power supply? Not sure!
Make sure that (2nd) diagram isn't controversial. (Nor the first for that matter.)
It caused a great stir down here a few decades back. (Industry people said it was "illegal".)
Yes, there are 240VAC relays (as in 240VAC coils/solenoids).
And I presume you mean the relay at the AC output of the DMS3... (At the input should be illegal.)
If so, no cutting should be necessary - the DMS3 AC supply passes thru, else the inverter is powering the loads and supplies the MEN.
IE - the APPROVED and Licensed-wired 240VAC relay coil is across the mains input of the DMS3 so that when energised, its contacts at the DMS3 AC output bypasses the inverter. [ POST EDIT - can now probably ignore this original addition... "as well as cutting the inverter's DC supply. (How is another issue.)" ]
When the mains AC drops, the inverter relay drops out, hence changing over to the inverter's AC & MEN point. That involves [ POST EDIT - can likewise probably ignore this original text... "DC connection and inverter start-up, and" ] a break of at least the relay changeover time (plus inverter start-up).
If mains is available, the batteries are being charged thru the PMS3 so the inverter overhead (idling or stand-by current) isn't an issue.
So forget that mains disconnecting the inverter's DC statement earlier. (Sorry, I should have thought about this...)
Whilst the mains is stable, you might manually switch off the inverter.
FYI regarding the 2nd diagram...
It is essentially an ac-ac UPS but without the rectifier aka charger.
It was from a telco spec for powering "essential" AC equipment off 48V exchange batteries that powered all the comms gear.
It was also an "on-line" system, ie, it was the inverter that normally supplied the load. (Unlike off-line UPS which switch to the inverter ONLY if there is no mains AC. In fact typical domestic PC UPS etc have a combined rectifier(charger) and inverter so it can only either charge, else invert. (LOL.... How do you know when the inverter has failed?))
Hence it would only switch-over if the DC failed or the inverter failed or was overloaded, ie, no switching delay due to loss of the mains AC.
However, the inverter would track the incoming mains and would allow a manual transfer if it was in sync.
It was the inverter that energised the "bypass" (mains-changeover) relay. (Energise when the inverter supplied the load.)
To cut a long story short, it was defined as a "no break" system where "no-break" was defined as "up to a 5ms break" (ie, for relay switching time, and- in theory - sensing time).
That was because under typical standards, a 5mS break could be tolerated by typical loads like PCs (... LOL... but not the IBM-XTs with the Mexican-made power supplies with brown rivets - like Sodium lamps etc they only tolerated a 1mS break if that! Oh - the memories... )
And if you wanted to force a break when the inverter wouldn't allow it (because the mains was not there or out of voltage/frequency tolerance or not in-sync), then you simply flipped the inverter's input DC breaker (and prepared an explanation if its loads blew up! )
Why did I recite the above?
Firstly because that's the way I'm used to thinking and you application is sort of inverted... ooops, opposite to that.
Hence I need to be careful and you need and others to check my sanity.
It also includes important issues such as what if the loads are switched over to an out-of-phase supply?
And I shall reiterate the universal Riot Act that you SHALL not perform any work that is Mandated to be done by qualified and Licensed personnel.
And that extends to 240VAC etc work that is NOT covered by such mandates.
I did try to investigate the DMS3 and your desire in more detail, but didn't get far.
However during this reply's "discussion", it seems the answer may be fairly straight forward. But it can't be that easy can it?
But hold back any implementation lest you die or get gaoled or kill someone else or cause damage etc.
Was what I outlined what you were thinking?
And are you UK based? (Or yUK LOL.)
Meanwhile I'll break and try a fresh look...
power supply? Not sure!
Ah yes, a fun question.
It can happen. F.ex, the North Island of New Zealand is (or was?) fed by a DC link from the South Island. Occasionally its phase could instantaneously change due to its DC-AC conversion. (It used to be about twice power year. I found out because AC-AC UPS would drop the mains.)
I'm not sure if Vancouver suffers the same. (It's undersea AC feeds were converted to DC decades ago to carry more power.)
And anyone with a typical PC UPS is likely to experience the same during a power outage.
Chances are that modern SMPS power supplies (used in PCs & TVs and most newer things like phone chargers etc) don't care - they just need a certain voltage of any shape.
Transformers etc won't like it, but they too are probably ok.
Plus if it's with a 4-5mS break (a typical relay or contactor switch-over time), that's about a 1/4 cycle or 90° break for 50-60Hz AC. It's not as if you have an instantaneous voltage change as I described above (where worst case maybe it could go from a peak -√2 x 240 = -340V to +340 within a milli-second. though I'd consider that extreme unlikely).
Alas I'm trying to delve deep in my memory banks... But all I see is brown rivets!
But even if it's a tracking in-sync inverter or UPS, a switching delay of ~5ms is ~ 90°, hence on average some 100V for a 240V AC supply. (Would it be the "average" of half the peak?)
And that's just the switching delay. How long does it take to sense and decide to switch? IE - the voltage difference could be greater...
[ That was never defined in aforementioned inverter nor UPS specs, mainly because being on-line, it was not that important an issue. (READ: too hard to define!) ]
Now and again when switching on power outlets you might here a crackle or an arc. From what I recall, that was sometimes due to the switch-on being near the peak voltage whilst the transformer etc load was magnetised in the opposite phase. (Other times due to bad switches or slow turn-ons, hence arcing.)
Last edited by OldSpark; 04-24-2012 at 10:19 AM.