For work I recently got one of those AMD 6400 dual cores, which runs idle at
50+ degC also that 300 + Million Transistor Graphics Card is up in that
temp. range...
If we don't talk about a 40 Gallon Household water heater, but instead some
smaller reservoir similar to what people used under the sink in years past,
at least there would be warm water for washing hands or hair - free
basically!
My solar concentrator for example that invests less than 3 kW during its
tracking hours manages to get my big tank often hotter than one can stand!
Temperatures above 43 deg C are considered that - too hot!

If people are using electricity to heat water, now they could feel good
about running their computer and crunching numbers all day!
Of course I have been proposing the same about waste heat of refrigerators,
recently I saw this article though:

A fridge that takes only 0.1
kWh a day
?
By
Tom Chalko, Mt Best, Australia, mtbest.net
First published in
Renew (Australia) issue 90
(2005)
There is no mistake in the title. This article
describes an inexpensive fridge that is 10 to 20
times more energy efficient than an average
fridge on the market. It also demonstrates that
the biggest limitations are our habits and
mediocre attitudes, not technology or cost.
My dream is to live a near-zero emission life.
Step by step I come closer to bring this dream to
reality. After all, the rainforest here at Mt Best is
so beautiful and so unique that I hesitate to
disturb it with any kind of pollution.
Insulating and double-glazing my RAL home
reduced my winter energy requirement to about
20 Watt per square meter of floor area. I do not
like "star ratings". I think they are misleading
and promote ignorance rather than understanding
of energy efficiency.
Reflective solar heating (described in issue 88 of
Renew) nearly halved my heating energy (and
the firewood) needs in winter. However, to
achieve a near-zero emissions I needed to find a
clean renewable source of energy to replace the
firewood altogether. The system of choice
became a geothermic storage heat pump system
that will be described in my next article.
However, since I generate my own electricity
from wind and sun, I needed to save some
energy to run the heat pump.
For almost 2 years I tolerated the fridge that I did
not like a little tiny bit. The list of things I did
not like about it is too long to mention here. If I
could eliminate this fridge, I would have enough
energy to run the heat pump=85
Chest fridge
Comparing the energy consumption of various
refrigeration devices available on the market I
noticed that well-designed chest freezers actually
consume less electricity than fridges of
comparable volume, even though freezers
maintain much colder temperatures inside. While
chest freezers typically have better thermal
insulation than fridges, there is another reason
for their efficiency.
Vertical doors in refrigeration devices are
inherently inefficient. As soon as you open a
vertical fridge door =96 the cold air escapes, simply
because it is heavier than the warmer air in the
room. When you open a chest freezer =96 the cool
air stays inside, just because it's heavy. Any leak
or wear in a vertical door seal (no seal is perfect)
causes significant loss of efficiency. On the other
hand, even if you leave the chest freezer door
wide open, the heavy cool air will still remain
inside. Have you ever wondered why chest
freezers in supermarkets have their doors either
wide open or not thermally insulated?
Designing refrigeration devices with vertical
doors is clearly an act against the Nature of Cold
Air. Shouldn't we cooperate with Nature rather
than work against it?
I become really curious just how efficient a chest
fridge can be. After contacting some leading
fridge manufacturers and discovering that no one
has ever made and tested a chest fridge, I
decided to make my own test. I bought a good
chest freezer and turned it into a fridge.
Turning a chest freezer into a chest
fridge
The main difference between a freezer and a
fridge is the temperature maintained inside.
Freezers maintain sub-zero (freezing)
temperatures down to =9625
o C, while fridges
operate somewhere between +4
o and +10o C.
Hence, turning a freezer into a fridge meant
changing the temperature control. Rather than
interfering with mediocre thermostat of the
freezer, I decided to install an external
thermostat to cut the power off when the
temperature of my choice has been reached.
For my research I bought a Vestfrost SE255
chest freezer with 600a refrigerant and a $40
battery-powered thermostat equipped with digital
temperature display and an internal 5A/240V
latching relay. The main feature of the latching
relay is that it consumes battery power only
during actual switching so that the thermostat
equipped with it is a true micro-power device
and its 2 AAA batteries last for many months.
Connection diagram (Fig 1.) is really simple.
Thermostat relay cuts the power to the freezer,
much like a light switch cuts the power to a
lamp. Thermistor (the temperature sensor) is
placed inside the freezer at the end of a thin 2-
wire flexible cable. I used the freezer drain hole
to pass the thermistor cable inside the cooling
compartment. An alternative is to insert it from
the top via the chest door. If the thermistor is left
near the bottom of the chest fridge =96 the
minimum fridge temperature is controlled by
thermostat. If the thermistor is located near the
top of the cooling compartment =96 the thermostat
will control the maximum temperature there. The
best position for thermistor is somewhere in the
middle.
It took me about 30 minutes to make all
connections. The most time consuming part was
removing the thermistor from inside the
thermostat (I cut it out from the circuit board
using wire clippers) and soldering it at the end of
a thin 2-wire flexible cable. I protected the
thermistor from moisture and mechanical
damage using shrink-wrap tubing and a tiny bit
of silicone.
The external thermostat can be installed
anywhere on the fridge or outside it. I have
decided to place it on the wall behind the fridge,
so that the temperature display was easy to read
at the eye-level.
I have also removed the interior light bulb, rated
15 Watts, because I avoid using energy wasting
devices as a matter of principle. I will consider
installing a LED interior illumination if I find a
reason for opening my fridge in the dark.
When I finished my connections I had a chest
fridge with a digital temperature display and a
temperature control at my fingertips.
Performance
I set the thermostat to +7
o C and switched on the
AC power via energy measurement gadget called
Sparometer. After about 2 minutes my
thermostat displayed +6.5
o C and the power to
the freezer was cut off. The temperature
continued to drop down to about +4
o C. I thought
that there was something wrong with the digital
display, because everything happened too
quickly. I took another thermometer and to my
surprise, it confirmed readings of the thermostat.
I watched the system for a few hours and then
decided to move the content of my old fridge to
the new one that I have just made. Since I have
never had a chest fridge =96 it took me some time
to arrange baskets and their content inside. I
placed the most frequently used items in top
baskets that slide on top edges of fridge walls. It
turned out to be a very practical idea. Not only
they are very handy there, but also I can take out
the entire basket, rather than taking out one item
at a time (a typical case with a vertical door
fridge).
In the first 24 hours my new chest fridge took
103 Wh (0.103 kWh) of energy. About 30% of
this energy was consumed during the initial
power up and re-arranging of the fridge content.
The room temperature varied from 21
oC during
the day to 15
oC at night. The fridge interior
temperature was kept between +4
o and +7o C.
The fridge compressor was working only for
about 90 seconds per hour. When the thermostat
intervened =96 the fridge consumed ZERO power.
The only active part was a battery powered
temperature display.
Results of my experiment exceeded all my
expectations. My chest fridge consumes as much
energy in 24 hours as a 100W light bulb does in
just an hour. Not only it is energy efficient. I
have never seen a fridge that was SO quiet. It
only works 90 seconds or so every hour. At all
other times it is perfectly quiet and consumes no
power whatsoever. My wind/solar system
batteries and the power sensing inverter simply
love it. With my new chest fridge I have power
to spare and I can use it to warm up my house in
winter with a heat pump. I wonder why no one
has ever thought of a chest fridge controlled by a
digital thermostat=85
What performance can I expect from my chest
fridge during hot summer days? In principle, the
energy consumption should be proportional to
thermal losses of the fridge, which in turn are
proportional to the temperature difference
between the inside and outside of the fridge. A
vertical door fridge has large additional losses
caused by opening the door and the associated
loss of cold air.
The power consumption data for my chest fridge
was measured for average 5
oC internal and
about 18
oC average ambient temperature (13oC
difference). If the average ambient temperature
rose to 31
oC, the temperature difference will
double to 26
oC. This in turn should double the
thermal losses and hence the energy
consumption. In this case I expect my fridge to
work about 3 minutes per hour.
In reality, doubling the temperature difference
causes slightly more than doubling of the energy
consumption, due to the reduced thermal
efficiency of the fridge heat pump (compressor
system). The larger the temperature difference
between the heat source and the heat sink, the
less efficient is the heat pump.
Fortunately for those who rely on solar power, in
hot summer months there is also more solar
energy to power the fridge...
It is obvious that a truly energy efficient fridge
does not cost any more money than a mediocre
one. It actually costs less. It also has extra
features, such as a digital temperature display
that gives you full control over the temperature
settings. So - WHY mediocre fridges are being
made? Why people continue to buy and use
energy wasting devices? Does anyone care?
Nearly every household on Earth has a fridge
that totally wastes at least 1 kWh of energy a day
(365 kWh a year). Some fridges waste 3kWh
each day. How much reduction in greenhouse
emissions can we achieve by banning just ONE
inefficient household device in just ONE
country? How many politicians debating for how
many years will it take to achieve such a ban?
Rather than waiting for someone to do
something I would like to volunteer to supply
modified chest freezers and/or freezer
modification kits to environmentally conscious
people of Australia. Let's do something in the
right direction right now.
Note:
Design details of my electronic thermostat
have been published in Renew magazine issue
93 (2005). The copy of this article is attached
below.
Fig. 1. Wiring diagram of the chest freezer
turned into chest fridge. The junction box is not
required if all connections are made inside the
fridge service compartment. Active (A)
connection passes via latching relay terminals
inside the thermostat.
Fig 2. Fully installed chest fridge with thermostat
on the wall. Note energy measuring gadget at the
bottom left.
Chest
freezer
Electronic thermostat
Junction box
thermistor
A
N
E
COM NC
My fridge-to-freezer
thermostat
By
Tom Chalko, Mt Best, tom@mtbest.net.
First published in
Renew, (Australia) issue 93,
2005
Since I published my article about my 0.1
kWh/day chest fridge in the issue 90 of
Renew I
keep receiving daily inquiries from all over the
world about the thermostat that I used to convert
my chest freezer into an ultra efficient fridge.
Apparently, my article has been posted and
discussed at quite a number of Internet Forums
and attracted attention of some academics and
researchers, some of whom contacted me
directly.
This article aims to answer those inquiries in
public. It describes details of the thermostat
system that I had to design myself, because I
couldn't find a ready-made unit that was able to
meet all my requirements.
The Jaycar QT7200 thermostat (described in
issue 92 of
Renew) that I tried first in November
2004, failed after about 6 weeks of use. Its relay
contacts fried up, because they were not rated for
inductive load of any kind to begin with. I had to
disconnect it in a hurry when the relay begun to
produce bright and loud plasma arcs that
illuminated my kitchen at night.
Although, in essence, the thermostat function is
very simple, design of a really good freezer-tofridge
thermostat system is not quite trivial.
There are some unexpected problems and
challenges that only become apparent when one
aims to design a system that works really well.
Let's begin by outlining general requirements for
the freezer-to-fridge thermostat.
Thermostat requirements
1. Reliability. Fridges need to be very
reliable household devices, simply
because our health depends on their
reliability. Excessive temperature
fluctuations due to any malfunctioning of
the thermostat accelerate food spoilage
and introduce the associated health risks.
The thermostat should work unattended
for years if not for decades.
2. Safety. The 240V power supply to the
fridge should be well insulated from all
low-voltage electronic components of the
thermostat.
3. Zero 240V power consumption during
the standby period (when the fridge
compressor is off). This requirement is
very important in the situation, when the
fridge is powered by an inverter that has
a power-sensing feature. Using zerostandby-
power appliances allows inverter
users to save up to 0.4 kWh per day just
by allowing their inverter to enter the
standby (sleep) mode at every
opportunity.
4. Hysteresis. The number of fridge
compressor starts per hour should be kept
as low as possible, not only to conserve
energy, but also to minimize the
compressor wear.
5. The thermostat should be easy to install
and should not require any modifications
to any freezer, so that a new freezer
warranty is not compromised in any way.
6. The thermostat should be simple and easy
to construct from readily available low
cost components
Zero-standby-power challenges
>From my experience with failing Jaycar QT7200
thermostat illuminating my kitchen with plasma
arcs it became obvious that a 240V relay used to
switch the fridge compressor on and off should
be properly rated for inductive load.
The general trend in modern industry is to
replace electro-mechanical relays and contactors
by solid-state semiconductor relays. In our case,
however, this clashes with the requirement 3.
Solid-state relays have significant capacitance
when off. This means that when they are
connected in series with a motor (a
resistive/inductive load) they allow a small
current to flow through the motor windings, even
when everything is off. This current causes a
continuous power loss of about 0.5 Watt as
measured with my Vestfrost freezer compressor.
Since semiconductor switches and relays cannot
meet the zero-standby-power requirement, we
have to consider them unsuitable for our
application.
Fig 1. Schematic of the zero-standby-power freezer-to-fridge thermostat. AC
supply earth connection (not
shown) must be carried from the AC power supply wire to the freezer supply
wire.
Hence, we need to settle for a properly rated
relay. After experimenting with a few brands and
designs, I decided to use OMRON G6RN-1A
DC12 relay. In addition to its ability to handle
small inductive load, and low-energy switching,
it has about 7kV insulation between its 12V coil
and its 240V contacts, which I consider essential
from the safety point of view. A number of them
work for 6 months now in thermostats that I built
with no signs of any deterioration in
performance.Due to the zero-standby-power
requirement, all electronics of our thermostat,
including the temperature-sensing system, need
to be powered from a battery-based power
supply. Since we also require our system to work
unattended for years (or decades) we have
another challenge to meet: design of a UPS
(Uninterrupted Power Supply) that can work for
many years unattended. The battery in this UPS
needs to be charged when the fridge compressor
is turned on.
The design
The schematic of the system that I currently use,
after making several not-so-interesting mistakes
and attempts, is depicted in Fig 1. It is a result of
a compromise between the minimal possible
power consumption, simplicity and the cost of
components. The temperature sensing system
consists of thermistor R1 (BC 2322 640 54103,
10k
@25=BAC) interfaced with an op-amp. The
LM324 quad op-amp chip that I selected has
quite low power consumption (<0.7 mA) and can
operate from single voltage power supply, which
greatly simplifies the design.
U1C and U1D serve as buffers, to minimize the
power consumption taken by the temperature
measurement and comparison system down to
negligible values. U1B is a summing amplifier.
U1A is a Schmidt trigger with easy to adjust
hysteresis (by changing R13), set here at approx
0.5=BAC. The diagram is annotated to be easier to
read. Capacitor C4 prevents radio signals that
appear on the long thermistor R1 cable from
interfering with functioning of the system.
The switch SW1 addresses the issue of powering
the system down (the center-off position) and
allows the thermostat to operate in two modes:
powered by mains 240V ("SW1 up", in which
case battery can be removed) and from the
battery ("SW1 down", the zero-standby-power
mode). The "SW1 up" mode also addresses the
issue of the initial charging of the battery.
Note the use of the micro-power LM2936 as a
5V regulator. Typically used LM7805 would by
itself consume 5 times more power than the
entire circuit and would prevent the entire
system to become classified as micro-power.
Using LM7805 would make battery discharge
cycles 5 times deeper and hence requiring 5
times larger capacity battery for sustained
operation, not to mention a bigger transformer to
keep it charged. It is interesting to note, that the
system in Fig 1 will work correctly even if
LM2936 is removed and bypassed (pin1 pad
connected to pin3 pad). This is due to the fact
that all key voltages in the system are
proportional to the LM324 supply voltage.
In "SW1 down" mode, the battery is charged
when the freezer relay and hence the compressor
are on, which, for my Vestfrost fridge, is
between 1 and 2 minutes per hour. The rest of
the time, the thermistor circuitry is powered up
by the battery, so that it does not draw any
current from the 240V supply.
During the system operation, the nominal 8.4V
NiMh battery voltage varies between 9.2V and
9.4V, so that in practical terms the battery
remains fully charged and hence can operate for
many years.
Fig 2. The printed circuit board of the thermostat
is double-sided and has insulating solder masks
to maximize the circuit safety and reliability. The
240V part contains only 2 pins and they are well
insulated from the rest of the circuit.
When choosing the transformer, we need to be
aware of its magnetizing current specifications
and choose the one with the minimum
magnetizing current, if possible. In my design I
used a cheap 2VA transformer with built-in
thermal fuse and the magnetizing current
<20mA. Since the battery charger section
(transformer TR1 and LM317 regulator) only
work 1-2 minutes per hour, their optimization
was not attempted.
Fig 3. Inside the thermostat. The thermostat
enclosure is waterproof. Top: transformer
surrounded by soldered and double-insulated
wires. Bottom: the 9V NiMh battery. Cables are
intentionally left longer so that the assembled
pcb can be removed for inspection without
disconnecting it from any part of the circuit.
Installation
The thermostat system described above is
designed to be installed on a power extension
cable that delivers power to the freezer. No
freezer modification is needed. The well-sealed
thermistor, soldered at the end of a thin cable of
sufficient length, needs to be inserted into the
freezer interior. This is best achieved using a
freezer draining hole.
Temperature measurement
I deliberately omitted temperature measuring and
display from my design in order to keep the
design as simple as possible (one IC design).
What helped me in this decision was an
abundance of the fancy temperature measuring
gadgets available on the market. Personally I use
the "dual thermometer" from Jaycar for less than
$20 with its temperature sensor well sealed with
silicone. It measures 2 temperatures: one inside
the fridge and one in the room outside it.
The weak point
The design above has one weak point. When the
240V power is not available and the fridge
interior temperature rises, the 60mAh battery
will power the relay coil up for about 4 hours
and then will go flat. This time can be extended
by using a larger capacity battery.
I have doubts if this issue requires attention,
because my view is that if the power goes down
for many hours, the content of any fridge, no
matter how advanced, will need a very careful
inspection and manual intervention. When the
power is restored, my system will require
switching to the "SW1 up" mode for a day or so,
so that the battery becomes fully recharged.
Fig 4. The thermostat at work. Bottom wires are
240V in-out wires entering the enclosure via
rubber grommets. The end of the top wire
connects to thermistor R1. Switch SW1 is
equipped with a waterproof boot. The
potentiometer shaft end is accessible for easy
temperature adjustment. A bit of a good tape or a
drop of silicone can be used to cover the
potentiometer shaft if the full waterproofing is
required.
Living with the chest fridge
My Vestfrost freezer-turned-fridge with the
thermostat described in this article maintains the
pre-set fridge interior temperature with the
accuracy of approximately =B11.5=BAC. The minimal
fluctuations in temperature help to minimize the
food spoilage. For example, one of my test jars
contained some yogurt that was still edible 6
months after its expiry date. Have you ever tried
to store half a jar of yogurt in a fridge for 6
months?
Please contact the author (tom@mtbest.net) for
prices and availability of the kit in your part of
the world.
Other published designs
The thermostat design published in the Silicon
Chip magazine (June 2005, Australia) has been
designed to draw about 60Wh per day
continuously from a 240V source. This power
draw comes from the 240V "plugpack" and the
triac relay. Implementing this thermostat to a
Vestfrost freezer-to-fridge conversion, would
cause my 0.1 kWh/day fridge power
consumption rise by about 60%. People who
relay on Solar Power would incur another waste
of 200-400Wh per day, because the mediocre
thermostat would prevent their inverter from
ever entering the standby (sleep) mode. In their
case, the amount of wasted power would be
equivalent of running at least TWO (and up to 4)
additional Vestfrost fridges equipped with my
thermostat. Surely, they would enjoy
experiencing the difference, wouldn't they?
-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist