I am trying to find out some information about how 230 volt UPSes are
wired up. In particular, I want to know how the output is wired with
respect to the hot/neutral/ground wiring relationship. I am intending
to get a "dual conversion continuous online" type of UPS, however, that
may not actually happen, so I need to know this with respect to both
types. I suspect the possibility these types may be different because
the dual conversion type has more opportunity to isolate things.
What I am looking for is knowledgeable information, not conjecture.
I am do plenty of conjecturing, so that is not in short supply.
I highly suspect, for safety and compliance reasons, the grounding wire
is passed through solidly.
It's the 2 power wires that I want to know about. There are three
different kinds of systems in the world:
1. A specific wire is always the grounded neutral wire, with outlets
and plugs polarized to ensure the correct connection. This can be
seen in countries like UK, Australia.
2. One of the wires is the grounded neutral wire, but outlets and
plugs are not polarized, so plugging in can result in either wire
being the grounded one. This can be seen in countries like
3. Neither of the wires is grounded. Each wire is at half voltage
relative to ground, in opposing polarity so the different between
them is the full voltage. Outlets and plugs may or may not
be polarized, but this would not matter. This can be seen in
countries like USA for 2 wire 240 volt circuits (e.g. what you
might plug a heavy duty air conditioner into).
I am wanting to operate the 230 volt UPS in the USA. What I want to
be sure of is that there is no possibility of causing a short circuit
through a backfeed, either in line interactive mode UPSes, or when
a dual conversion UPS is in bypass mode.
I believe that would only be possible if one of the output wires is
grounded while both are being passed through electrically. Since that
would pose a hazard in the German wiring, I suspect it would not be the
case. But I cannot rule out there may be another wiring strategy that
would be safe in both UK and Germany, but not safe with the USA wiring.
A dual conversion UPS could be considered a "separate derived system"
(in terms of electrical code), and probably wired as such. But how
would it be grounded? I suspect one output wire is grounded and thus
the other will be 230 volts relative to ground. If the switch that
transfers to bypass mode is the "open transition" type, that should
still be safe even with a USA power source, ehough it will result in
a change in the voltages relative to ground when bypass in engaged.
There are UPSes designed for higher voltage specifically in the USA,
but these tend to be much larger capacity (over 3000 VA), specified
for 208 volts (instead of 240 volts), and usually are three phase type.
What I need is 1500 to 2500 VA single phase.
I know most computer power supplies can be safely operated on 240 volts
even with the USA style wiring. The power cutoff switch breaks both
wires, and no assumption about which wire is made in any aspect of the
power supply wiring. This is required in Germany due to the Schuko
plug being able to rotate 180 degrees. I cannot recall ever seeing a
computer power supply that cannot be operated on "220" (as labeled)
either by means of a voltage change switch or autoranging.
Still, the big question is how the output from the UPS is wired to,
or otherwise made relative to, ground. These units don't generally
use isolation transformers, so whatever is done, it is done as part
of, or in conjunction with, the output DC/AC conversion circuit.
Is there anyone using the "international" 230 volt UPSes on American
style 240 volt (center tapped neutral) circuits?
Do you see or have any reason to anticipate any safety issues?
Actually no. Even dual conversion ups pass the neutral straight through so
they are not separately derived sources.
We test UPS for both US and European usage. If you connect a 240V UPS to a
US style 240V supply, you can get large voltage swings from output line to
ground when the UPS goes to battery. IMO, from our test results, the only
safe way to use a 240 EU UPS on a US 240V supply is with an isolation
transformer. The neutral connection of the UPS really needs to be the
neutral and not a phase conductor.
Remember, that even when the UPS is OFF, the neutral conductor WILL BE
ENERGIZED if you connect it to a US 240V service.
Charles Perry P.E.
|>I am trying to find out some information about how 230 volt UPSes are
|> wired up. In particular, I want to know how the output is wired with
|> respect to the hot/neutral/ground wiring relationship. I am intending
|> to get a "dual conversion continuous online" type of UPS, however, that
|> may not actually happen, so I need to know this with respect to both
|> types. I suspect the possibility these types may be different because
|> the dual conversion type has more opportunity to isolate things.
|> A dual conversion UPS could be considered a "separate derived system"
|> (in terms of electrical code), and probably wired as such.
| Actually no. Even dual conversion ups pass the neutral straight through so
| they are not separately derived sources.
The question is, how can they know which wire is the neutral? That is
not identifiable when the incoming power uses a non-polarized plug such
as the Schuko. Remember, the plug can be rotated 180 degrees. These
would be UPSes generally designed for markets including Germany where
L-N is 230 volts, but you don't know which line is neutral.
Of course in my case I'd be using a NEMA-6-15 to C13 or to a C19. The
output power would be to C13 outlets, and C14-C13 cords would be used
to connect computers to it.
| We test UPS for both US and European usage. If you connect a 240V UPS to a
| US style 240V supply, you can get large voltage swings from output line to
| ground when the UPS goes to battery. IMO, from our test results, the only
| safe way to use a 240 EU UPS on a US 240V supply is with an isolation
| transformer. The neutral connection of the UPS really needs to be the
| neutral and not a phase conductor.
When you say "goes to battery" I assume you mean a "line interactive UPS"
as opposed to a "continuous dual conversion". I do expect a L-L voltage
coming out in the 220-240 volt range. And I expect the hot wire coming
out will be 220-240 volts relative ground when in conversion operation.
So I realize I cannot depend on a lower L-G voltage as I might in a USA
style wiring. OTOH, it's wrong to depend on the L-G voltage; only the
L-N (or L-L if applicable) voltage should be considered.
I presume these "international" power supplies are safe with either a
"0-0-230" volt structure, or a "230-0-0" voltage structure, or a "115-0-115"
voltage structure. Any reason to believe otherwise?
When you say that an isolation transformer is needed, how are you expecting
that transformer secondary to be wired? Would it be 100% floating? Or
would one of the 2 wires (any center tapped not used to emulate EU power)
be bonded to ground?
| Remember, that even when the UPS is OFF, the neutral conductor WILL BE
| ENERGIZED if you connect it to a US 240V service.
Which conductor is the neutral conductor if you rotate the plug 180 degrees?
You believe the off switch of such a UPS is single pole, not double pole?
No. A true double conversion UPS does not run on the battery at all times.
The exact voltage at which it transfers to battery power depends on the
manufacturer, model, and sometimes even settings.
Another thing, don't think that a double conversion UPS will protect a load
from all outages, voltage sags, or voltage swells. We find MANY units that
will drop the load for certain input conditions that are not out of the
A lot of what you ask depends on the size of UPS. We have done extensive
testing on UPS that are 2kVA and larger. Most of these do not plug into a
two pin Schuko, and they definitely have a terminal that is to be connected
to neutral. I will have to ask one of my test engineers about the smaller
units. I don't remember evaluating any.
Charles Perry P.E.
|> When you say "goes to battery" I assume you mean a "line interactive UPS"
|> as opposed to a "continuous dual conversion".
| No. A true double conversion UPS does not run on the battery at all times.
| The exact voltage at which it transfers to battery power depends on the
| manufacturer, model, and sometimes even settings.
I know that. That's why I had to ask you which you mean. But I do know
a double conversion is running on DC, paralleled with the battery by some
means. Hence it is deriving AC power from the DC. It is not a straight
| Another thing, don't think that a double conversion UPS will protect a load
| from all outages, voltage sags, or voltage swells. We find MANY units that
| will drop the load for certain input conditions that are not out of the
Theoretically faulty units, then. Maybe lots of the products do that.
But then I'd say lots are faulty.
OTOH, I have see double conversion UPSes that cut the input AC to DC
conversion, effectively going to battery, at the slightest twitch in the
AC input. But the load remained solidly powered.
| A lot of what you ask depends on the size of UPS. We have done extensive
| testing on UPS that are 2kVA and larger. Most of these do not plug into a
| two pin Schuko, and they definitely have a terminal that is to be connected
| to neutral. I will have to ask one of my test engineers about the smaller
| units. I don't remember evaluating any.
The Schuko could support up to 16 amps, though typical circuits using it
in many countries are limited to 10 amps. So I can easily imagine that
a UPS needing more than 8 amps could abandon the Schuko because in some
places a dedicated circuit might be needed and those would be typically
If a UPS does pass neutral through, then it cannot bond it to the ground
at the output, and hence is not a separately derived system. However, I
see no reason to actually pass the neutral through on a dual conversion
unit. Pass the ground through and ground one of the converter outputs so
you have made a grounded conductor (technically not a neutral if the whole
system is just 2-wire) and properly referenced the system to ground.
I just spoke to a tech support person at Powerware, and asked about the
wiring issues. He did say that if the neutral was reversed on 120 volt
models, they would refuse to start and give a wiring error status. But
when I asked about the 230 volt models, and mentioned the possibility of
systems with no neutral being supplied, he offered the information that
their "high voltage" models "work fine on the dual hot wiring of 240 volts
in the USA".
I'll call some other manufacturers later in the week.
But they do pass the neutral straight through. Strange but true.
And I bet I can make those same units fail to supply the load. All have
control algorithms programmed by faulty humans ;-)
Sure they say that. Only testing provides the real answers. You would be
surprised at how little REAL knowledge manufacturers have of their products
under less than ideal power conditions. We often end up teaching them about
shortcomings in their products. The process runs like this:
Them: There is no way our product does that!
Us: Here is the test data and the test protocol
Them: Oh, we never tested that, it can't happen anyway
Us: Here are the results of approximately 100,000 monitor-hours of power
quality monitoring in the US and across the world.
Them: Oh, I guess we need to look into it.
Always a fun time!
Charles Perry P.E.