[OT] Generator testing

Ben Scott dragonhawk at gmail.com
Wed Sep 9 11:40:46 EDT 2009


On Wed, Sep 9, 2009 at 11:17 AM, Hewitt_Tech <hewitt_tech at comcast.net> wrote:
> When I was in the Army Signal Core I worked
> at communications facilities that used battery strings (48V) that
> provided station power by feeding very large inverters.

  That's similar how US commercial telco COs (central offices) do
things.  Except they just design most of their equipment (including
landline phones) to run off 48 VDC and so don't need as much inverter
capacity.  Under normal operations, they just keep the batteries
topped off from city power.  If city power fails, the batteries just
start discharging -- there are no transfer issues because there is no
transfer.  Then they have on-site generators to start recharging if
city power doesn't come back quickly.  They also use two battery
systems in each building, with two sets of wiring, each on opposite
sides of the walls and racks, so if something goes wrong and blows up
one battery system, things keep running.  It's a very robust design,
and a rather different mindset from what one often sees today.

  In my request for stories, I was especially interested in AC
generators and transfer systems, because that's something I don't have
much knowledge of.  UPSes you can buy at Wal-Mart these days.  ;-)
There's been a little (mostly idle) talk of a generator for our front
office at $DAYJOB, though, and I know next to nothing about the
practical aspects.

  I know many UPSes don't like the output from some generators, for
example -- they stay on battery (which runs down and then drops the
load).  So I know just going to Home Depot and buying something off
the shelf is probabbly a bad idea.  But I wouldn't know what to look
for if I was asked to review a proposal.

  Pointers to good websites, books, etc., are also welcomed.  :)

-- Ben


More information about the gnhlug-discuss mailing list