208V*sqrt(3)= 360V

If you go phase-to-phase with a 250V rated supply it will go BANG, big time.

There is no reason why a supply rated for 360V input would be any more efficient than one rated for 240V.
Do you think that USA equipment running at 110V is much less efficient than UK equipment running at 230V? It is not. The designer decides what efficiency is needed/wanted and designs accordingly.

You might wan to very that the switchmode supplies have a range switch, or are truly universal.  As agreed on by utilities and appliance producers, the generally accepted nominal 60Hz voltage limits for ‘208V’ equipment are 184-220V.

Another advantage of 208V switchmode operation is elimination of neutral triplen currents.  In most cases, “stretching the rules” should be avoided, like operation from 277V branch circuits. Also, if the popular IEC320 C13/14 connectors are used on the cord end opposite the wall plug, take caution in plugging the cord end into a component not intended for {or switched to/jumpered for} a higher voltage—like a monitor or printer.
  

logbook:

208V 3-phase system means it has 208V L-L.
L-N is 208/sqrt 3 = 120V. A standard system in the USA.

The covention is to specify a 3 phase system with its L-L voltage.

So two Lines of a 208V 3 -phase system will provide 208V system not 360.

rbulsara is right of course, but what that means to you hrc, is that yes you can do that IF the power supply can truly handle it. And, as Busbar noted, there may be a range selector switch, jumper or dip switch to do so. Most SMPS's have one somewhere because 90-250V is too wide a range for the regulator circuit to cover on it's own without greatly affecting efficiency.

"Venditori de oleum-vipera non vigere excordis populi"

Something to check is the front end diode bridge rating and heatsinking arrangement. The conduction angles for a three phase rectifier are shorter than those for a single phase design. The heating effect per diode is therefore lower.

The potential problem is probably more likely on bigger supplies but is still worth checking.

-----------------------------------

Start each new day with a smile.

Get it over with.

Sounds like it has a boost pfc. This would allow the wide
range input without a selector switch. If you run it at 208
your currents will be less. The efficiency should be a little higher too, do to the reduction of the I squared R losses. Your best bet is to find out who made it and contact them on how to wire it or what else is needed to run off 208. Most have internet site too.

I've been in contact with the power supply engineering group and discussing both the pf and efficiency of the units.

From what I have gathered, switchers typically will rectify the input AC thru some sort of bridge then run the DC thru the transformers and switcher controllers.  Because the 208VAC is from two legs of the three phase supply, there really isn't two lines out of phase, since there is no reference to ground, so essentially the bridge "sees" the rms value of the 208 or around 293VAC.  Its now a question of how does the supply react to a higher voltage on the DC side of things (efficiency and power factors).  More work to do in the lab to check these out vs. what the theory says.

I appreciate all the input!

rbulsara and jraef

thanks for correcting me.

I must admit that when I talk about 3-phase systems I think in terms of the line-to-neutral voltage. The incoming stuff to our factory is standard 230V, 3 phase. We are mostly running single-phase equipment distributed evenly across the phases, so everything is 230V rated. It’s useful to note the ambiguity in the usage which caught me out

Phase to phase will at least work for the next guy!  While common, most utilities don't favor 208.  If the machine goes somewhere else, there may not be a neutral in the three phase.