Re: UPS and generator interaction?
Oh, another detail. Some 98% of the UPSi around are "standby" units. They sit and trickle their batteries until the line fails, then quickly kick in. They take 'n' hours to recharge when the line returns. But there exist another genus. These 'full time' units ALWAYS run the load from the UPS inverter; and have a big AC line->DC "battery charger" -- big enough so as to keep up. The advantage is a very high degree of line isolation. Any surge, sag, glitch, spike may affect the AC->DC side of the equation, but will have to get past the battery plant and inverter for the load to see it. Note it does not even care what the input frequency is. I know of one large unit that was sent to Mexico City. At the time, it was 50Hz, but there was some announced plan for the city to go to 60, Real Soon Now. The UPS "battery charger" ran on anything between say 40-70 Hz, but the inverter made 60.0 Hz, period. Such units are not common or cheap. In the low end, Sola Corp used to make some in the low (1-3?) KVA range. Top end, how many KVA do you need? I think the one going to Mexico City was 500KVA. If your power is really rotten [Here, Guyana comes to mind...] you may want to spend more up front. Side thought, but not a NANOG topic. What in your data center really cares if your generator puts out 57 or 63 Hz, not 60.0? Why? -- 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
David Lesher wrote:
Side thought, but not a NANOG topic. What in your data center really cares if your generator puts out 57 or 63 Hz, not 60.0? Why?
Some clocks get a little nutso. Because they are powered by AC synchronous motors with gearing that assumes 60 Hz. (or 50 Hz, as the case might be.) Some fans and other devices also use synchronour or induction motors with similar engineering assumptions. -- Requiescas in pace o email
Some UPS's do also. We had a site that needed to have some work done on the building mains. Small, in the country, yada, yada, yada. Owner had a tractor mounted generator with a gearbox in front of it. Pretty stable once you tweaked it in because of the HP of tractor (small load too, like 4kVA, gen was like 20kVA). Problem was the gearbox in front of the generator between it and the tractor was not standard, at least in the PTO realm. Tractor owner thought it was correct for a 540 rpm PTO speed. It wasn't. Forgot to check the output with a Voltage and Frequency meter. Smelled a distinct odor of burning electrical components. Line Interactive APC UPS's don't do well with about 160 volts and 83 Hz input. Apparently, you can only interact so much .... Honest to God, APC replaced it because they said it should have tripped off line with the line being that out of whack ... Eric
-----Original Message----- From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu]On Behalf Of David Lesher Sent: Tuesday, March 30, 2004 10:32 AM To: nanog list Subject: Re: UPS and generator interaction?
Side thought, but not a NANOG topic. What in your data center really cares if your generator puts out 57 or 63 Hz, not 60.0? Why?
-- 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
Line Interactive APC UPS's don't do well with about 160 volts and 83 Hz input. Apparently, you can only interact so much ....
Interacting with the line requires battery power. They can't charge and supply power at the same time, unlike traditional online UPSes. Once their batteries (LI type) go, they can't do anything.
Honest to God, APC replaced it because they said it should have tripped off line with the line being that out of whack ...
APC is very good about replacing equipment that might be their fault. Even after warranty. I've never seen another company with as smart, techncial staff available quickly. Too bad no one has any real experience with their big gear to see if its supported as well. Deepak Jain AiNET
Once upon a time, Deepak Jain <deepak@ai.net> said:
APC is very good about replacing equipment that might be their fault. Even after warranty. I've never seen another company with as smart, techncial staff available quickly. Too bad no one has any real experience with their big gear to see if its supported as well.
We've got a APC power strip (20A vertical "zero U" mount with 14 outlets) that failed (possibly from heat). The circuit breaker tripped and would not reset. Their response was that we had too much plugged in (never mind that we switched to a different 20A strip and it is working fine on a 20A breaker) and that while they would replace it, if the replacement failed they would NOT replace it. I prefer APC for small UPSes, but I'm not impressed by support on a simple power strip. -- Chris Adams <cmadams@hiwaay.net> Systems and Network Administrator - HiWAAY Internet Services I don't speak for anybody but myself - that's enough trouble.
We've got a APC power strip (20A vertical "zero U" mount with 14 outlets) that failed (possibly from heat). The circuit breaker tripped and would not reset. Their response was that we had too much plugged in (never mind that we switched to a different 20A strip and it is working fine on a 20A breaker) and that while they would replace it, if the replacement failed they would NOT replace it.
I prefer APC for small UPSes, but I'm not impressed by support on a simple power strip.
That sounds like fairly decent support for a power strip. I would have let them replace it and see if the same thing happened. If you were overloading it, the breaker should be resetable. However, a cheaper power strip may not have as touchy a breaker or allow more power than it should. They have replaced full MATRIX 5000 units, batteries and all for me a few times without quibbling and without any kind of extended service contract on them. I'm sure someone will have already suggested that you put an ammeter on the load to see what is actually happening. While I have never used APC's Zero U strip, I have heard good things about Baytech's version (which I've also not used). YMMV. Deepak
On Tue, 30 Mar 2004, David Lesher wrote:
Side thought, but not a NANOG topic. What in your data center really cares if your generator puts out 57 or 63 Hz, not 60.0? Why?
Some UPSes such as the Best FerrUPS series and other voltage regulators and line conditioners that use a ferro-resonant transformer where there's an L-C tuned circuit as part of the power transformer. Other motor loads may care to some extent. Analog electric clocks will run slow or fast, no big deal. Lower freqencies are harder on marginally designed transformers which may not have enough core material. -- Jay Hennigan - CCIE #7880 - Network Administration - jay@west.net WestNet: Connecting you to the planet. 805 884-6323 WB6RDV NetLojix Communications, Inc. - http://www.netlojix.com/
participants (6)
-
Chris Adams -
David Lesher -
Deepak Jain -
Eric Germann -
Jay Hennigan -
Laurence F. Sheldon, Jr.