At 02:52 PM 9/2/2006, Jim Brown wrote:
>On Sat, 02 Sep 2006 11:48:13 -0500, Mike Niswonger wrote:
> >Try switching out the wall wart power supply.
>Yes. I include these power supplies in the list of potentially
>bad boxes. Also, try using ferrite chokes to suppress radiation
>from the power supply leads. The trash on the power leads can
>originate either on the Ethernet box or the power supply (or
>Another solution is to have one suitably-sized linear supply
>trickle-charging a battery that runs the gear.
The one challenge I've found is that most wall-wart powered gear doesn't
run off 12V. 5V is pretty common, as is 7.5, and there's some 10s and 18s,
too. For the lower voltages, one could run a linear 78xx type regulator on
a big heatsink, but, as Jim says later, this is kind of wasteful and expensive.
There are a number of fairly RF quiet DC/DC converters around with decent
efficiency (85%+) available from Mouser and Digikey, to accomodate all the
various voltages you need. I just got a Cincon 9-36V input, 5V, 4A module
that appears to be very quiet for about $60. At least, I don't hear any
new birdies on the radio that's connected to the equipment that's connected
to the module. I suspect that newer modular designs tend to be pretty quiet
because a)they're not quite as price sensitive as a consumer wallwart and
b)they're going into something which presumably will need to be quiet, and
the designer doesn't want to fool with adding parts to filter a noisy
converter. A "built into the consumer gadget" converter is much more likely
to be noisy, because of the price sensitivity.. consumer stuff tends not to
use modular converters.
Maybe someone will come up with a project for QST with a little board that
has a high efficiency switcher with widerange input ( 9-36 is nice.. runs
on either 1 or 2 car batteries nicely) and a settable output (say by
choosing a couple resistors or components) and which is RF quiet (perhaps
by putting it in a good packaged with filtered connectors?). A battery
power to wall wart converter, if you will.
>adopting this strategy as an alternative means of UPS for my
>computer networking infrastructure. In addition to the obvious
>noise reduction possibilities, conventional UPS units are power
>hogs. I've become quite conscious of this after moving from
>Chicago where I paid $0.10/kWh, to California, where I'm paying a
>graduated rate that hits $0.35/kWh.
It's even worse than that at the higher tiers, I suspect (as I read my $350
electric bill here in Southern California). The recent 15% rate hike was
not across the board (as one might think), but almost entirely on the high
tiers, where the percentage increase is substantially higher. (29.75 c/kWh
for generation for kWh over 200% of baseline+ 6.73 c/kWh for delivery) But
that's ratepayer politics...
TowerTalk mailing list