<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.5730.11" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>>The per-port regulators would be non-isolated.
Probably feeding off<BR>>an internal 48V bus.
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Yes, so that will be 90 to 95% efficient, but it is
fed from an isolated supply that at best will be 90%, probably less. Those
numbers must be multiplied, giving 81 to 86% overall efficiency--and I am
assuming best in class, which only a switch at a high price point would
offer.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>>Well, the bulb has a peak
permissible operation temperature of about<BR>>160 degrees C ... so likely no
extra cooling required
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>True in open air. However, inside a metal
enclosure as I described to make it applicable to the situation, the enclosure
would probably get warm enough to be a fire hazard. The light bulb might
still be fine after the building burned down! :-)</FONT></DIV>
<DIV><FONT face=Arial size=2>I've designed products and put them through
UL, CSA and CE safety testing.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV>>It'd still be a 400W PSU if it supplies 400W</DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>You can play various games with the numbers, such
as choosing input versus output power to write in the spec. The label will
have to indicate 500 W of power consumed, by law. No matter how one plays
with the numbers, an 80% efficient supply that delivers 360 watts will consume
450 W, and turn 90 W into heat. If it is rated for 400 W output, it is
probably in a portion of its range at 360 W where it is achieving near optimal
efficiency and the above math would apply.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>In cases where half the loads drew less power
and/or were not POE loads, the consumption would go down, and so would the
heat. However the manufacturer can't design for that case and if they
don't provide automatic fan control (which apparently most switches don't have),
the fan must be designed for worst case, which in the above example is 100 W of
heat. Also, at 25 to 50 % load, the efficiency will probably be lower
because it was optimized for a larger load, and because some losses are fixed
(not load dependent).</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV>>This whole thread is getting stupid</DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>If some increase in understanding occurs, then it
isn't a waste. No it won't change the behavior of any switches, but it
might help people understand why.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>Wilton</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV></BODY></HTML>