mechanical interlock option - just clip a contactor on each side of the interlock unit and job done.

If that 'most' is true, I'm having a devil of a job finding them. Also,one tends to end up with pretty large assemblies (particularly with the Schneider/Telemechanique ones) if one has to use two contactors and a interlock unit.

If you don't care which supply the load runs from, then simply connect each coil to the supply that unit switches via a NC contact on the other contactor (or the interlock unit). If one supply is present, then that contactor will pull in and supply the load. If both supplies are present, then whichever was present first will continue to supply the load.

...and if both supplies appear simultaneously? I do have some difficulty in understanding what mechanical interlocks do in that situation, hence...

The addition of a single relay (timer if you want a delay before switchover happens) will create a priority where the system will switch to a specific supply when both are present.

Indeed- I've already acknowledged that.

For using a single relay, then there are types where clearance is guaranteed - ie it's impossible for any NO contact to close while any NC is closed, or vice-versa. I believe these are referred to as "guided contact".

As I've said, I would have expected them to exist - but have not yet found any. Maybe searching for "guided contact" (a term I've never heard before) might do the trick.

With a standard relay, it's possible for (say) a NO contact to weld and hold together when the relay drops out - which means you end up with one NO contact and one NC contact closed. So you end up (in this situation) with a live from one supply and a neutral back through the other.

Indeed. In my arrangement, I look to the 'RCD functionality' of the RCBOs to deal with that one.

As to the earth, surely switching the earth is a really bad idea - prohibited by regs in the UK ? Would this be one of those situations where going TT and providing your own earth would be better ?

Totally agreed. As I said, in my situation, the issue does not arise (only one earth). Even in the OP's case, although there are 'two supplies', there may not be two earths. Whatever, as you say, switching the earth is a definite no-no .

Kind Regards, John

 

[SimonH2]

If that 'most' is true, I'm having a devil of a job finding them.

It took me only a few seconds to come up with the three I linked to !

Also,one tends to end up with pretty large assemblies (particularly with the Schneider/Telemechanique ones) if one has to use two contactors and a interlock unit.

They aren't that big - but it does vary.

...and if both supplies appear simultaneously? I do have some difficulty in understanding what mechanical interlocks do in that situation, hence...

The mechanical interlock will physically prevent the slower unit from pulling in, even though it will have it's coil energised. That is what they are designed to do !

The addition of a single relay (timer if you want a delay before switchover happens) will create a priority where the system will switch to a specific supply when both are present.

Indeed- I've already acknowledged that.

The relay won't prevent both pulling in together. The configuration I'd be looking to use would put the supply of the timer relay across the priority supply. When it pulls in (either directly or after a time delay) it opens the coil supply to the second contactor (on the non-priority supply). Thus if you apply both power supplies simultaneously, you will apply power to both contactor coils until either one of them is pulled in, or the relay has pulled in.

 

[bernardgreen]

The mechanical interlock is designed to prevent both contactors closing when there is a fault in the control circuit.

An AC coil has a lower impedance when the armature is not pulled in because the magnetic path is not closed. As a result the current in the coil of the contactor that has been prevented from closing will be higher than normal.

This could lead to overheating if that contactor coil remains energised with it armature not pulled in.

 

[JohnW2]

The mechanical interlock will physically prevent the slower unit from pulling in, even though it will have it's coil energised. That is what they are designed to do !

I realise that - but my uncertainties are about what happens if neither is significantly slower than the other - is there not a possibility of the mechanical interlock mechanism simply 'jambing' with neither contactor having changed state if both coils are energised simultaneously (as would happen, in my case, on restoration of power after an all-phases power cut)?

Kind Regards, John

 

[JohnW2]

The mechanical interlock is designed to prevent both contactors closing when there is a fault in the control circuit. ... This could lead to overheating if that contactor coil remains energised with it armature not pulled in.

Theoretically true (although I'm not sure the coil impedance would be all that much different with the armature pulled in/not, would it?) - but, in practice, that situation would presumably only arise if the electrical interlock had failed to do its job. Short of adding in further redundancy into the control circuit, I'm not sure what you're suggesting one can/should do about that possibility.

Kind Regards, John

 

[JohnW2]

[... (although I'm not sure the coil impedance would be all that much different with the armature pulled in/not, would it?)

I've done an experiment and have to agree that the difference can be quite substantial. With a small relay with a 230V AC coil, the current (from my mains supply, ~245V) was about 10.3 mA when normally energised, but about 16.5 mA when energised but with the armature prevented from moving. That's roughly a 1.6-fold increase in current, which would correspond to about a 2.6-fold increase in power dissipation - which I agree might be enough to result in overheating.

Kind Regards, John

 

[SimonH2]

... is there not a possibility of the mechanical interlock mechanism simply 'jambing' with neither contactor having changed state if both coils are energised simultaneously

Good question.
In part I suppose it depends on the mechanical design, and also how far it allows the armature to move before blocking. If the movement is enough to open the interlock contacts then it wouldn't an issue - at worst they would pull in and drop out until one of them got ahead of the other, and I really can;t see you finding an interlocked set well enough matched for them not to have enough of a bias.

One way to avoid it completely would be a delay-on relay on one of the contactors. However, that still theoretically leaves a small window if both supplies are restored, but one with a delay equal to the on-delay of the relay.
Adding a priority relay - coil across one supply, NC contact in series with the other contactor coil - would guarantee operation regardless.

So shopping list is :
2 contactors, one interlock unit, one relay, some terminals, bit of DIN rail to mount them on, and a box to put them in (plus sundries like cable glands etc).

 

[JohnW2]

... is there not a possibility of the mechanical interlock mechanism simply 'jamming' with neither contactor having changed state if both coils are energised simultaneously

Good question. In part I suppose it depends on the mechanical design, and also how far it allows the armature to move before blocking. If the movement is enough to open the interlock contacts then it wouldn't an issue - at worst they would pull in and drop out until one of them got ahead of the other, and I really can;t see you finding an interlocked set well enough matched for them not to have enough of a bias.

Quite .... as you say, 'it depends' - particularly on whether or not the mechanical interlock allows enough movement to open the interlock contacts (if they don't the whole thing could stay 'jammed', with neither relay having changed state - hence improper functionality and, possibly, the overheating problem raised by Bernard). ...

One way to avoid it completely would be a delay-on relay on one of the contactors. .... So shopping list is :
2 contactors, one interlock unit, one relay, some terminals, bit of DIN rail to mount them on, and a box to put them in (plus sundries like cable glands etc).

... and all that just because it's impossible or very difficult to find an (ideally 'compact') dedicated changeover relay/contactor with decent contact separation/travel and what you described as "guided contacts". I remain surprised that such dedicated animals either don't exist or are very difficult to find, since it's not a particularly unique or rare functional requirement we're talking about.

For my application, I'm perfectly happy with what I've got (as described on Sunday) - which had a comparable 'shopping list' (and assembly work) to what you propose. It's been in service for around 20 years without problems (during which period it has satisfactorily and correctly dealt with a good few losses of one or both phases) - and, although I realise that 'working without problems' does not prove everything, unless I've missed something, I'm pretty happy (per what I wrote) that the design has addressed most possible scenarios/malfunctions/failure modes. Do you see any significant potential problems with it (which don't involve invoking exceedingly unlikely combinations of circumstances!)?

Kind Regards, John

 

[SimonH2]

Of course, some contactors have NC contacts and/or the option of adding NC aux switch blocks. You'd have to check both the rating of the NC contacts (usually significantly lower than the NO main contacts) and whether there is reliable break-before-make operation (particularly in the case of welded contacts).
Having looks at relays with guided contacts, it seems these are mostly designed for control circuits and have fairly low switching capacity.

As to what happens with an interlock if both sides are energised together - you'd have to ask the manufacturer. But I'm sure they must have considered it.

 

[JohnW2]

Of course, some contactors have NC contacts and/or the option of adding NC aux switch blocks. You'd have to check both the rating of the NC contacts (usually significantly lower than the NO main contacts) and whether there is reliable break-before-make operation (particularly in the case of welded contacts). Having looks at relays with guided contacts, it seems these are mostly designed for control circuits and have fairly low switching capacity.

All I can say is to repeat my surprise that I don't seem to be able to find any 'readily available, off-the-shelf' dedicated changeover contactor - which would remove all these issues and the need for any discussion.

As to what happens with an interlock if both sides are energised together - you'd have to ask the manufacturer. But I'm sure they must have considered it.

If a satisfactory changeover configuration were available, this question would obviously not arise - 'interlock' would be an intrinsic feature of the contact arrangement. As for the mechanical interlock, unless they can guarantee that it would always allow enough movement for at least one of the sets of interlock contacts to open, it's very hard to see how they could be certain how the mechanical interlock would behave (and sure that it could not 'jam') under these circumstances.

Kind Regards, John