[ale] APC UPS

hbbs at attbi.com hbbs at attbi.com
Thu May 8 12:59:56 EDT 2003


Did anyone find out if that generator was DC?  Reason I ask is that once
undriven, a flywheel naturally can't stay at the same speed and therefore can't
run an AC generator so as to make it continue to put out 60Hz.

Once place where I used to manage a data center had a very old, very dead, very
big UPS that hadn't been used in many years; the machines were running off of
several small UPSses.  There was no generator and there had been buzz and hype
at the home office about getting one, but when I was charged (pun) with
replacing that big UPS with a new one, I worked out a plan to hook it up to a
Linux box with a 16-port serial card on it, and I was going to program the Linux
box to progressively shut down and power off stuff in the room so as to save
power for the production AlphaServer cluster, the Frame Relay router, and the
Ethernet concentration gear until the very end, and best I could tell, that
would take about two or three hours - longer than any of the power outages the
building had experienced in the previous three years.  

I was laid off a week and a half after I put the new UPS in, and a few months
later, they bought a natural-gas-powered generator at a cost of about $50,000,
having opted not to spend the $5000 or so that my solution would have cost. 
They should have stenciled my name on it as a memorial.  

- Jeff
> On Wed, May 07, 2003 at 03:23:42PM -0400, Robert E. Karaffa, II wrote:
> > I've always been led to believe that one wants a UPS that can talk to the
> > computer so that when power fails and the humans are away, the computer can
> > take care of itself.
> 
> My comment was to suggest that, while desirable for automatic shutdown
> when the UPS battery is running down, it is not critical.  The chance
> of file system damage from an abrupt shutdown of a desktop or other
> Linux system that is not very active when the humans are away seems
> to be in the 1%-5% range.
> 
> Thus, it is far better to have a UPS connected for the 95% of power
> interruptions that are sufficiently brief that the UPS can "ride them
> through" and risk the occasional abrupt loss of power than to not
> use any UPS.
> 
> To respond the the question earlier on power conditioners, in most
> areas the voltage is sufficiently stable that conditioners do not
> seem necessary.  UPSes are so cheap these days that there's no reason
> for conditioners in any case.
> 
> Btw, I was totaly amazed by E-Delta's flywheel motor-generator setup.
> Wow!
> 
> > -Bob K.
> 
> -- Bob T.
> 
> 
> > on 5/7/03 2:55 PM, tfreeman at intel.digichem.net at
> > tfreeman at intel.digichem.net wrote:
> 
> > > On Wed, 7 May 2003, Transam wrote:
> 
> > >> On Tue, May 06, 2003 at 09:55:13AM -0400, Hogg, Russell E wrote:
> 
> > >>> I'm considering a UPS for home use.  I'm doing some reading now about what
> > >>> numbers matter and what is worth paying for when it comes to a UPS.
> 
> > >>> I'm curious what sentiment the group at large has.  Surely a lot of you 
> know
> > >>> more about UPS(es) than I.
> 
> > > <<snip usual excellent response>>
> 
> > >> Regarding surge suppressors (that one should have on peripherals), the
> 
> > > Question time - There used to be power conditioners which were
> > > intermediate between a surge suppressor and a UPS in price. You didn't see
> > > them often, but I haven't seen any in ages. They could tame _big_ power
> > > surges, big short voltage drops, longer slight voltage drops, and in
> > > general clean up the line power to something pretty nice.
> 
> > > Question is are these things still around for sale and worth the effort?
> 
> > > Thanks in advance.
> 
> > >> cheap ones are essentially worthless.  They have merely a diode to
> > >> absorbe
> 
> > > <<snip again>>
> > >> number of spikes, who needs it.  Most outages are for only a few seconds
> > >> so there probably is not a need for most even to bother with the "wire
> > >> the UPS to the computers to shut down the computers stuff".  It IS nice
> > >> to have for commercial use.
> 
> > > Given that I've had two multiday outages and a single day outage in the
> > > last six months, I probably need to reimplement that "shutdown wire"
> > > stuff.
> 
> > -- 
> > **************************
> > Robert E. Karaffa, II
> > Technical Director
> > Emory University
> > Flow Cytometry Core Facility
> > 954 Gatewood Dr.
> > Atlanta, Ga 30329
> > voice: 404/712-4429
> > e-mail: rkaraff at emory.edu
> > web:  http://www.emory.edu/WHSC/MED/RESEARCH/FLOWCYT/
> > **************************
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://www.ale.org/mailman/listinfo/ale
_______________________________________________
Ale mailing list
Ale at ale.org
http://www.ale.org/mailman/listinfo/ale





More information about the Ale mailing list