Note: Descriptions are shown in the official language in which they were submitted.
~ ~633~7
35-EL-1441
- 1 -
FAIL-SAFE ELECTRONICALLY
CONTROLLED DEFROST SYSTEM
=
BAC~GROUND OF THE INVENTION
1. Field of the Invention:
The invention relates to refrigeration
systems, and more particularly to frost-free refrigeration
systems for home use in which an automatically con-
trolled defrost hea-ter-is provided incorporating a
defrost heater.
2. Description of the Prior Art:
Previously~ the defrost cycle of a frost-free
refrigerator was controlled by a timing motor connected
directly across the compre~sor motor. Thus, whenever
~he compressor motor ran, the defrost motor ran, and
after about eight (8) hours of compressor run-time,
the de~rost timer reached the defrost mode. The
compressor was shut off, while maintaining power to
the defrost timer. At this time the defrost heaters
were energized as well as a condenser fan, and the
frost build-up was melted and collected in the evaporating
pan in the machinery compartment to be evaporated. The
defrost heaters use a lot of energy, and this method
of defrost cycle control is sub-optimal in situations
where the doors of the refrigerator are not opened for
extended periods of time, e.g., when the owner is on
vacation or at night. Under these circumstances, no
~ ~3347
35-EL-1~41
2 --
moisture enters the refrigerator, and the length of
time between defrosts may be extended without the fear
of excessive frost build-up. With the advent of
microprocessor controlled appliances, it is possible
to detect these situations and lengthen the defrost
cycle appropriately. A shortcoming of the electronic
control system was that the de~rost system was not
protected from catastrophic failure of the electronics;
if the electronics failed in such a way as to disable
power to the defrost timer, the defrost system would
not cycle Two modes of defrost failure can be identified.
First, it is possible for the system to fail with the
defrost heaters off, in which case a frost build-up
would occur restricting the air flow between the freezer
compartment and the fresh food compartment, causing an
eventual overheating of the fresh food side. Even more
detrimental is the mode in which the system fails with
the defrost heaters on and the compressor disabled, in
which case the refrigerator rapidly overheats~ The
; 20 bi-metal thermostat in series with the defrost heaters
will interrupt the power, but the compressor will
remain disabled.
- 5UMMP.RY OF THE IMVENTION
Accordingly, it is an object of the pre~ent invention
to provide in a refrigeration system an improved
automatically controlled defrost system.
It is a further object of the present invention
to provide in a refrigeration system an automatically
controlled defrost system ha~ing reduced power
consumption.
It is another object of the present invention to
provide in a refrigeration system an electronically
controlled defrost system of high reliability.
It is another object of the present invention to
provide in a refrigeration system an electronically
controlled defrost system of high reliability and reduced
~ ~33~7
35-EL-1441
-- 3 --
power consumption.
These and other objects of the invention are
achieved in a self-defrosting refrigaration system
including an evaporator subject to frost build-up; an
electrically powered, motor driven compressor for
circulating coolant through the evaporator. These
elements are powered from a conventional alternating
current source under the control of a thermostat to
maintain the refrigerator temperature at a desired
value, a defrost timing switch setting a minimum
defrost period, a defrost delay relay, an electronic
decision element or microprocessor for computing a
suitable amount of extension of the defrost period
consistent with the circumstances and a fail-safe
coupling circuit installed between the decision element
and the defrost delay relay.
The defrost timing switch includes a timin~ motor
and a two condition switch permitting either compressor
operation or defrost heater operation when the thermostat
is closed as a function of the time that the timing
motor is energized. The defrost delay relay includes an
operating winding and a pair of normally closed contacts
which open when the operating winding is suit~bly
energized. The thermostat contacts, the timing motor
and the defrost delay relay contacts are serially connected
between the ac input terminals so that energization of the
timing motor only occurs when both the thermostat and the
defrost delay relay contacts are closed.
The electronic decision element is responsive to
input information supplied by the user or other sensors
for determining the extension of the defrost period
beyond the minimum period set by the defrost timing switch.
It produces an output signal having a duration dependent
on the input information and equal to the computed delay.
The decision element is coupled by the fail-safe coupling
circuit to the relay operating winding for opening the
334 ~
35-EL-1~41
- 4 ~
relay contacts to achieve the desired extension in the
: period between defrostings. In a preferred form, the
electronic decision element is a microprocessor (integrated
circuit) producing a sequence of output pulses,
characteristic of the decision element, whose presence
is generally indicative of proper operation of the
- decision element.
The ac source is of a conventional low frequency
(50-60 hertz, while the output signal of the decision
element is at a substantially higher frequency (1 kHz).
The fail-safe coupling means is then designed to provide
frequency dependent coupling to provide response to the
periodic pulses produced by the decision element and
non-response to direct current or low frequency
alternating current quantities at the same coupling
interface.
The fail-safe coupling circuit includes an amplifier
responding to the pulse sequence from the decision
element and a pulse detection circuit, to which the
amplifier output is supplied, and will operate the
defrost delay relay when amplified pulses are supplied
thereto.
In a preferred from, the amplifier is a pulse
ampli~ier having a predetermined input threshold and
a binary OlltpUt characterized by a first and second
output state, dependent on whether or not the input
signal exceeds the threshold. Consistently, the decision
element produces a sequence of periodic pulses whose
amplitudes exceed the threshold of the pulse amplifier.
The pulse detection circuit comprises a capacitor
having a first terminal coupled to the output of the
pulse amplifier and two rectifiers. The rectifiers are
serially connected in like polarity across the operating
winding. The second terminal of the capacitor is connected
to the rectifier interconnection. The rectifiers are
connected in a sense to charge the capacitor through
: 1 ~63347
35-EL-1441
-- 5 --
one rectifier when the pulse amplifier output is in a
first output state and to discharge the capacitor through
the other rectifier and the operating winding when the
pulse amplifier output is in the second output state to
operate the defrost delay relay. The capacitor value
is selected to provide sensitive relay operation at the
frequency of the microprocessor pulse sequence while
being insensitive to periodic waveforms at the fundamental
or ripple frequencies of the ac power source.
In a preferred form, the pulse amplifier includes
two transistor amplifier, one being on when the other
is off under the control of the decision element and
effectively interconnecting the first terminal of the
capacitor to either the positive or the negative terminal
of the dc bias source. Thus, when the "first" amplifier is
off and the "second" amplifier is on, a charging voltage
is applied from the positive terminal of the bias supply
to the capacitor. Similarly, when the "first'l amplifier is
on and the "second" amplifier is off, the capacitor is
discharged to the negative terminal of the bias supply
in a path including the relay winding and which energizes
the r~lay.
BRIEF DESCRIPTION OF TE~E DR~WINGS
The novel and distinctive features of the invention
are set forth in the claims appended to the present
application. The invention itself, however, together
with further objects and advantages thereof may be best
understood by reference to the following description
and accompanying drawings in which:
Figure 1 is a simplified block diagram of a
household refrigeration defrost control system having
special "vaction mode" setting for reducing the
frequency of defrosting and conserving power when the
owners intend to be away;
Figure 2 is an electrical circuit diagram illustrating
`" f ~S3~ ~
35-EL~1441
-- 6 --
a portion of a refrigerator control system in which a
microprocessor is coupled to a defrost delay relay through
a novel fail-safe coupling circuit;
Figure 3 illustrates a portion of a refrigerator
control system similar to that in Figure 2, with an
alternative form of fail-safe coupling circuit;
Figure 4 is a simplified block diagram showing
the interconnections between the microprocessox, a
control console on the refrigerator and the fail-sae
coupling circuit; and
Figure 5 is a simplifier block diagram of a
microprocessor suitable for use in a refrigerator
control system.
DESCRIPTION OF T~E PREFERRED EMBODIMENT
In Figure 1, a simplified electrical block diagram
of an automatically controlled defrost system for a
refrigeratox is shown. The system includes a
mi~roprocessor to extend the interval between defrostings
dependent on the circumstances for optimum power con-
servation. In addition, the microprocessor is
connec~:ed into the defrost control circuit through a
fail-safe coupling circuit in such a manner that
failure of the microprocessor or fail-safe coupling
circuit provides only a loss of the power conservation
feature and does not significantly affect the reliability
of defrost operation in the non power conservation mode.
The conventional components of the electrical
block diagram include the plug 11, which is used to
connect the refrigerator into a conventional 120 volt,
60 hertz appliance grounding receptacle (a part of the
house wiring), an electrically powered compressor
m~tor 12 for circulating coolant through an evaporator
in order to cool the refrigerator, a thermostat 13
designed to turn on the compressor motor when the
refri~erator temperature is too high and turn it off
when the refrigerator temperature is too low, a defrost
3 3~ 7
35-EL-1441
-- 7 --
heater 14 for removing frost from the evaporator which
builds up during the cooling cycle, a defrost timer 15
designed to operate the compressor for one interval
and the defrost heater for another interval, and a
defrost thermostat 16 designed to turn off the defrost
heater when the frost is removed from the evaporator.
In a conventional refrigerator, the foregoing
elements 1-16 (not including the deErost delay relay 17),
comprise a conventional highly reliable defrost systemv
The5e elements (1-16) are interconnected in the following
manner. The plug 11 includes a pin or terminal Tl for
connection to the "hot" connection in the house wiring
receptacle, a pin or terminal T2 for connection to the
neutral connection in the receptacle and a pin or
terminal T3 for connection to the ground connection in the
receptacleO The plug terminal Tl is connected to the
"hot'7 bus 18 of the refrigerator power wiring, the plug
terminal T2 is connected to the neutral ~us 19 of the
~' refrigerator power wiring and the plug terminal T3 is
connected to the ~ Ea~ of the refrigerator for gounding
it to the ground in the wiring receptacle. The
thermostat 13 is a conventional vapor filled bellows
activated device having a pair of contacts Sl and S2
which close when the refrigerator temperature is above
a desired value. The thermostat contact Sl is connected
- to the "hot" electrical bus 1~ for application of power
to the circuit and the thermostat contact S2 is connected
to the defrost timer, which in turn controls the
compressor 12 and defrost heater 14, which are the loads
of the circuit. The defrost timer 15 consists of a timer
motor having terminals Ml and M2 and a signal pole, double
throw switch having a pole contact S3 and stationary
contacts S4,S5. The timer pole contact S3 and the timer
motor terminal Ml are both connected to the thermostat
contact S2. The other terminal of the timer motor (M2)
is connected through the contacts (S9, S10~ of the defrost
1 1~33~7
35-EL-14~1
-- 8 --
delay relay 17 and~ which for purposes of the present
discussion are assumed to be closed, to the neutral
refrigerator bus 1~. The first stationary contact S4
of the defrost relay is connected to one terminal (M3)
of the compressor motor 12, while the other motor
terminal (M4) is returned to the neutral bus 19. The
stationary contact S5 of the defrost relay is connected
to one terminal of the defrost heater 14. The defrost
thermostat is conventional, electrically equivalent to
a single pole, single throw switch as illustrated.
The contact S6, which contacts S7 under "cold"
conditions, is connected to the other terminal of the
defrost heater 14 and the other-contact S7 is connected
to the neutral refrigerator bus 19. This mode of
connection allows the heater to be turned off earlier
than allowed by the timer once the frost has been
removed from the evaporator.
The foregoing elements of the defrost system are
set conservatively to insure against frost accummulation.
Typically, the timer is set so that the compressor will
run between five and eight hours, a long enough interval
to insure that under the most humid conditions and maximum
door openings that the accumulated frost on the re~rigerator
in question will not be greater than the defrost heater
: 25 can remove in one heating period. The defrost heater is
set to operate for a fixed maximum period, typically
thirty minutes. by means of the timer, as noted above.
The period is often shortened by the deforst thermostat
which senses when the evaporator coils are frost free
and turns off the heater. If the heater is turned off
after perhaps 25 minutes, the timer then continues in
the defrost position for the 5 minute balance of the
period allocated for frost removal. At the end of the
defrosting period, the de.~rost timer than allows the
compressor to come back on and reinstitute the cooling cycle.
~ ~63347
35--EL--1441
_ g _
The power conservation components of the electrical
block diagram oE Figure 1 include the defrost
delay relay 17, the fail-æafe coupling circuit 20,
which is depicted in detail in Figure 2, the microprocessor
21, the lnput shift register 22, and the switches 23, 24
and 25. These components delay the defrost period for
power conservation without significantly decreasing
the reliability of the overall defrost system.
The power conservation elements of the block diagram
are interconnected as follows: The defrost delay relay
17 is a device which consists of a pair of normally
closed contacts 59 and S10 and a dc operating winding
having terminals Wl and W2 which opens the contacts
when suitable energization is supplied. Energization for
the relay is supplied by the coupling circuit 20, which
makes connection to the terminals Wl and W2 of the relay
winding. The fail-safe coupling circuit 20 responds
to a control signal ~rom the microprocessor 21. (The
power conn~ctions to the blocks 20 and 21 are not fully
shown in Figure 1.)
The microprocessor 21, which may be used to control
several refrigerator functions, responds to several inputs.
The first input is a connection to the "hot" refrigerator
bus 18, which provides ~ 60 hertz signal indicative of
the passage of time that the refrigerator has been
connected to the ac source. Other inputs are provided via
the input shi:Et register 22. They include a switch 23,
which sets the microprocessor into a vacation mode (on
condition, a switch 24 which sets the microprocessor into
a vacation mode (off) condition, and a door switch, which
responds to a door opening to produce a vacation mode
(off) condition. The operation of these elements will
be taken up in greater detail in which follows.
One embodiment of the fail-safe coupling circuit
20, including the microprocessor 21 and the defrost
delay relay 17, ~e illustrated in Figure 2. The
~ 1633~7
35-EL-1441
-- 10 --
coupling circuit 20 is of maximum reliability with a
minimum components count. In the vacation (off) mode
achieved by operating the vacation mode (off) switch
25 or opening the refrigerator door, which actuates
the door switch 23, the input shift register 22
provides a control input to the microprocessor 21,
which causes it to produce no output to the coupling
circuit. In the event that the vacation mode (on)
switch 24 is operated, the input shift register 22
provides a control input to the microprocessor which
^~ causes it to produce ~e~ output for a fixst period of
typically 20 minutes followed by a typically 30 minute
period in which 1 kHz rectangular pulses of about 1.2
volts amplitude are produced, followed by another 20
minute period in which the microprocessor output is
zero, followed by a second 30 minute period during which
the pulses occur. The cycle rep-r1ea-t~ until the defrost
timer operates to turn off the compressor. ~The pulses
in the microprocessor output are ~plifi~ by the
coupling circuit 20 and applied to the operating winding
of the defrost delay relay 17. In the indicated
sequence, the relay contacts are held open for 30
minutes sequence. The timer motor whose terminals
~Ml, M2) are serially connected with the defrost delay
relay contacts is thus prevented from advancing for
30 minutes in each 50 minute sequence. As a consequence,
the defrost cycle which would have provided 5 hours of
compressor time between defrostings in the normal mode,
is now extended to 12-1/2 hours in the vacation mode.
The operation of the fail-safe coupling circuit
20, which amplifiers the pulse output of the micro-
processor to a suitable level to operate the relay 17,
and which includes a suitable pulse dete~tion circuit
for dc relay operation, may now be explained with
reference to Figure 2. The coupling circuit may be
regarded as consisting of an included rectifier-
3 ~ ~
35-EL-1441
-- 11 --
capacitor power supply (30, 31) for providing a filtered
low voltage (e.g. 24 volts in Figure 2), a pulse amplifier
consisting of the three transistors 32, 33, 35, diode 34 and
load resistances 36, 37, and a pulse detection circuit
consisting of capacitor 38 and rectifiers 39, 40. The
pulse amplifier has an input threshold below which no
output is produced and above which full output is
produced. The input -threshold is typically about 1.2
volts and the output is slightly less than the dc supply
voltage.
The dc power supply 30, 31 of the coupling circuit
is designed to operate from 24 volts of unfiltered dc
available at terminal 29, and which i5 normally obtained
from a separate step-down transformer from the 120~ 60
cycle main. The anode of the rectifier 30 is coupled to
terminal 29~ The cathode of rectifier 30, which becomes the
positive terminal of the internal dc supply, is coupled to
one terminal of the filter capacitor 31. The other
terminal of the capacitor 31 is connected to a ground
connection common to the 24 volt transfcrmer secondary
and the coupling circuit wiring. The filter capacitor 31
is typically 20 microfarads and prQvides a substantial
~t of energy storage.
The pulse amplifier of the fail-safe coupling
circuit 20, which includes the three transistors 32, 33
and 35, has an input terminal corresponding to the base
electrode of the transistor 32 to which the 1 k~z output
of the microprocessor is connected, a common or ground
terminal, and an output terminal 41 from which an amplified
1 kHz pulse is derived for coupling to the pulse detection
circuit (38, 39, 40).
The elemen-ts of the pulse amplifier are interconnected
as follows. The transistor 32 has its emitter connected
to the base of -the transistor 33 and its collector
connected through a first load resistance 36 to the
positive terminal o~ the internal dc supply 30, 31. The
`~ ~633~7
35-EL-1441
- 12 -
collector of transistor 32 is also connected to the anode
o~ diode 34 whose cathode is connected to the base of
transistor 35. The emitter of transistor 33 i9 connected
to the ground connection. rrhe collector of transistor
33 is connected to the emitter of transistor 35. The
collector o.f transistor 35 i5 connected through a second
load resistance 37 to the positi~e terminal of the
internal dc supply 30, 31. The output of the pulse amplifier
appears at the interconnection 41 between the collector
of the transistor 33 and the emitter of transistor 35.
The foregoing circuit ~32-37) amplifiers pulses from
the microprocessor. Assuming that microprocessor 21 is in
a first, quiescent condition (the zero state), and has an
output at or near zero, the input junctions of transistors
32 and 33 are at near zero bias and both transistors are
off. At the same time, the input junction of transistor
35, whose emitter may be assumed to be coupled to ground
through an as yet not fully described load, and whose base
i8 returned to B+ through diode .34 and resistance 36,
becomes forward biased, turning on transistor 35 and
coupl~.ng the nearly full dc voltage of the internal dc
supply to the amplifier output terminal (41). In the
second, active condition corresponding to the one state
of the microprocessor, a pulse exceeding the 1.2 volts
input threshold of the amplifier is applied. The
microprocessor output is sufficient to forward bias the
serially connected input junctions of transistors 32 and
33, causing both:transistors to conduct. Conduction by
transistor 32 reduces the bias applied to the base of
transistor 35 toward ground potential turning transistor
35 offO Conduction by transistor 33 forces the potential
at output terminal 41 to fall to near ground potential.
Considering both states dynamically, when a 1 kHz pulse
is applied from the microprocessor to the pulse amplifier
at a value exceeding approximately 1.2 volts, an output
pulse of the same frequency at slightly under 24 volts will
3~ Y
35-EL-1441
- 13 -
appear at the output terminal 41 for coupling throuyh the
detection circuit to the operating winding of relay 17.
The pulse detection circuit 38, 39, 40 couples the
output energy from the pulse amplifier to the operating
winding of the relay in an essentially dc form. The
input terminal of the pulse detection circuit is a first
terminal of the 2.2 microfarad capacitor 38. The second
terminal of the capacitor 38 is connected to the cathode
of diode 39 and the anode of diode 40. The anode of diode
39 is connected to the Wl terminal of the relay actuating
winding and the cathode of the diode 40 is connected to the
grounded terminal W2 of the relay actuating winding.
The detection circuit functions in the following
manner. Assuming that the microprocessor is in a one state
in a pulse train and that the transistor 35 is conductive,
the capacitor 38 receives a positive charge of approximately
24 volts on the electrode connected to the terminal 41.
The charging path includes in succession the diode 30, the
56 ohm load resistance 37, transistor 35, capacitor 38
and diode 40. When the capacikor 38 is charged, which
normally occurs before the "1" state of the pulse is ended,
the transistor 35 becomes less ~-conductive. When the
"O" state of the pulse occurs, transistor 35 is turned
off strongly by conduction of transistors 32, 33. At
output terminal 41, the transistor 33 provides a
conductive path to ground. This allows the positive
charge on the capacitor 38 to begin discharging through
a path which includes from ground, the relay actuating
winding, diode 39, capacitor 38, and the transistor 33
(collector and emitter) to ground to complete the circuit.
The rate of capacitor discharge is limited by the
inductance of the relay winding. When the next "1" state
of the microprocessor occurs, the capacitor 38 is re-
charged through the charging path. During the next "O"
state, the capacitor is again discharged thrGugh the
relay winding. The amount of energy coupled into the
I lB33~
35-EL-1441
- 14 -
capacitor from the pulse amplifier for each conduction
period is fixed by the capacitor 38 and B~. The amount
of energy coupler per conduction period into the detection
circuit by the capacitor tends to be independent of
frequency so long as the duration of the on time of the
pulse exceeds about 5 times the time constant of the
charging circuit.
During the period that the pulse amplifier is in a
zero state in a pulse train, the capacitor 38 discharges
through a circuit which includes the inductance of the
relay operating winding. The circuik is designed for
roughly optimum energizing current in the relay. With a
suitably optimized load, the pulse detection circuit then
becmomes frequency dependent. In other words, the
average current equals the charge per conduction pulse
times thè pulse repetition rate. With a constant charge
per pulse the average current becomes directly proportional
to the pulse repetition rate. Practically, the relay will
usually operate for pulse at a repetition rate somewhat
below 1000 per second, e.g., 500 Hz but will not operate
for unfiltered ripple waveforms such as the 50 Hz or-120
Hz corresponding to the ripple frequencies that might
occur with filter capacitor failure. Pulses at these
ripple frequencies are incapable of coupling sufficient
energy to the capacitor to actuate the relay. Meanwhile,
the capacitive coupling prevents the relay from being
sensitive to dc quantities.
In the foregoing Figure 2 embodiment, it is highly
unlikely that the electronics can open the delay relay
unless there are no electronic failures. If the
microprocessor or its supporting power supply were to fail,
the square wave would not be present and no current would
flow through the relay. If transistors 32, 33, 35 or diode
34 were to fail, the square wave would not appear at the
capacitor 38 and no current would flow. For the system to
fail, would require rectifier 39 to fail as a short circuit,
33~1~
35-EL-1441
- 15 -
rectifier 40 to fail as an open circuit, capacitor 38
to fail as a short circuit and transistor 35 to be in
saturation or shorted. The ~24 VDC supply must also be
operational at the time. Under any other combination
of failures other than the one described, delay relay
17 will not open and the defrost system will remain
intact.
The fail-safe coupling circuit is also of a
conservation electrical design. The in~ut threshold
which establishes which of two output states the pulse
amplifier is in, is established by the input junction
drops of the two transistors 32 and 33, and is of high
reliability. Since the pulse amplifier produces a binary
output in which there is a high state and low state, any
gradual deterioration in the transistorgain will not
produce circuit failure since there is a very substantial
excess gain. The remaining capacitors, resistors and diodes are
are all of high reliability, and when selected with con-
servative ratings, they provide a high reliability
configuration comparable to the 20 year design life of the
mechanical refirgerator components.
~ second embodiment of the fail-safe coupling
circuit is illustrated in Figure 3. When similar
components are recited in Figure 3, primed reference
numerals have been employed. The Figure 3 embodiment
is in a low cost commercial form employing a commercially
available Darlington array and includes a capacitor in
parallel with the defrost delay relay operating winding
to reduce relay chatter and a resistance in the discharge
path of the coupling capacitor to preclude excessive
currents during discharge. In the pulse amplifier/ the
Darlington elements of the array, which bear the reference
numerals 51 to 55, and a separate Darlington 57 are the
active elements. The input of the first Darlington 51 is
coupled to the output terminal of the microprocessor 21 and
the output thereof (an inverting output) is coupled to
~ ~L633~7
35-EL-1441
- 16 -
the four inputs of the Darlingtons 52 to 55. The output
of 51 is coupled through resistance 58 to the B~ bus 59.
Three of the Darlinyton, 52, 53 and 54, have their
inputs and their outputs paralleled, the paralled output
being serially connected through the 12 ohm current limiting
reistance 56 to the output terminal 41' of the pulse
amplifier. The Darlington 55 has its output coupled to
the base of a separate Darlington 57, the base electrode
of which is coupled through load resistance 60 to a B+
bus 59. The output collector of the Darlington 57 is
coupled through a 12 ohm resistance 61 to the B+ bus 59.
The emitter of the Darlington 57 is coupled to the output
terminal 41'. The pulse detection circuit is similar
to that shown in Figure 2 except for the addition of the
capacitor ~ .
The foregoing pulse amplifier and pulse detection
circuits function in much the same manner as their
counterparts in the first embodiment. When the output
of the microprocessor 21 is in a high or "1" state,
the output of the first Darlington 51 is low and the
output of the second rank of Darlington (52 through 55)
is high. Noting that the Darlingtons 52, 53 and 54 are
internally grounded, and have collectors which are
unconnected to B+, this state produces an open circuit
condition between resistance 56 and the internal ground
of the array. Simultaneously, the Darlington 55 has
an open conne~tion between the output circuit and ground.
This allows the separate Darlington 57 to be turned on by
current flowing through resistance 60. In the microprocessor
one state, therefor, the voltage at output terminal (41')
of the pulse amplifier approaches the B+ potential (12
~olts~. This allows the capacitor 38' to be charged
through the diode 40' in the same manner as in the first
embodiment.
~ssuming that the output waveform of the microprocessor
l 1633~7
35-EL~1441
- 17 -
is a pulse sequence, the next interval produces a "0"
state. With a "O" state at the input of stage 51, a
"1" output is applied to the input of the second rank
of Darlington amplifiers 52 and 55. This causes strong
conduction through the Darlingtons 52, 53, 54 which
discharges the capacitor 38' through a path including
the resistance 56, the three Darlington 52-5~ and ground.
The balance of the path includes the rectifier 39' with
a significant portion of the initial current source being
conducted through the capacitor 62 in shunt with the
winding of relay 17'. At the same time that the three
Darlingtons 52 to 54 are connecting the output point
41' to ground, the Darlington 55 is also connecting.
Darlington 55 connects the base of the separate Darlington
57 to ~round, turning it off.
Thus, the pulse amplifier in the second embodiment
alternately connects the output terminal 41' through-a low
impedance to B~ and through a high impedance to ground
in the capacitor charging state (when the microprocessor
is in the "1" state) and to B+ through a high impedance
and to ground with a low impedance during the period that
current is being supplied to the relay (when the micro-
processor is in the "O" state). The operation of the
Figure 3 emb~diment is accordingly very similar to that
of the first embodiment but has the advantage of
permitting lower costs in the context of other control
requirements in the refrigeration system.
Returning to system operation, when the relay
energizing pulses occur at a sufficiently high rate, e.g.
1,000 pulses per second, sufficient current flows to
energize the defrost delay relay 17' and it is held in an
open position which prevents the defrost timer 15 from
advancing. The defrost timer, which is under control of
the microprocessor, is designed to be on for a first
specified period (20 minutes) and off for a second
specified period (30 minutes) so as to defer the period
~ ~33~
- 18 35-EL-1441
oE the next defrost, as previousl~ described, when in the
vacation mode.
The mlcroprocessor is designed to produce a signal
which indicates with a high degree of probability that it
is functioning properly. While that signal could take a
number of different forms, the 1,000 pulse per second
rectangular format has that basic property, and is initiated
when a suitable input is provided by operation of the
vacation mode on switch.
The refrigerator in the present arrangement is provided
with a main control console 69 as illustrated in Figure 4,
which is mounted upon the refrigerator door. The control
console is a unit which accepts information from the user
and also displays information to the user. Information is
accepted by means of input switches and information is
supplied to the user by means of output displays. As illus-
trated in Figure 4, the microprocessor 21 and the main con-
trol console are linked by a plurality of connections includ-
ing a main bidirectional data bus and buses for the display
board and data shift clock and direction data. The arrangement
permits the cycling of information from the input switches on
the main console to the microprocessor and from the micropro-
cessor to the displays on the main control console under micro-
processor control. More particularly, data from the input
switches are "strobed" into a parallel to serial shift
register (22) located in the control console and that data
is then read serially into the microprocessor. The parallel
to serial conversion minimizes the number of actual con-
nections between the microprocessor and the control console.
Two of the bits in the latter data stream are from the
vacation mode "ON" and vacation mode "OFF" switches which
are manually pressed by the user. The microprocessor
interprets the status of these bits and performs a time
integration function requiring switch closure for several
successive cycles before recognizing that the vacation
mode switch is "ON". This process is referred to as a
. . .
33~ ~
35-EL-1441
- 19 -
"debounce" function.
Once the vacation mode "ON" condition has been
recognized, the microprocessor is programmed to duty
cycle modulate the defrost timer by alternately
opening and closing relay 17 as earlier described. It
accomplishes this by applying a 1 kHz square wave to the
fail-sa~e coupling circuit (2) when the relay is to be
open and removing the signal when the relay is to be
closed. The generation of the lkHz square wave may best
10- be understood by reference to Figure 5 which is an
illustration in block diagram form of a typical micro-
processor. The MK 3870 is a commercially available example
of such a microprocessor.
The microprocessor consists of 13 blocks (71-83)
generally linked by a main data bus (70) and connected
to the control console through N input-output ports (71).
One o~ these input-output ports is connected to the input
shift register 22 ~ithin the control console. The
microprocessor includes an arithmetic logic unit (72)
and an accumulator and status register (73). The micro-
processor employs a ROM address register (74) recirculating
information through an adder (75) whose other input is
coupled to the main data bus 70. The block 74 supplies
~n output via a program ROM (76) to the main data bus.
Also coupled to the main data bus is an indirect scratch-
pad address register (77) whose output is coupled to a
~cratchpad RAM (78), the latter recharging data with the
main data bus. The microprocessGr is under control of
the control logic (79) and instruction register (80).
Timining is achieved through a timer (81) and interrupt
logic (82). A crystal LC, or RC, clock is shown at 83.
In order to generate a square wave pulse at one of
the output ports, the program ROM 76 first outputs a logic
level (hi:gh or low) to the appropriate input output port.
The program ROM then leaves the port in that state for a
length o~ time T which can be accurately timed by the
1 1~334~
35-E~-14~1
- 20 -
internal timer (81). At the end of the "T" interval
the program ROM then forms the logical complement of that
bit, which is then outputs to the input-output port for
an additional time. If the process is repeated ad-
infinitum a square wave results at the input-output
port. For a 1 kHz square wave T is 500 ~eresees~.
The presence of a square wave is an indication
of high probability that the portion of the microprocessor
controlling the defrost delay relay 17 is operating
properly~ The most probable result of a catastrophic
component failure in this part of the circuit is that
the input-output port would become locked in one of two
permissible states. This is true of digital electronic
components in general; they generally fail either shorted
or open circuited. Because of the complexity of a
microprocessor, the two states generally assumable upon
failure are (to a first approximation) equally likely.
By requiring a square wave output, the system is able to
react to a failure in either mode.
Figure 5 is representative of microprocessors in
generaIly. All of the functional blocks are used in
generating the high frequency square wave with the
principal exception o~ the ~nused input-output ports.
The blocks used in generating the square wave are also
used in performing all of the other functions of the
refrigerator logic such as the calender indication,
door open alarams, etc. Since this is a synchronous
system, a catastrophic failure will most probably cause
the machine to "stop" in some state, and hence a square
wave will not be produced. For more subtle failures such
as a bad ROM, or a scratchpad failuret it may be possible
for the machine to generate a square wave of the correct
frequency, but for such to happen, multiple failures of a
non-catastrophic variety are required, a situation which
has a low probability of occurrence. The only single
I t $33~ 7 35-EL-1441
- 21 -
failure likely to cause this condition is the failure
of a particular bit of the scratchpad register (that
bit which tells the mic.roprocessor to generate the
square wave), and even in this case, additional soft-
ware can be used to detect and ignore this co.ndition.
The program which generates the square wave isresident in the program ROM (76). The status of the
square wave is kept in the scratchpad RAM (78), and the
timer (81) and interrupt logic (82) are used to generate
the time base T. In addition, blocks (74, 75, 77, 79, 80,
83) are used in the transfer and manipulation of data
and program instructions.