If you were Registered and logged in, you could reply and use other advanced thread options
|
Posted by Joseph on March 4, 2007, 4:30 pm
> >
> > Your grasp of the painfully obvious has at least been
> > consistent over the years, as has your overinflated pride in it.
>
> I suppose for those with short attention spans the way I contributed
> the link to my own article could unwittingly qualify as an
> unnecessarily distraction sufficient to cause spontanious diversions
> from the intended subject.
I love it when you talk dirty... :)
>
> But also, you know what, you and I both have high IQ's. You probably a
> higher EQ than I and possibly, in some areas, also a higher IQ than I.
> But the content of my recent articles is borne precisley out of this
> realisation. That what is obvious to me, and a few others, is just not
> obvious to the majority. If the content of my articles are obvious to
> you then obviously they are not aimed at you.
>
> My articles are for the majority. They are for the people :)
>
> You and I know just how clever "the people" are. It was precisley the
> people who voted Bush and Blair in. The consequences were not obvious
> to the people. My articles are aimed at precisely those who have
> different opinions about what is obvious :-)
>
|
|
Posted by Marc O'Brien on March 4, 2007, 10:49 am
On Mar 3, 11:44 pm, .p.jm@see_my_sig_for_address.com wrote:
> wrote:
>
> >People, if you don't yet subscribe to the UK's RAC magazine then
> >here's a good reason why you should. They publish my articles in the
> >associated RAC Service Engineer Magazine. Of course this argument can
> >be substantiated - everyone should read my articles and we all doubt
> >anyone would disagree.
>
> Bwahahahaha :-)
>
>
>
> >Thoughts on better application of mini splits in server rooms. It has
> >only been out a few days but already I've received a hand full of
> >phone calls from suppliers and contractors wanting to discuss the
> >principles further.
>
> >The link is to a low resolution scan of the most recent issue's
> >article.
>
> >http://www.hvacrforums.com/uploads/design/closecontrol/rac_se_spring_...
>
> >Let me know what you think.
>
> Your grasp of the painfully obvious has at least been
> consistent over the years, as has your overinflated pride in it.
The image linked to below shows me, more clearly, walking through
strip curtains which along with the perspex sheets seen spread across
the top of the racks all help to contain the cold air within the cold
isle.
Not seen before and therefore not likely "painfully obvious" :-)
http://www.hvacrforums.com/uploads/design/closecontrol/contained_cold_isle.jpg
|
|
Posted by on March 4, 2007, 11:46 am
wrote:
>On Mar 3, 11:44 pm, .p.jm@see_my_sig_for_address.com wrote:
>> wrote:
>>
>> >People, if you don't yet subscribe to the UK's RAC magazine then
>> >here's a good reason why you should. They publish my articles in the
>> >associated RAC Service Engineer Magazine. Of course this argument can
>> >be substantiated - everyone should read my articles and we all doubt
>> >anyone would disagree.
>>
>> Bwahahahaha :-)
>>
>>
>>
>> >Thoughts on better application of mini splits in server rooms. It has
>> >only been out a few days but already I've received a hand full of
>> >phone calls from suppliers and contractors wanting to discuss the
>> >principles further.
>>
>> >The link is to a low resolution scan of the most recent issue's
>> >article.
>>
>> >http://www.hvacrforums.com/uploads/design/closecontrol/rac_se_spring_...
>>
>> >Let me know what you think.
>>
>> Your grasp of the painfully obvious has at least been
>> consistent over the years, as has your overinflated pride in it.
>
>The image linked to below shows me, more clearly, walking through
>strip curtains which along with the perspex sheets seen spread across
>the top of the racks all help to contain the cold air within the cold
>isle.
>
>Not seen before and therefore not likely "painfully obvious" :-)
>
>http://www.hvacrforums.com/uploads/design/closecontrol/contained_cold_isle.jpg
>
Now let's see your data re : 'with and without', IOW, basic
'cold aisle' measurements without strips, vs with. Of particular
interest would be a widely measured ( IOW, at each server rack, low,
mid, and high ) set of datapoints of server entering and leaving air
temps. Empirical data, not theories. If you can demonstrate an
actual delta of let's say ~ 3 F or more leaving-server average temp (
with all other factors being equal, IOW to purely demonstrate the
effect of the strips and nothing else ).
IOW, standard 'computer room cold aisle airflow design' vs
'with your strips'. You have a test case installation available, it
would seem, so simply take the measurements and document them - you
could do it in a day or less.
Measure the leaving air of each server fan, at the server back
panel, in the fan air stream, and note the numbers. Now, remove your
strips ( just tie them back real good ) and other barriers, wait an
hour or 2 for thing to equalize, and take the numbers again. Post
them here.
If the differences are < 3 F, you have a paper-wonder. If
not, you have a wonderful sales point / proof I just gave you for
free.
BTW, measure DB only. Nothing else matters to the computers (
assuming we don't get into static charge discussion - let's stick with
actual server cooling ).
This test assumes, of course, that the ductwork / registers
etc are properly designed and installed to begin with such that the
barriers are not merely compensating for a bad initial system design.
One rule of thumb about 'great ideas that are practically free
and no one ever thought of' - someone else probably DID, and there's
probably a REASON it's not in common practice.
I'm reminded of the many ' improve your gas mileage' things
for sale. One I saw on the 'net was a little stamped metal plate with
blades stamped down to turbulate the air entering the carbs. 'Just
slip this in, and increase your mileage and horsepower by30 % or more
!!! Only $ 69.95 plus shipping !
Now, it was a piece of maybe 18 gauge sheet metal with the
most basic stamping operation you could imagine. The kind of thing
that would cost GM or Ford or such maybe $ 0.03 each( or maybe half
that much ). If they could increase their mileage / HP ratings (
which are HUGE sales / marketing points ) that much for that little,
how long do you think they'd wait ? In fact, they wouldn't even have
to buy the part, just re-design the intake manifold a tiny bit, and
have ZERO cost per unit to do it.
But they don't. Hmmm...... why ? Maybe because
it's BULLSHIT :-)
--
Click here every day to feed an animal that needs you today !!!
http://www.theanimalrescuesite.com/
Paul ( pjm @ pobox . com ) - remove spaces to email me
'Some days, it's just not worth chewing through the restraints.'
'With sufficient thrust, pigs fly just fine.'
HVAC/R program for Palm PDA's
Free demo now available online http://pmilligan.net/palm/
|
|
Posted by on March 4, 2007, 12:01 pm
On Sun, 04 Mar 2007 11:46:13 -0500, .p.jm@see_my_sig_for_address.com
wrote:
> Measure the leaving air of each server fan, at the server back
>panel, in the fan air stream, and note the numbers. Now, remove your
Get the entering air for each while you're at it, too. Then
show the deltas for that also.
--
Click here every day to feed an animal that needs you today !!!
http://www.theanimalrescuesite.com/
Paul ( pjm @ pobox . com ) - remove spaces to email me
'Some days, it's just not worth chewing through the restraints.'
'With sufficient thrust, pigs fly just fine.'
HVAC/R program for Palm PDA's
Free demo now available online http://pmilligan.net/palm/
|
|
Posted by Marc O'Brien on March 6, 2007, 2:24 pm
On Mar 4, 4:46 pm, .p.jm@see_my_sig_for_address.com wrote:
> wrote:
> >On Mar 3, 11:44 pm, .p.jm@see_my_sig_for_address.com wrote:
> >> wrote:
>
>
> >> >The link is to a low resolution scan of the most recent issue's
> >> >article.
>
> >> >http://www.hvacrforums.com/uploads/design/closecontrol/rac_se_spring_=
.=2E.
>
> >> >Let me know what you think.
>
> >> Your grasp of the painfully obvious has at least been
> >> consistent over the years, as has your overinflated pride in it.
>
> >The image linked to below shows me, more clearly, walking through
> >strip curtains which along with the perspex sheets seen spread across
> >the top of the racks all help to contain the cold air within the cold
> >isle.
>
> >Not seen before and therefore not likely "painfully obvious" :-)
>
> >http://www.hvacrforums.com/uploads/design/closecontrol/contained_cold...
>
> Now let's see your data re : 'with and without', IOW, basic
> 'cold aisle' measurements without strips, vs with. Of particular
> interest would be a widely measured ( IOW, at each server rack, low,
> mid, and high ) set of datapoints of server entering and leaving air
> temps. Empirical data, not theories. If you can demonstrate an
> actual delta of let's say ~ 3 F or more leaving-server average temp (
> with all other factors being equal, IOW to purely demonstrate the
> effect of the strips and nothing else ).
Sorry I didn't get back earlier. Been busy.
The strip curtain and perspex sheeting idea was a pleasant surprise to
me. Neither were my idea. They were the idea of one of our clients. A
customer in his early 20's who so far has grown his data hosting
business even faster than we have grown our own 3 business's. We
started out about the same time.
We had worked together already planning the latest new data room
layout to establish the traditional hot and cold isle setup which we
ourselves had only come to know of through analysing, with our client,
the problems experienced on two previous premises. Admittedly, it was
our client who had the most free time to research the principles. The
traditional hot and cold isle setup was something we had overlooked on
both their previous premises where split systems were used. The new
site is now using proper close control down flow units with SHR's of
around 0.99. But these new CRAC Units (Computer Room AC's) were still
seeing return air temperatures mid range between the floor discharge
areas and the hot isle hotter spots. Ideally we wanted the hottest air
stream in the rooms to be in those areas nearest the CRAC intakes. And
the coldest air streams consistent in temperature between the CRAC
discharge and the server inlets. When the hot and cold air mixes
before returning to the CRAC's then the CRAC's set points have to be
turned lower to ensure cool air onto the servers. Too, if hot air
mixes with the cold before going onto the servers then again the
CRAC's set points have to be turned down further to compensate.
So the first problem was obviously that cold air entering the cold
isles was still, here and there, bypassing the servers excessively. We
talked about it for a while. But didn't come up with any ideas. When
we talked about the problem we did delve a little into the concept APC
and Trox are using for high density blade servers. Traditional server
systems produce about 2.5kW per rack, that is per frame. Blade servers
are producing in excess of 20kW per rack. With the the newer blade
servers you just cannot achieve sufficient server cooling without
physically bolting the CRAC unit onto the back of the rack to ensure
the very higher volumes of air required are drawn assuredly straight
through the servers. The traditional layout of CRAC's located either
side at the end of hot isles is just plain futile when it comes to the
newer blade servers.
What APC have done now for blade servers is an enclosed room at the
back of the racks but as part of the racks. That enclosed room is
where all the cables come in. On the other side of the enclosing room,
is the bolt-on CRAC system. So, the air at 20=B0C in the server room
enters the blade servers, it exits the blade servers at about 40=B0C. So
the enclosed room is itself at about 40=B0C. This high velocity air then
gets drawn directly into the CRAC unit which expels it back into the
server room at about 20=B0C. In other words, APC have originally bolted
a CRAC directly onto the back of the blade servers but then realised
there is a cable access hindrance. So they've then separated the CRAC
from the back of the blade server racks just enough to allow a body to
walk behind the racks. Then they have re-enclosed the cable access
space with a wall and door either side and a perspex roof so that the
server rooms lights can still light up the area behind the racks but
the hot air leaving the servers is not diluted by cooler room air
before entering the CRAC.
It was from the discussion on APC's methods of cooling the new blade
servers that our client came up with the idea of strip curtains and
perspex to enclose his cold isles. APC enclose the hot isle. Our
client has enclosed the cold isles.
Some related notes...
http://www.apcmedia.com/salestools/TDOY-5U362W_R2_EN.pdf
> IOW, standard 'computer room cold aisle airflow design' vs
> 'with your strips'. You have a test case installation available, it
> would seem, so simply take the measurements and document them - you
> could do it in a day or less.
>
> Measure the leaving air of each server fan, at the server back
> panel, in the fan air stream, and note the numbers. Now, remove your
> strips ( just tie them back real good ) and other barriers, wait an
> hour or 2 for thing to equalize, and take the numbers again. Post
> them here.
Actually, I am keen to get back in soon and take readings similar to
what you ask. I like gathering numbers.
The idea in my article is simply that the same ideas can be used to
improve the performance and reliabilty of mini splits in data rooms.
Essentially, I explain why mini splits are not suited but how the
situation can be improved. We need to just buck up and accept, people
do you use mini splits in data rooms, this wont stop, so, let's just
make more sense of the habbit.
|
Page 2 of 5 < 1 2 3 > last >>
| Similar Threads | Posted | | Re: Bi-Flow Gas-Line Filter Driers & Inverter Mini-Splits | March 13, 2008, 2:45 am |
| Application Engineer Opportunity - Multiple Roles including Dubai | February 7, 2008, 1:31 pm |
| Unexpectedly cold rooms in zone-valved big house | February 8, 2007, 12:50 am |
| Seeking Suggestions/Tips before covering up the central A/C for the season. | August 27, 2007, 8:01 pm |
| Mini split Wiring help please | August 19, 2006, 11:40 am |
| Mini split issue | October 18, 2006, 7:55 pm |
| Which Mini Split for Coastal Area | August 19, 2006, 3:00 pm |
| Mini-split with auxiliary heat | November 11, 2006, 11:00 pm |
| Mini-Split Universal remote | November 13, 2006, 2:08 am |
| Running a mini refrigerator upside down? | February 20, 2008, 2:09 pm |
|
|
Yahoo! 
Google 
Windows Live 
del.icio.us 
digg 
Netscape 
XML Sitemap