You also have to make sure your PC is set to *neither* "turn on" nor "resume previous state" when power is restored. Nothing gets me more anxious than watching the lights flicker, PC going off, then on, then off, then on... just let the damn thing go *off* and *I* can figure out when it's best to turn it back on!
:-/
I suppose it's a matter of taste, and infrastructure reliability. Hard to know what's going on in Seattle. It sounds as if restoration is somewhat by experiment. Where I live, a prolonged outage, if it's not load shedding (which has so far thankfully been rare), it's an equipment failure, and the technicians are able to fix it definitively before restoring the power. The guys on the ground seem to know what they're doing, for which one has to be grateful.
I don't like expecting to do something on my PC and finding that it's been powered down (PCs in this houseold are left running all the time), so they're configure in memory mode - they boot up after an outage if they were running before. We just power the monitors down when they're not in use.
Sylvia.
I have lived in places where most of the electrical infrastructure was "hanging from poles" and, in those places, it seemed like outages caused by "downed lines" (think: ice) often were a noticeable blink (off) of the lights followed very quickly by a real outage. Perhaps one is caused by an "unexpected event" and the other is the system's *designed* reaction to that event (e.g., a breaker deliberately tripping).
When I lived in those places, I had machines that had mechanical power switches -- "on" was ON and "off" was OFF! The time between each of these transitions (blink off, blink back on, then final outage) was fast enough that you couldn't reach the power switch to *prevent* the PC from coming back on -- and then off -- again.
We try to be concious of our energy use (abuse?). Since the machines use a fair amount of power (triple redundant power supplies and fans, etc.) we try to turn things off when they aren't actively "doing something". On the other hand, there are often 5 or 6 machines running concurrently out of necessity. (no need for space heaters in those conditions -- and something to definitely avoid in the Summer months! :< )
Not very environmentally friendly are we? Most modern PeeCees consume a hundred watts or more when idle. I hit the sleep button on mine and in 10 seconds it goes down. Hit the button again and it takes 10 seconds to wake and is ready to use. Since we have two in the house it saves energy and we're more in tune with the ecology. It also may help explain why our batteries last a long time.
Yes we are. We're happy to have a nuclear power station built.
Sylvia
Some high-end UPS units, like the APC SmartUPS rack mount ones, will let you set a turn-on delay for this purpose.
I've had mixed luck with that, too. What I do now is log the battery voltage and percent charge, and judge by how far they drop during the self-tests whether the pack is on its way out.
My empirical experience is that there are three issues here:
I've worked at two sites that had APC SmartUPS 3000 XL units. These are
3 kVA rack-mount UPSs that take eight 12V gel cells in series-parallel to make a 48V pack.At the first site, the server room was pretty hot and the power was really unreliable. We'd fully discharge the pack every couple of months on average. Typical battery pack temperatures were around 100 F. We were putting in new batteries every two to three years.
At the second site, we practically never had outages and the server room was relatively cool. We got five years out of those batteries before the voltage started to drop precipitously during the weekly self tests.
BTW, some of the early SmartUPS 3000 units don't use a battery tray and have all the batteries slide into the front of the UPS single-file. It's a hell of a job to change them out when a battery fails and swells up, wedging itself in place! Later ones have a removable tray/cartridge, much easier.
How hot is hot? I commented elsewhere in this thread that the battery compartment of my UPS is about 10 degress Celsius above ambient. I'm wondering whether I should extend the leads, and put them outside the case.
Sylvia.
Generally cooler is better, but that doesn't sound at all excessive unless the room the UPS is in is also very warm.
Generally[0], UPSes will produce the rated output for 20 minutes, so if yours worked for an hour, that suggests that your load is only 333VA.
[0] At least, that's the figure I've extracted from various UPS spec-sheets over the years.
--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------
Gel cells "prefer" a lower float (and charge) voltage at higher temperatures. If your UPS doesn't have a thermistor circuit to reduce the float voltage at the higher battery temperature, keeping the battery cooler might extend its life by avoiding overcharging.
--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
I've been involved with enterprise power solutions for the past decade. That includes the APC Matrix line. The 3750 watt 5000va model had a full load runtime of 25 minutes.
Hi!
I suppose that's true of higher end UPS units. However, consumer grade units run at their full load may only last 3-7 minutes by the manufacturer's own admission. (I have a feeling that their admission is also somewhat inflated from the reality.)
That's why I suggest anyone buying an average UPS get one that is larger than they need. The inverter will also be a lot happier since it isn't being run so close to its limits.
William
If this hasn't been discussed already. I would say the batteries have top priority. Regardless of what charging rate is used, brand, etc.. In the last 15 years or so...I've had numerous brands such as; APC, Cyberpower, Trip-lite and some other "X" brands as well, all of them with many years of good service. For me, the average lifetime of a SLA battery has been about 3 years. Several have passed the 4+ year mark, and a few for several months. I consider this to be GOOD. I have had "fairly" consistent input power, nothing unusual, mostly several seconds, and maybe once in awhile 10-15 minutes of outage. Once in a blue moon it goes out for several hours; therefore I shut everything down, once that annoying beep gets past 15 minutes or so. IMHO. I would say the QUALITY of the battery is the most important factor. A well built battery, will be able to handle those odd charging rates a lot better. The best batteries (again, IMO) have been Panasonic & Yuasa. I have seen others fair well too, but I cant recall their brands. The best would be made in Japan, Taiwan and USA. (in no particular order) Some Chinese have fared well, but they are not on the top of my list. I can only imagine the quality will increase since most manufactures will (if they haven't already) eventually move their assembly to mainland China. Bottom line: The quality of the battery is almost everything in a UPS. Just my 25 cents worth (I added 20 for inflation:) Cheers Dave
Hmmm... I'm not sure I would agree "unconditionally"; if the UPS's are designed in such a way that they prematurely age batteries, then I'd rather figure out how to *fix* that -- no sense "wasting" what are otherwise good batteries!
I use Panasonics exclusively. I haven't found any of the "off brands" to be significantly cheaper or "better". I've had associates use all sorts of different brands in an attempt to save a few pennies here or there but usually with no better results than me.
Not the best choice.
I have a SmartUSP 1200 {that's 1200VA in their hype..}. I have used external batteries for ~10 years. I had two 50AH deep cycle batteries until they died. Now I have some ~18 AH ones. The 50's would take days to recharge, but so what?
Car batteries are a poor choice; get deep-cycle batteries.
-- A host is a host from coast to coast.................wb8foz@nrk.com & no one will talk to a host that's close........[v].(301) 56-LINUX Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.