DIYbanter - View Single Post

Don Y[_3_]

On 12/28/2015 12:24 PM, wrote:
On Mon, 28 Dec 2015 10:54:56 -0700, Don Y
wrote:

On 12/28/2015 8:32 AM, dpb wrote:
On 12/27/2015 4:40 PM, Don Y wrote:
On 12/27/2015 8:45 AM, dpb wrote:
On 12/27/2015 8:24 AM, Don Y wrote:
I can't seem to get this right. : Can't put it on the
roof with the other sensors as the deliberate exposure that
the roof affords those others interferes with the accuracy
of the temperature reported.

And, mounting it on the back porch ("up" out of the way)
shades it from direct sunlight but "keeps it warm" in the
colder weather due to heat trapped by the roof of the porch
(the same roof that *shades* it from the Sun).

So, it seems like the trick is to find shade without an
"overhead covering"? (contradiction in terms??)

I'll try the porch -- but much closer to the ground (farther
from the heat-trapping ceiling)?

You've got a special situation; if you're serious about grove
protection I'd
suggest getting in touch w/ the UofFL citrus folk--

grin Not a "grove", just half a dozen trees.

There's only so much we can realistically do; can't deploy smudge pots
as there are too many structures nearby; can't deploy fans as the trees
aren't arranged in a "grove" (we'd need several small fans, instead).

We opt to cover the trees and use small heat sources to help encourage
convective air flow (has to work in case it's also raining).

I've covetously eyed the "blow up figures" that folks put out for
holiday decorations thinking that I might be able to repurpose the
fans used in them. They *should* be able to tolerate rain, right?

OTOH, they are covered by a large inflatable fabric sack so I'm
not sure how well they actually fare in a downpour! (do people
actually have them inflated when its raining??)

The more practical solution is to move to higher ground farther
from the wash.

Or, stop growing fruit that doesn't like the weather!! :

[I've repeatedly threatened SWMBO that when the next REALLY cold
spell comes through and whacks the trees, I'll use them in the
smoker!! : ]

I've not been to grandparents' old place in TX Valley in years to know what
they now use (of course, it's two-three generations removed from grandparents
now) but for temp measurement I'd think you'd get by pretty nicely with just a
covered box near where you really need the detail info; as you point out if
you're really after that local micro-difference to protect individual tree(s)
it may not be enough unless you go for it in an all or nothing fashion whenever
you do get an indication (likely the most reasonable alternative). But, as
another I think noted, I've got a Davis Instruments weather station that seems
to be quite accurate simply with the sun shielding they come with "out of the
box" -- it sits out in the midday sun in SW KS at 100+ F w/o measuring
ridiculously high temps that occur from directly-exposed thermometers.

The standard thermometer on the north side of the house at the top of the porch
also has always seemed pretty good -- it's 8' or so, roughly from actual
outside wall of the house so the heating effects don't seem very influential as
far as can tell...

Radiate heat from the (masonry) house isn't the problem, IMO.
Rather, it is the heat trapped under the roofline that elevates
the temperature registered by the thermometer/sensor.

I would like to come up with a "solution" that others can
recreate (as I assume others will be interested in much of
the technology that I've developed -- and I'm not interested in
going into manufacturing! : )

COTS "weather stations", IME, are just "curiosities". They give
the user/owner a sense of what the weather is like -- without having
to be precise (note "precise" and "accurate" have different technical
meanings).

E.g., an indoor thermometer doesn't have to be accurate at all!
Do you *really* know if it is 68F in the house? 71? What's
the temperature at the other end of the hallway? In the kitchen?
etc.

All you want from the "thermostat" is to keep the temperature AT
a particular point -- and, some way of referring to that point
(we could call it "foogle" just as easily as "68") as well as
points ABOVE and BELOW ("elgar" is just below "foogle" and "gismatch"
is just above! : )

OTOH, if you know that a particular fruit is hardy to 32F (or 28F),
then you don't want to play guessing games wondering if 32 is
"elgar", "foogle" or "gismatch"!

You also want to know the *actual* relationship between these so
you can perform reliable mathematical operations on them: "How much
energy do I have to expend to create a local "foogle" microclimate
when the outdoor temperature is "elgar"? Or, "dingbrill"? At which
point will I no longer be able to influence the microclimate based on
these calculations??

[Someone with the sort of interest I'm concerned about will be
willing to sort out *proper* placement of a sensor instead of
just relying on what's convenient]
Have you tried the chunk of 6 inch plastic pipe??
You get a chimey effect that moves the sir through the pipe when the
sun shines - negating any radiant heat absoption, you get full time
shade, and no heat entrapment. Put it on a pole out close to the
trees. It will be about the most accurate temperature record you can
get, across all variables. Mount the temp sensor in the middle of the
pipe. 2 or 3 feet long is adequate - longer won't hurt

I don't think that would be a viable option for most folks.
It's one thing to have to make an effort to "put it in the right place";
another thing entirely to have to locate something that is that physically
large and visually unappealing! Likewise, I need to come up with
a smaller "rain detector"... one that isn't as visually obtrusive
(and, ideally, requires less maintenance).

[People don't want to make "careers" out of this sort of thing! When
was the last time you calibrated your household thermostat?]

I'm going to investigate radiation shields and see if I can come up
with an approach that is easily reproduced that affords greater
flexibility in siting. I can make a bunch of sensors, calibrate
them (to each other) and set them in different shields located in the
same general area. Take data, then relocate the group to other
locations and see how they perform compared to each other, etc.

I'll talk to someone at the university and see what they say
about setting up a "control" against which to measure them.
That way, just let a machine do all the data collection in a
more repeatable fashion instead of playing wack-a-mole with
different sensors, placements, etc.