Data Collection ...
Part # 2 (Jan 14/06)...
I
want to take some accurate and meaningful data readings off the Passive Solar
Collector, and the "Min/Max - In/Outdoor" digital thermometer is an
aid toward that direction. The "Outdoor" or remote sensor is actually
placed down into the top of the center column of the collector, while the
"Indoor" sensor is built into the body of the display (effectively
measuring the air temp at about 4ft off the ground.
The
Black unit mounted below is measuring the true Outdoor temperature via a remote
sensor but does not have the "Min/Max" recording function.
With
this arrangement I recorded the display every 30 minutes from 6:30AM right
through to 6:00PM of the same day.
Let
me tell you, that type of stuff sucks, for the first 2 or 3 hours I was just
barely within a 5 minute window of the 30 minute reading period... for the rest
of the day I was so concerned that I'd miss a reading that I progressively did
less and less in the shop, until I started to count the half-hours until the sun
would be completely set and all was dark.
The
worst part of that exercise was that the day turned out to be dull and overcast,
although it did have promise early on.
I
took my readings and pumped them into Excel and created the graph shown to the
left.
Yellow
is the output of the collector, Green is the inside temp, Blue is the outside
temp and Red is my subjective interpretation of the ambient light conditions on
a scale of 0 to 5, with 0 being darkness and 5 being light cloud cover...
...it
never got what I would consider direct unobstructed sunlight (and still hasn't
to this day).
After
reading several web-sites of others that have built passive collectors, one site
introduced the idea of a Data-logging device that would sample the temperature
at a predetermined frequency and log the values for later evaluation.
This
seemed like the perfect thing for me, with the exception of the cost...
Typically I was finding that for 3 to 4 sensors, software and what ever else to
get into this setup the price would be between $200 and $300 after shipping,
exchange between US and Can funds etc...
The
solution was found in the form of a Dallas Semiconductor IC - DS1615 an
all-in-one IC that can store thousands of entries, Realtime Clock, and a choice
of interface options. Dallas Semi was even good enough to send me 2 (two)
samples to ensure that I would be happy with their product...
The
schematic pictured above is the "Rube-Goldberg" implementation that I
ended up using. In my haste and enthusiasm for the DS1615 I neglected to order
the "Modern" RS-232 transceiver DS-275 per Dallas Semi's application
notes.
What
little I could recall about PC architecture was that the original RS-232 ports
had discrete driver/receiver chips in the form of the venerable 1488 & 1489.
These workhorse IC have fallen out of favour due to their requirement for a +/-
12V power source and 
individual
space requirement, while the DS-275 is a 1 (one) chip solution that can
communicate on a single 5V supply.
These
dusty old 486 class mainbrds were hauled out to supply virtually everything
except the DS1615.
Battery
clip, 32.768Khz crystal, power supply connector, even the 1488 & 1489 to
interface to the PC serial port.
This
is the assembled and working proto-type...
It
may not be pretty but it works...
The
Button battery if new could power the DS1615 for 10 years (per spec sheet), The
white power connector is only used during programming of the unit and retrieval
of data.
The
3-pin connector in the upper right is the serial interface.
The
entire setup is shown to the right including an old AT style power supply that
has to have a hard disk connected to it to act as enough of a load to
power-up...
But
like I said, it does work...
This
graph is the data collected through the evening and morning of our living room.
The
software also is available as a "Free" download from Dallas Semi.
The
unit was programmed to record a sample every 5 minutes.
The
quality of the graphics is reduced for this site and is NOT a reflection on the
actual output from the program.
I
saved the raw data to a txt file and was able to import it into Excel and the
graph as shown below.
As we
heat exclusively with wood, you can see when the last load was put in the
furnace around 9:00PM (21:00), luckily it was not a cold night as at 6:00AM it
was still about 19.5C.
The
curve from about 6:30AM onward through to about 11:00AM is a great
representation of what one armload of wood will provide with a forced air wood
furnace.
With
a working proto-type in hand, I'll order 5 or 6 DS1615, and perhaps even the
proper interface chips so that I can recover the desk space lost to the PC
Power-Supply.
The
Gee-Whiz factor of this project and the potential it holds has been a lot of
fun. Once the other sensors are completed (2 or three weeks) I'll be in a much
better position to assess the utility of the Passive Solar
Collectors.