charlie said:
There WAS server based software that could actually start the P/C's
shutdown sequence over an Ethernet connection. Since I retired, I
haven't bothered to keep up with such things.
If your UPS software lets you run a command before it shuts down the
local host on which that software is running, you could have it run the
shutdown.exe program.
Run "shutdown.exe /?" in a command shell to see the parameters for it,
one of which is to specify the "computername" (or hostname) of the host
to which you want to issue the shutdown command. If there is more than
1 remote host to which you want to submit the shutdown command, create a
script file, like a .bat file, that contains multiple shutdown commands,
one for each remote host, and specify that .bat file in the UPS config
for a command that it executes before it shuts down the local host.
Getting the UPS to turn
itself off was built-in some models. It was based upon a percent of
charge left in the UPS batteries. Supposedly it was to keep the
batteries from being too deeply discharged.
Guess I haven't used a UPS that doesn't shut itself down when it no
longer has a sufficient charge to carry the output load. If it cannot
supply the output load, why stay powered up?
Also, you might want to look at configuring the UPS so it does *NOT*
restore output power when its input power is restored. Often after a
power outage and when the power comes back up, it may bounce a couple
times. That is, power comes up, drops in a short time, and may repeat
this a couple times before the power source actually becomes stable.
You don't want your hosts starting to power up and boot into an OS and
load startup apps but then lose its power (because the UPS has had no
power so it isn't charged so it cannot supply any power to the loads
between these bounces). Your host would power-bounce along with the
bouncing of restored source power to the UPS. That's a consideration
only when the loads will power up when power is restored to them. With
computers and Windows, and assuming you are using the newer ATX-style
PSU (not the AT-style with a hardwired Power switch), the computer won't
power up just because power was restored to it. However, that doesn't
apply to your powered speakers, printers, monitor, or whatever else you
attach to the UPS. Not all consumer-grade UPS'es have a config setting
to not supply output power when input power is restored.
Some of the "Gaming" P/Cs have as large as 1000 or 1200 Watt power supplies.
The one I'm using has a 900W 80%Eff P/S with two AMD/ATI HD video cards,
and can draw up to about 750W peak while playing newer games.
The usual power draw in office apps is no where near the P/S capability,
Some of the really fancy graphics in newer games can use multiple video
cards, (two to four)and each video card can draw somewhere around 300W,
when at max resolution and clock rates.
How do you know what is the power draw by all devices attached to the
UPS? Does the UPS software tell you the current load on it? Many do.
The max power ratings for your computer components doesn't tell you what
they are actually drawing.
I quit buying replacement UPS batteries. Seems that many of the newer
UPS models use batteries that have sensors and/or taps built-in.
(Similar to laptop battery circuitry inside the battery case)
Naturally, it costs almost as much to buy new batteries as it does to
replace the whole UPS.
Not what I've found at all. For an $80 UPS, the battery replacement at
BatteriesPlus.com (who is NOT a cheap battery source) is $32. That's a
hell of a lot cheaper then buying a whole new UPS. I had a 2.5kVA UPS
(very pricey when first purchase, like ~$1800) with a huge and heavy
isolation transformer and generating true sinusoidal output (so it could
be used with non-computer gear that doesn't have switched power
supplies) and the 2 motorcycle-sized batteries cost a total of $65.
That is a lot cheaper than buying another 2.5kVA sinusoidal isolated UPS
even at today's cheaper prices where they run $500 an up.
LIFe UPS batteries for high reliability use can cost six or seven times
the cost of the usual lead acid UPS battery. (~$700, and that's a recent
price quote)
I always use the same-type battery when replacing it. Why? Because the
charger circuit inside the UPS was designed for THAT type of battery,
not some other chemical composition and different charging behavior. If
the UPS had a sealed lead-acid battery then that's the type I get for a
replacement. I wouldn't be putting a lithium battery in as a
replacement for a lead-acid one since the charging curve, rate, top-off,
and trickle charge for them are different. You'd burn up a lithium
battery using a charger meant to be used with lead-acid batteries plus
the UPS designed for lead-acid batteries would let them drop down to
zero volts but lithiums should never be dropped that low. Trickle
charging of lead-acid batteries should not be used with lithiums. Even
the phases in the charging cycle is different. Li-FE batteries are very
sensitive to over-discharge which will happen if the charger in the UPS
was designed for a lead-acid battery. Lithium iron sulfate, Lithium-
iron, or Li/Fe batteries are a replacement for alkaline zinc-manganese
batteries, not lead-acid batteries.
You do pay through the nose to have a Li/Fe (more likely Li/Fe/PO4)
battery that has 2000 charging cycles before it reduces to 80% of its
capacity along with a 10-year storage life. However, if your source
power is dropping 200 times a year and for long enough to fully
discharge the UPS battery, you need to consider using an alternate power
source, like your own generator or contact your power company on why you
are losing power every other day.