On Oct 9, 7:08 am, "RBM" rbm2(remove wrote:
My assumption is that one uses a UPS to avoid power interruptions, giving
you a short time to safely save data and shut down. How is a system that you
have to manually throw a switch, assuming you are even there at the time,
going to help?

Greetings,

The addition of the DPDT switch makes the installation more
professional and more robust with increased applications and value
after the conclusion of the project with minimal extra work. Do you
think it is fair to call the flanged input and DPTP switch a "single
circuit generator input with manual transfer" when showing the house?
Do you think it will increase the home value more than the cost of the
switch?

I don't plan on ever flipping the switch in my installation. When the
power goes out I get notified on my phone. If it is out for very long
I have to plug the UPS into a generator or else the systems go into
hibernation. When the power comes back on the systems come out of
hibernation and continue as they were.

Hope this helps,
William

On Oct 9, 9:45 am, DerbyDad03 wrote:
On 9 Oct, 02:13, " wrote:



On Oct 8, 9:24 pm, DerbyDad03 wrote:

On Oct 8, 9:07 pm, " wrote:

On Oct 8, 7:27 pm, "John Grabowski" wrote:

" wrote in message

ups.com...

Greetings,

I have two computers, one in the basement and one on the third floor,
which I would like to run off the same plug-in UPS (expensive, so I
don't want to purchase another one). There is already conduit going
up to the third floor. What do I need to properly (to-code) install a
UPS protected outlet on the third floor which shares the plug-in UPS
in the basement? Is there a code-compliant solution? Are you allowed
to plug a cord into a UPS which goes into a junction box, is spliced
to THHN, and then into the house wiring?

Thank you for your time,
William

PS: I am not asking for instructions on how to install an outlet. The
outlet that the UPS plugs into in the basement is already GFI
protected. There are enough spare cubic inches in the box to support
an additional device and wiring.

Perhaps the installation of a flanged inlet at the UPS location would give
you the ability to use permanent wiring in the wall. I'm not sure what wire
size you need, but they come in 15 and 20 amp and perhaps 30 amp straight
blade and twistlock.. Depending on the inlet configuration that you use,
you might need to make your own cord set-up to connect the flanged inlet to
your UPS. At the third floor location you could just have a regular
receptacle or something different like a twistlock to indicate that it's
special.

Thank you very much David, John, and RMB for attempting to help me
create a NEC complaint installation. I am going to proceed with
project "central UPS" as one poster put it. I am going to use a red
plate and a label from a labeling machine which states the receptacle
is connected to the basement UPS. I will use twist lock flanged
connectors (in the basement) and 12 AWG wire throughout.

One line explanations for hecklers:
=====================================
Anti-rackmount coalition: I have clients because I give them what they
want.
People who don't believe in discount rates: The miracle of compound
interest is real and I don't actually mind your disbelief because
someone has to pay all that interest.
UPS TCO math: I elaborated in my Pete C. reply post.
Derby Dad: I'll be sure and eat the high-fiber box to after consuming
the cereal contained therein.- Hide quoted text -

- Show quoted text -

I don't recall saying anything about your fiber intake. I did,
however, ask a question about the code compliancy of your beautifully
finished attic.

Fair enough.

To the best of my knowledge it is electrical code complaint so long as
it is not considered a bedroom (it doesn't have arc-fault
protection.) There may also be more than 5% voltage drop if you fully
loaded the existing circuit (but that is a FPN).- Hide quoted text -

- Show quoted text -

Once again, I am feeling misunderstood.

My original question was not limited to NEC compliancy. I also asked
about egress and any other codes that might be applicable to the
intended use of the space: height of sloped walls, joist sizing for
live loads plus furniture, ventilation, etc.

I don't really care if it's code compliant in all areas - that's your
choice - and as I've said before, making it NEC compliant is
commendable, especially from a safety perspective. I'm just curious as
to whether or not you've put as much effort into all the other codes
that might apply as you seem to be to putting into the UPS
connections.

To answer your question I have put disproportional effort into this
posting. I have done so because I personally don't like it when I
spend the time to help someone out and they don't bother to reply. I
have also learned a little something and hopefully so have the other
participants. I consider the room reasonably safe for my family and
self in all regards. I apologize but I do not wish to revisit the
entire scope of the project.

The practicality or cost effectiveness of your installation are things only
you can judge. I only offer my two cents to help, and in the process I
usually learn a few things


" wrote in message
oups.com...
On Oct 9, 7:08 am, "RBM" rbm2(remove wrote:
My assumption is that one uses a UPS to avoid power interruptions, giving
you a short time to safely save data and shut down. How is a system that
you
have to manually throw a switch, assuming you are even there at the time,
going to help?

Greetings,

The addition of the DPDT switch makes the installation more
professional and more robust with increased applications and value
after the conclusion of the project with minimal extra work. Do you
think it is fair to call the flanged input and DPTP switch a "single
circuit generator input with manual transfer" when showing the house?
Do you think it will increase the home value more than the cost of the
switch?

I don't plan on ever flipping the switch in my installation. When the
power goes out I get notified on my phone. If it is out for very long
I have to plug the UPS into a generator or else the systems go into
hibernation. When the power comes back on the systems come out of
hibernation and continue as they were.

Hope this helps,
William

On Oct 9, 2:30 pm, "RBM" rbm2(remove wrote:
The practicality or cost effectiveness of your installation are things only
you can judge. I only offer my two cents to help, and in the process I
usually learn a few things

" wrote in message

oups.com...

On Oct 9, 7:08 am, "RBM" rbm2(remove wrote:
My assumption is that one uses a UPS to avoid power interruptions, giving
you a short time to safely save data and shut down. How is a system that
you
have to manually throw a switch, assuming you are even there at the time,
going to help?

Greetings,

The addition of the DPDT switch makes the installation more
professional and more robust with increased applications and value
after the conclusion of the project with minimal extra work. Do you
think it is fair to call the flanged input and DPTP switch a "single
circuit generator input with manual transfer" when showing the house?
Do you think it will increase the home value more than the cost of the
switch?

I don't plan on ever flipping the switch in my installation. When the
power goes out I get notified on my phone. If it is out for very long
I have to plug the UPS into a generator or else the systems go into
hibernation. When the power comes back on the systems come out of
hibernation and continue as they were.

Hope this helps,
William



yes. you (and several others) have been helpful
thank you!

" wrote:

On Oct 8, 10:10 pm, "Pete C." wrote:
" wrote:

On Oct 8, 6:44 pm, "Pete C." wrote:
Tony Hwang wrote:

Pete C. wrote:

" wrote:

On Oct 8, 4:11 pm, "Pete C." wrote:

HeyBub wrote:

wrote:

Greetings,

I have two computers, one in the basement and one on the third floor,
which I would like to run off the same plug-in UPS (expensive, so I
don't want to purchase another one). There is already conduit going
up to the third floor. What do I need to properly (to-code) install a
UPS protected outlet on the third floor which shares the plug-in UPS
in the basement? Is there a code-compliant solution? Are you allowed
to plug a cord into a UPS which goes into a junction box, is spliced
to THHN, and then into the house wiring?

Thank you for your time,
William

PS: I am not asking for instructions on how to install an outlet. The
outlet that the UPS plugs into in the basement is already GFI
protected. There are enough spare cubic inches in the box to support
an additional device and wiring.

It will cost more to run another wire from the basement to the stars than
the $75 for a UPS.

I'm pretty sure copper prices aren't that high yet, but yes, while a
central UPS is nice, with today's UPS prices distributed UPSes are more
practical.

A new UPS for the third floor would cost $300-400,

Exactly what UPS are you looking at? You certainly don't need a UPS
anywhere near that cost to run a single computer, particularly these
days with LCD monitors. You're not trying to UPS a laser printer are
you???

consume additional
electricity,

Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

and take up space.

The UPS to handle one computer will be about the size of a shoe box, and
shouldn't be any issue to position along with the computer itself.

Then I would have to replace the
batteries in both of them every few years at an increased cost instead
of just the one.

Not likely. My big UPS with my server rack in my garage is now about 10
years old, on it's original set of batteries and is showing no sign of
any problems. The server rack loads it to about 15% capacity and just a
couple weeks ago there was a power failure where it ran some 60 minutes
on battery until I switched to generator power when the UPS reported 20
minutes of battery remaining.

The upstairs UPS, replacement batteries, and power
consumption (including cooling power) would probably come to $750 or
more over its lifetime.

Wildly exaggerated estimate, unless you're counting it's lifetime as 30
years and even then it's probably a high estimate.

If a 42U rack costs $1000 delivered then just
the space in the rack the new UPS would take up could be considered to
cost $100.

What planet are you on buying new racks delivered? Racks are a dime a
dozen used, often free and there is no reason to buy new unless you are
building a showcase data center at which point you still need redundant
UPSes, redundant generators, redundant air conditioning, access floors,
etc.

The dual-UPS solution is NOT the most economical solution.

It is by far, once you open your eyes to the real world and not some
fantasy idea of a shiny enterprise data center.

Spend the $100 of a decent UPS at Fry's, Costco, etc., plug it in
upstairs and get on with your life.

(The data centers I deal with have UPSes and generators rated in MW
total capacity, and $250K/month electric bills)
Hi,
Let me guess, OP'er came here to argue because he does not have real
something to do himself. If he has a wish list, dive in and do it
instead of wasting time arguing with everyone. Military installations
I used to work on went out some times with all that high tech redundancy.
As far as I remember no system has 100% up time. Maybe 99.9999%

The "five 9s" i.e. 99.999% is the gold standard used most places.
Perhaps the military adds an extra decimal place. But that 0.001% is why
you also mirror / RAID all disks, do daily backups and send them off
site, have redundant power supplies on all your CPUs and storage,
redundant network links, etc. Perhaps the OP is planning on hosting a
pay net porn site that will loose money if it's down, in which case it
should be hosted at a real data center anyway.

"What planet do I come from?" Let's try a planet where people use
numbers instead of emotions to make decisions.

I use numbers as well, not emotions.



A search for "rack mount ups" at froogle.com sorted by relevance
prices the five most relevant at:
$825, $440, $490, $504 and $650

Rack mount UPS = same guts as non rack mount UPS in a rack mount case
for 2X the price. Try a non rack mount UPS on a $20 rack shelf, that's
what I've used for years and it performs perfectly.



I clicked on the first one and it has a "670W Nominal Output Power",
and provides about 20 minutes of standby time at half load (325W)
which hardly seems excessive. My "kill-a-watt" meter puts my power
usage at just under 400W so I would get a little less than 20 minutes
with new batteries in the unit and less once the batteries age.

If you take the cheapest unit at $440 and pretend the wiring cost was
$140 you come to a $300 initial savings. If your discount rate is 8%
and the project life span is 10 years $300*1.08^10=$647.58.

$647.58 saved based on initial cost alone

Again, you are not making rational choices for the UPS, you're going on
cosmetics. You can get a perfectly acceptable Trip-Lite brand 1KVA UPS
at Costco for $100 add a rack shelf from another source for $20 and
you're done.



Then take a look at batteries:
http://www.newegg.com/Product/Produc...ggested&Descri...

First off, *never* buy replacement UPS batteries from the UPS
manufacturer, they are severely overpriced. Get replacement UPS
batteries from a local wholesale battery distributor, where they are
much cheaper for the same battery. I've used Mid State Battery in the
past and a set for a UPS rarely exceeded $40 with tax. The UPS is out of
warranty at that point so it's not like you have to use the "branded"
batteries to maintain warranty.

Notice the manufacturer states that they are good from 3-5 years.
This makes since because it matches the life expectancy of lead-acid
car batteries and also matches my personal experience.

Car batteries operate under very different conditions from UPS
batteries. Car batteries cycle from freezing temperatures to 150 degree
temperatures and back in the winter. Car batteries don't sit on a float
charger most of their life. Car batteries are called upon to produce
several hundred amps for 30 seconds and then recharge. There is pretty
much no valid comparison between UPS battery life and car battery life.



If the cost of the batteries+shipping for the unit is $50+ (and it is
more) two replacements over a 10 year life expectancy will add another
$100 or so to the total cost.

Again the life expectancy you are going by is not correct. It may be a
good number for a large corporation to use when replacing batteries in
hundreds of UPSes, but it's silly for a home user where you have a
couple UPSes and can easily load / runtime test them annually. As I
noted, the batteries in my big (1.42KVA) UPS are a decade old and still
performing perfectly under load.



We are up to $747.58. If the additional UPS uses only 20W and 10W for
cooling 30W at $0.10 per kWh over 10 years comes to $263 which brings
us to about $1010 difference. If the space is free and somehow I have
overestimated by $260 we are still talking about a $750 difference.

The standby consumption is the only place you're even close to reality,
and the $2.19/mo is pretty insignificant.



If my math is wrong, please show me. Otherwise please help with my
ORIGINAL question. ... and NO ... I am not switching from rack mount
equipment to non-rack mount for a whole variety of reasons that are
even more off the topic of home repair than I have been drawn to by
the erroneous responses about UPS system TCO. Please answer the
original electrical question!

If you want to waste money for some sort of ego / cosmetic reasons, by
all means feel free to do so, certainly people waste money of far more
silly things. Don't try to "cook the books" to try to justify the extra
cost for cosmetics as some sort of savings though.

My central server, network gear and a few other extraneous items along
with the big UPS are in a standard equipment rack in my garage. The rack
is used and was free, most of the equipment in it is not rack mount and
is on rack mount shelves, though the main ethernet switch is rack mount.

As for the original question:

NEC Article 645.10 "Disconnecting means" indicates that a means of
disconnecting power to all electronic equipment in the information
technology room must be provided as well as similar disconnecting means
for HVAC and closing smoke / fire dampers, and the controls must be
grouped and located at the principle exit doors.

NEC Article 645.11 indicates that a UPS of 750VA or lower capacity
installed within the information technology room is exempt from article
645.10 disconnecting means requirements.

These are the sections of the NEC that reference UPS systems, and
clearly aren't intended to be applied to a residential environment.
There are other articles for separately derived systems and multiple
power sources, none of which are directly applicable.

Ultimately it comes down to the discretion of your local building
official (AHJ) to interpret and apply the code and their own expertise
to determine what they will require / accept. Since there is no directly
applicable NEC code you are entirely at their mercy as to what they'll
let you do.

The AHJ may simply insist you use a separate UPS so that the
disconnecting means are clearly recognizable to the local fire
department responding to your server fire. They may accept the remote
outlets if the connections are entirely isolated in separate conduit
from the regular household circuits. It's entirely up to the AHJ and
nobody on the 'net can give you the precise answer.

The NEC is available online for your review, someone posted the link
here recently I believe.



PS: I do have systems at a hosting facility. These systems do not
need to be at a colocation facility or they would be. JEEESH!!!!

That's nice. Do you have a backup generator to keep things online for
more than the ~20 minutes of the UPS(es)? If not, that's a better place
to spend money than purty racks.

Greetings Pete,

Your math is wrong. It uses prices for non-rackmount equipment and
generics.

Bull. Trip-Lite is not a generic brand, and as I noted, the rack mount
UPSes from a given company are the same guts as the non rack mount
versions, only at a much higher price. If you don't want to recognize
that reality, you're probably the type to also buy $100 power cords to
make your CD player sound better.

This is not the class of equipment which I am using for
this installation. I know you have strong pro-generic anti-rackmount
feelings but please try to put them aside.

Your idea that non-rack mount equipment is somehow a different class is
truly laughable. Rack mount has nothing whatsoever to do with the
quality of a piece of equipment, it is only a mounting standard and as I
noted from most any given manufacturer the exact same guts are available
in rack mount and non rack mount versions.

This is not what all my
clients like to see me use even in the dev environment and they are
right by definition.

If your trying to dress a set to impress clients, admit that it is a
cosmetics issue, don't try to justify the cosmetics with distorted and
just plain false calculations.


Insisting on rack mount / name brand equipment does not equate to
cooking the books. Period.

Trying to make it appear that there is some advantage other than
cosmetic to buying overpriced rack mount equipment is indeed cooking the
books.

Your entire argument stems from the fact
that I could do it cheaper if I used non-rack mount generics. You
have not pointed out a math error!

I have pointed out a common sense error. If you need to try to make
things cosmetically flashy to try to impress technologically ignorant
clients that is an entirely different issue.

Lawyers could probably drastically
reduce their clothing / dry cleaning costs if they wore jeans and a
white t-shirt but clients wouldn't like it. When client perception is
taken into account it is cheaper to pay for the suits.

There you go, the entire basis for your rack mount fetish - cosmetics to
try to impress clients.


Yes. I have a backup generator and speaking of backup generators with
"central UPS" I can simply plug this generator into the power inlet in
the basement through the basement window (indirectly by plugging the
central UPS into the generator). Without "central UPS" I have to run
the generator cord in through the basement window and then snake it up
three flights of stairs with extension cords!

That's pretty lame, I have my generator located next to my detached shop
and connected at my main panel with an approved interlock kit. I don't
have to drag any cords anywhere to bring it online. Sounds like a better
area for you to invest your time, effort and money in than running a UPS
line upstairs. Get a standalone UPS for upstairs and put it behind a
nice brushed anodized aluminum rack mount filler panel.


That said, thank you for your NEC notes.

--William

mm wrote:

On Mon, 08 Oct 2007 22:14:14 GMT, "Pete C."
wrote:


Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

Then why is my one-computer UPS so hot if I leave it on while not
using the computer. Not just in one spot, but over the entire top
surface and some of the sides.


Put a 7W night light in a closed shoe box all day and see how hot it
gets. Look at a 10W soldering iron melting solder. Heat will build over
time from even a quite small source if the package design doesn't
provide very good dissipation. My main UPS runs very cool, buy it does
have a full time fan at low speed, which switched to full speed when the
UPS switches to inverter.

On Oct 10, 12:44 am, "Pete C." wrote:
mm wrote:

On Mon, 08 Oct 2007 22:14:14 GMT, "Pete C."
wrote:

Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

Then why is my one-computer UPS so hot if I leave it on while not
using the computer. Not just in one spot, but over the entire top
surface and some of the sides.

Put a 7W night light in a closed shoe box all day and see how hot it
gets. Look at a 10W soldering iron melting solder. Heat will build over
time from even a quite small source if the package design doesn't
provide very good dissipation. My main UPS runs very cool, buy it does
have a full time fan at low speed, which switched to full speed when the
UPS switches to inverter.

You chose two examples which were both invalid. The tip of the 10W
soldering iron has a tiny surface area to dissipate the heat from. In
the enclosed example you don't talk about dissipating heat but the
heat trapped in the box. mm says the entire surface area of his UPS
is hot and dissipating heat. A better example would be that even a
40W fluorescent tube is not hot to the touch.

" wrote:

On Oct 10, 12:44 am, "Pete C." wrote:
mm wrote:

On Mon, 08 Oct 2007 22:14:14 GMT, "Pete C."
wrote:

Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

Then why is my one-computer UPS so hot if I leave it on while not
using the computer. Not just in one spot, but over the entire top
surface and some of the sides.

Put a 7W night light in a closed shoe box all day and see how hot it
gets. Look at a 10W soldering iron melting solder. Heat will build over
time from even a quite small source if the package design doesn't
provide very good dissipation. My main UPS runs very cool, buy it does
have a full time fan at low speed, which switched to full speed when the
UPS switches to inverter.

You chose two examples which were both invalid. The tip of the 10W
soldering iron has a tiny surface area to dissipate the heat from. In
the enclosed example you don't talk about dissipating heat but the
heat trapped in the box. mm says the entire surface area of his UPS
is hot and dissipating heat. A better example would be that even a
40W fluorescent tube is not hot to the touch.

Perfectly valid examples of how even a small amount of heat generated
can build over time if it's not being dissipated. I don't know what
planet you're on, but a 40W fluorescent tube most certainly does get hot
with extended use in a normal environment. Not as hot as a 40W
incandescent, but far hotter than ambient.

On Tue, 09 Oct 2007 12:17:56 -0400, mm
wrote:

[snip]


Of course that would be the speed of light in copper. I held a lamp
up to; one side of a copper sheet, and it hasn't gotten through to the
other side yet. Should I wait another hour.

[snip]

That reminds me of "slow glass" which is physically the same as
regular glass, but optically it's 10 light-years thick. When you look
into it you see what was on the other side 10 years ago. :-)
--
76 days until the winter solstice celebration

Mark Lloyd
http://notstupid.laughingsquid.com

"I have found Christian dogma unintelligable. Early
in life I absented myself from Christian assemblies."
-- Benjamin Franklin

"Pete C." writes:

Your entire argument stems from the fact
that I could do it cheaper if I used non-rack mount generics. You
have not pointed out a math error!

I have pointed out a common sense error. If you need to try to make
things cosmetically flashy to try to impress technologically ignorant
clients that is an entirely different issue.

He should also avoid showing his setup to his technologically
sophisticated clients. He doesn't want one saying "Why did you run a
circuit all the way from the basement instead of just adding a new UPS
in the attic? That would be cheaper and give more run time."

Dave

mm writes:

Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

Then why is my one-computer UPS so hot if I leave it on while not
using the computer. Not just in one spot, but over the entire top
surface and some of the sides.

What make and model? There *are* true UPSes that convert line current
to DC and then back to AC continuously, running the inverter all the
time. The supposed advantage is that there's no discontinuity in the
output waveform when the input line voltage goes away, but these UPSes
are larger and heavier and dissipate significant power continuously.

Or you could have a unit with a ferroresonant voltage regulator, which
wastes about 20% of its full load output rating in heat continuously.

But most people use standby power supplies that switch on only when AC
fails. These dissipate little power when on standby (just enough to
charge the batteries and run the electronics). Voltage regulation can
be provided by tap-switching transformers, again with little heat
generated (some SPSes include this function).

I have a variety of APC SPS units, and none are more than slightly warm
to the touch when operating in standby.

Dave

On Wed, 10 Oct 2007 17:55:39 +0000 (UTC), (Dave
Martindale) wrote:

mm writes:

Unless it's a hybrid UPS with ferroresonant line conditioner, it's
standby / float charging power consumption is pretty negligible.

Then why is my one-computer UPS so hot if I leave it on while not
using the computer. Not just in one spot, but over the entire top
surface and some of the sides.

What make and model? There *are* true UPSes that convert line current

It's a Tripplite, but I don't know the model. (It's on the floor and
one of the cords from something plugged into it is too short to pull
the UPS out from under the desk, so it's very hard to see the model
number.) It's about 8" high by 5"wide by 7" deep and iirc has one 4,
or 6, AH battery inside. It has two tiny slide switches and 3 tiny
lights at the top front. Much harder to turn on and off than the
earlier one that had a big rocker switch at the very top.

to DC and then back to AC continuously, running the inverter all the
time. The supposed advantage is that there's no discontinuity in the
output waveform when the input line voltage goes away, but these UPSes
are larger and heavier and dissipate significant power continuously.

Or you could have a unit with a ferroresonant voltage regulator, which
wastes about 20% of its full load output rating in heat continuously.

But most people use standby power supplies that switch on only when AC
fails.

Bought mine used without a manual, but I think I dl'd the manual from
the manuf. website. But that might be on the previous harddrive.

These dissipate little power when on standby (just enough to
charge the batteries and run the electronics). Voltage regulation can
be provided by tap-switching transformers, again with little heat
generated (some SPSes include this function).

I have a variety of APC SPS units, and none are more than slightly warm
to the touch when operating in standby.

Dave