 |
Description  |
|
|
FIELD OF THE INVENTION
The present invention generally relates to the field of remote
condition-sensing equipment, for example, two-wire remote
temperature-sensing transmitters. More particularity, the present
invention is directed toward the problem of detecting a broken wire in a
three- or four-wire resistance temperature device (RTD) temperature
measuring unit.
BACKGROUND OF THE INVENTION
Two-wire transmitters are commonly used to monitor various conditions at
remote locations. For example, to measure the liquid level in a tank at a
remote processing plant from its central control room, a two-wire
transmitter at the remote location is typically connected in series with a
power supply and a load at a central location through two transmission
wires. As the condition being monitored by the transmitter varies, the
effective series resistance across the transmitter also varies so as to
produce a corresponding change in the current drawn by the transmitter. An
industry standard has developed in a large number of applications, wherein
the current through the two-wire transmitter loop varies from 4-20
milliamperes (mA), wherein 4 mA is the minimum amount of current required
to power the remote transmitter.
Volume, pressure, liquid level, and temperature are just some of the
conditions which are typically monitored using two-wire transmitters.
Temperature, however, is one of the conditions which often must be
measured with precision. It is well known to utilize a resistance
temperature device (RTD) for this purpose. The RTD is typically immersed
in the medium, the temperature of which is to be measured, such that the
resistance of the RTD will vary with the temperature changes of the
medium. Utilizing either a table of resistance-temperature values or a
polynomial equation to represent the relationship between the RTD's
resistance and temperature, the actual temperature is then calculated from
the measured resistance value of the RTD.
If the RTD is connected to the two-wire transmitter via two wire leads,
then the RTD resistance measurement would necessarily include the
resistance of the wire leads. For more accurate temperature measurements,
a four-wire RTD system is often employed, i.e., two wires from each
terminal of the RTD are connected to the two-wire transmitter. Two of the
wires are used to pass current through the RTD, and the other two wires
are used to sense the voltage developed across the RTD during the
measurement. In this manner, the RTD's resistance is measured without
passing current through the same wires that sense the voltages, i.e.,
without including the voltage drop of the lead wires. In still another
version of an RTD system, a three-wire RTD is used, wherein such
lead-length compensation is performed by measuring the voltage difference
between only one voltage sensing lead and the current return lead.
Numerous other RTD configurations are also possible, a few of which will
be described below.
A problem often occurs whenever one of the wires to the RTD breaks or has
an intermittent connection. Although a broken wire in the RTD's current
path wires would immediately be apparent at the two-wire transmitter as an
over-ranging, i.e., infinite, resistance measurement, a break in the
voltage sensing wires may only slightly affect the resistance measurement
by the amount of lead-length compensation being performed. In other words,
depending upon the condition sensor configuration and the particular lead
wire that is broken, a remote measurement system may appear to be
functional yet be providing inaccurate readings for quite some time before
the broken wire is discovered.
A need, therefore, exists for an improved remote measurement system which
addresses the problem of detecting a broken wire in a three- or four-wire
RTD temperature measuring unit.
OBJECTS AND SUMMARY OF THE INVENTION
Accordingly, it is a general object of the present invention to provide an
improved remote resistance measurement system which is particularly
adapted for determining whether a fault exists in the system.
Another object of the present invention is to provide a computerized remote
resistance measuring circuit having multiplexed inputs which can
individually isolate any of the connecting wires to the remote RTD
sensors.
A further object of the present invention is to provide a
microprocessor-controlled two-wire transmitter having the capability to
determine if an intermittent exists at the transmitter input terminals,
and indicate to the user precisely which terminal has the intermittent.
These and other objects are achieved by the present invention, which,
briefly described, is a sensing circuit for a condition sensor having at
least three sensor connection wires, the sensing circuit comprising: a
multiplexer circuit having at least three input terminals for connection
to the three sensor wires, at least one output port, and at least two
address lines; a current source for applying power to the condition
sensor; a first circuit for determining an electrical characteristic of
the condition sensor as measured at the multiplexer circuit output port;
and a second circuit for controlling the address lines, for determining if
any connection from the condition sensor via the three sensor wires is
defective and thereby providing a fault signal, and for providing an
indication in response to the fault signal. In the preferred embodiment,
the indication includes a message on a visual display informing the user
as to which input terminal has the faulty connection.
According to the preferred embodiment, a microprocessor-controlled remote
resistance measurement system is provided wherein the connection leads to
the three- or four-wire resistance temperature devices are multiplexed via
a four-channel analog multiplexer at the input of the unit. A separate
two-channel multiplexer is also used to multiplex a fifth input for a
reference resistor. The output of the multiplexer is coupled to a
voltage-to-frequency converter, wherein the frequency output is utilized
as an input to the microprocessor-based controller. The microcontroller
can check for broken wires by addressing the multiplexers to individually
isolate any of the connecting wires to the remote RTD sensors. The output
of the multiplexer is monitored in a Test Mode by connecting an known
impedance to the multiplexer output to determine if any RTD connections
are defective. If, for example, one of the voltage sensing wires is
broken, the known impedance will cause an erroneous frequency reading into
the microcontroller, which will then provide an indication on the display
for determining exactly which wire is broken. Only two address lines are
used to control the five multiplexer channels through the use of a
function selector circuit and a two-stage measurement cycle.
In addition to detecting faulty RTD connections, the multiplexer circuitry
of the present invention also provides the following advantages: (1) it
allows for independent measurements of a number of RTD sensors using a
single two-wire transmitter; (2) it provides the capability to measure the
value of one RTD independently from the value of another RTD, so that each
device can be separately linearized; and (3) it provides for more accurate
resistance calculations through the use of a non-grounded reference
resistor at the transmitter.
BRIEF DESCRIPTION OF THE DRAWINGS
The features of the present invention which are believed to be novel are
set forth with particularity in the appended claims. The invention itself,
however, together with further objects and advantages thereof, may best be
understood with reference to the following description when taken in
conjunction with the accompanying drawings, in which:
FIG. 1 is a general block diagram of the computerized remote resistance
measurement system of the present invention, wherein a four-wire RTD
configuration is shown;
FIG. 2a is a simplified schematic diagram of the multiplexing circuit of
FIG. 1, wherein three address lines are provided by the optoisolator;
FIG. 2b illustrates representative waveforms for the optoisolator address
lines to illustrate the operation of the circuit of FIG. 2a;
FIG. 3a is an alternate embodiment of the multiplexing circuitry of FIG.
2a, which has been modified to utilize only two address lines;
FIG. 3b shows representative waveforms at various points of FIG. 3a
illustrating the circuit's operation;
FIG. 4a is another simplified schematic diagram for the multiplexing
circuit of FIG. 1, wherein an additional multiplexed input is utilized to
more accurately determine the value of the reference resistor;
FIG. 4b illustrates representative addressing waveforms for the operation
of the multiplexing circuit of FIG. 4a;
FIG. 5 is a detailed schematic diagram of the preferred embodiment for the
multiplexing circuit of FIG. 1, wherein five multiplexed inputs are
controlled by only two address lines;
FIG. 6a is a schematic diagram illustrating the input circuit configuration
for the multiplexer of FIG. 5 when a two-wire RTD sensor is used without
lead-length compensation;
FIG. 6b is a waveform timing diagram for the operation of the circuit of
FIG. 5 when used with the input configuration of FIG. 6a;
FIG. 7a is a schematic diagram illustrating the input configuration for the
multiplexer of FIG. 5 when a four-wire RTD sensor is used;
FIG. 7b illustrates the timing waveforms for the four-wire sensor
configuration of FIG. 7a;
FIG. 8a is a schematic diagram illustrating the input configuration for the
multiplexer of FIG. 5 using a three-wire RTD sensor;
FIG. 8b illustrates the timing waveforms for the three-wire RTD sensor
configuration of FIG. 8a;
FIG. 9a is a schematic diagram illustrating the input configuration for the
multiplexer of FIG. 5 having a three-wire dual-RTD sensor configuration
using no lead-length compensation;
FIG. 9b illustrates the timing waveforms for the input configuration of
FIG. 9a;
FIG. 10a is a schematic diagram illustrating the input configuration for
the multiplexer of FIG. 5 having a five-wire dual-RTD sensor configuration
utilizing lead-length compensation;
FIG. 10b illustrates the timing waveforms for the five-wire dual-sensor
configuration of FIG. 10a;
FIG. 11a is a schematic diagram illustrating the input configuration for
the multiplexer of FIG. 5 having a three-RTD sensor configuration;
FIG. 11b illustrates the timing waveforms for the three-RTD input
configuration of FIG. 11a;
FIG. 12 is a flowchart illustrating the specific sequence of operations
performed by the microcontroller of FIG. 1 in accordance with the practice
of the preferred embodiment of the present invention; and
FIG. 13 is a flowchart illustrating the various interrupt operations
performed by the microcontroller in the preferred embodiment.
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 represents a general block diagram of the preferred embodiment of
the present invention. The computerized remote resistance measurement
system 10 of the present invention includes a two-wire transmitter 12,
typically at a remote location, connected in series with a power unit 14,
typically at a central location. The power unit 14 comprises a battery 15
connected in series with a load 16, both of which are connected in series
with a current modulator circuit 17 in the transmitter, thus completing
the two-wire current loop. The battery 15 is typically 24 volts DC, while
the resistance of the load 16 varies widely depending upon the
application.
In the preferred embodiment, the current modulator 17 is configured to use
the industry standard of 4-20 mA, however, the former standard of 10-50 mA
may also be used. On the other hand, if the transmitter unit 12 is
centrally located, then the load 16 and the current modulator 17 may be
omitted such that the transmitter is powered directly from the battery 15.
In either case, battery power is routed to a local power supply circuit
18, from which power is routed to an isolated power supply circuit 20. The
local supply 18 provides power to a microprocessor-based controller
(microcontroller) 22, its associated electrically-erasable programmable
read-only memory (EEPROM) 23, and the associated display circuitry, while
the isolated supply 20 provides power to the remaining transmitter
circuitry. The local/isolated supply arrangement will be described in more
detail in conjunction with FIG. 5.
The transmitter 12 has four input sensor connection terminals, X, A, B, and
C, which are connected to an RTD 24 via connections wires 25X, 25A, 25B,
and 25C, respectively, using a number of different sensor configurations
as will be seen below. One such input configuration is shown in FIG. 1,
wherein a single four-wire RTD 24 is connected to the transmitter 12. A
current controller circuit 26 powers the RTD 24 via the current-carrying
wires 25X and 25C connected to terminals X and C, while terminals A and B
are connected to measure the voltage across the RTD 24 via the respective
voltage sensing wires 25A and 25B. In the general block diagram of FIG. 1,
current would flow from the current controller 26, out of the transmitter
from terminal X, through wire 25X and the RTD 24, and would return via
wire 25C to terminal C back to the current controller 26. Although the
invention is adapted to use a wide variety of RTDs, i.e., having nominal
resistances ranging from 10 ohms to 4000 ohms, a typical RTD used with the
present invention is the widely available 100 ohm platinum bulb type RTD.
Moreover, many of the principles of the present invention may also be used
with other types of condition sensors which vary their capacitance,
inductance, or magnetic fields in accordance with temperature, position,
liquid level, dielectric constant, etc.
Each of the input terminals X, A, B, and C are also connected to the inputs
of a multiplexer circuit (MUX) 28 as shown. Since the voltage sensing
terminals A and B are connected to very high impedance inputs at the MUX,
practically no current flows into terminals A or B during the measurement
process. Hence, using the four-wire RTD configuration shown, the voltage
measured across terminals A and B is the precise product of the excitation
current flowing from terminals X to C, and the resistance of the RTD
sensor 24. Hence, the resistance of the current-carrying wires 25X and 25C
connected to terminals X and C do not enter into the resistance equation.
In this manner, lead-length compensation is inherently being performed
such that a much more accurate resistance measurement can be obtained.
Multiplexing control signals are provided by the microcontroller 22 to the
MUX 28 via an optoisolator 30. As will be explained in detail below, the
multiplexer circuit 28 provides the ability to individually select any one
of the RTD connection wires 25 in order to measure, at the MUX output
port, the voltage developed from each of the input terminals X, A, B, and
C to a ground reference point. The output voltage signal of the MUX 28 is
connected to a voltage-to-frequency (V-to-F) converter 32, which provides
frequency output data to the microcontroller 22 via another optoisolator
34. A typical voltage-to-frequency converter, which could be used as
V-to-F converter 32, is disclosed in the 1990 Linear Applications
Handbook, published by Linear Technology, in Application Note 14, pg. 9.
Basically, the microcontroller performs the functions of frequency-to-ohms
conversion, ohms-to-temperature conversion, and temperature-to-pulse-width
modulation (PWM) conversion to drive the current modulator 17. The
microcontroller 22 automatically switches between a Measurement Mode,
wherein the resistance of the RTD is calculated and temperature
information is provided to the user, and a Test Mode, wherein the
microcontroller 22 directs the MUX 28 to check if any of the wires 25 to
RTD 24 are broken. In the Measurement Mode, the MUX 28 provides voltage
information to the V-to-F converter 32, which, in turn, provides frequency
information to the microcontroller 22. This frequency information, as well
as information from user-accessible mode switches 38 and a set of
factory-programmed wire jumpers 40, is processed by the microcontroller 22
to provide the temperature information to the user via a visual display
36. The temperature information is also transmitted over the two-wire link
via the current modulator 17. In the Test Mode, the frequency information
provided to the microcontroller 22 is used in conjunction with the
addressing information provided by the microcontroller 22 to the MUX 28 to
check for broken wires and to inform the user of precisely which wire is
broken via the display 36. A communications test port 41 may also be
connected to the microcontroller 22 to allow for automated factory
calibration procedures. A Motorola 68HCO5 is used as microcontroller 22 in
the preferred embodiment. The EEPROM 23 is used to store calibration
information used in the system.
More specifically, in accordance with the operation of the Test Mode, the
microcontroller 22 directs the MUX 28 via optoisolator 30 to check whether
any one of the wires 25X, 25A, 25B, and 25C is broken, such that this
fault information is provided to the user via the display 36. As will be
seen below, if wire 25X from the RTD 24 to terminal X is broken, or if
wire 25C from the other terminal of the RTD 24 to terminal C is broken, no
current will flow through the RTD. Even in the Measurement Mode of
operation, this condition will immediately be noticed by the
microcontroller 22 when it receives either a zero frequency value or an
over-range frequency value from the optoisolator 34.
However, if one or more of the voltage sensing wires 25A or 25B from the
RTD 24 to terminals A or B is broken, current will still flow through the
RTD device, and a zero or over-range frequency value will not be detected
in the Measurement Mode. Moreover, since the input terminal A or B would
now be floating, one can not predict what the voltage level out of the MUX
will be. If it remains within the approximate range of appropriate voltage
sensing values, the V-to-F converter 32 may continue to output a nominal
frequency value which appears to be normal. Therefore, the present
invention implements the Test Mode to detect whether any of the RTD
connection wires, particularly the voltage sensing wires 25A and 25B, are
faulty by connecting a known impedance to the MUX output. If one of the
voltage sensing wires 25A or 25B is broken, the known impedance will load
down the output of the MUX such that either an over-range or zero
frequency value will be present at the input to the microcontroller. Since
the microcontroller 22 is also controlling the addressing to the MUX 28,
the microcontroller can determine exactly which wire is broken and display
this information to the user. The following description provides a
detailed explanation of both the MUX circuitry involved and the software
program followed to perform this fault analysis.
FIG. 2a is a simplified schematic diagram of one embodiment of a
multiplexing circuit 42 which serves as the MUX 28 of FIG. 1. The
multiplexing circuit 42 includes a four-channel analog multiplexer 44, a
current source 46, a reference resistor R.sub.REF 48, and a known
impedance 52 switchably connected to the output Z of the multiplexer 44
through a diode 50. In this embodiment, three address lines, OPTO.sub.1,
OPTO.sub.2, and OPTO.sub.3, are provided to the multiplexing circuit 42
from the microcontroller 22 via the optoisolator 30. As explained below,
the microcontroller 22 controls these address lines such that the
multiplexing circuit 42 connects any one of the four input terminals X, A,
B, or C to the multiplexer output port Z, with and without switching in
the known impedance 52.
As can be seen from FIG. 2a, the current controller 26 of FIG. 1 basically
comprises a constant current source 46 and a reference resistor 48,
labeled R.sub.REF. The reference resistor R.sub.REF is used to precisely
determine the value of the current flowing through the RTD sensor 24,
labeled R.sub.1. Since the same current flows through the reference
resistor 48 as through the RTD 24, the current I through the RTD sensor
R.sub.1 is:
I=V.sub.R1 /R.sub.1 =V.sub.REF /R.sub.REF (1)
where V.sub.R1 is the voltage drop across R.sub.1, and V.sub.REF is the
voltage drop across R.sub.REF. In the four-wire RTD configuration shown,
the voltage drop across R.sub.1 is equal to the voltage measured at
terminal A to ground minus the voltage measured at terminal B to ground,
i.e., V.sub.R1 =V.sub.A -V.sub.B. Similarly, the voltage drop across
R.sub.REF is equal to the voltage measured at terminal C to ground, i.e.,
V.sub.REF =V.sub.C. Hence,
R.sub.1 =V.sub.R1 /I=V.sub.R1 /(V.sub.REF /R.sub.REF), or (2)
R.sub.1 =(V.sub.A -V.sub.B)/(V.sub.C /R.sub.REF), and finally (3)
R.sub.1 =[(V.sub.A -V.sub.B)/V.sub.C ]R.sub.REF. (4)
If a two-wire RTD sensor were used such that there were no connection wires
from R.sub.1 to terminals A or B, then
R.sub.1 =[(V.sub.X -V.sub.C)/V.sub.C ]R.sub.REF, (5)
although no lead-length compensation would be performed. In other words,
using a two-wire RTD sensor, the sensor resistance is proportional to
V.sub.X -V.sub.C. When using a four-wire sensor, the sensor resistance is
proportional to V.sub.A -V.sub.B. If a three-wire RTD configuration were
used, wherein no connection wire exists from R.sub.1 to terminal B, the
RTD sensor resistance would be proportional to V.sub.A -V.sub.C, and a
different calculation would be performed to compensate for lead length. In
any case, note that the excitation current I provided by the current
source 46 is no longer part of the resistance equation, since the final
resistance equation is a function of ratios of voltage values. As will be
explained below, the microcontroller 22 utilizes these proportional
voltage values to calculate the resistance of the RTD sensor 24 in the
Measurement Mode. In the Test Mode, the known impedance 52 is switched in,
such that the measured value of the RTD sensor is now being affected by
the resistance of the known impedance connected in parallel.
FIG. 2b illustrates representative waveforms for the address lines
OPTO.sub.1, OPTO.sub.2, and OPTO.sub.3 from the optoisolator 30. Using
these waveforms, the operation of the multiplexing circuitry 42 of FIG. 2a
will now be described. As illustrated in the waveform diagram, a complete
measurement cycle is comprised of a Measurement Mode and a Test Mode.
During the Measurement Mode, the multiplexer address line OPTO.sub.3
always remains low, while it remains high through much of the Test Mode.
With a low voltage from OPTO.sub.3 applied to the resistor 52, the diode
50 is reversed biased, such that the known impedance has no effect on the
measurement of the output voltage at Z.
At time t.sub.1, the microcontroller 22 places a high voltage level on
address line OPTO.sub.1, while OPTO.sub.2 and OPTO.sub.3 remain low. Since
OPTO.sub.1 is connected to multiplexer address port a.sub.0, and since
OPTO.sub.2 is connected to address port a.sub.1, a binary `01` is applied
to the multiplexer 44 such that multiplexer data port d.sub.1 is connected
to the output port Z. Hence, at this time, the voltage level at the output
port Z of the multiplexer 44 represents the voltage level apparent at
input terminal B as measured from ground. This is shown as Z=B at the
bottom of FIG. 2b during the time interval t.sub.1 -t.sub.2. At time
t.sub.2, the address line OPTO.sub.2 goes high, such that a binary `11` is
applied to the multiplexer 44. Accordingly, the multiplexer selects its
data port d.sub.3 such that the voltage level at output port Z is equal to
that of the input terminal X, i.e., Z=X. At time t.sub.3, a binary `10` is
used to address multiplexer data port d.sub.2, such that Z=A. Finally, at
time t.sub.4, a binary `00` is used to address multiplexer data port
d.sub.0 such that Z=C. Accordingly, all four input terminals X, A, B and C
have been individually selected during the Measurement Mode.
In the Test Mode at time t.sub.6, address line OPTO.sub.3 goes high, such
that resistor 52 is now connected to the output port Z at 45 through the
forward-biased diode 50. Note that the diode 50 is serving the purpose of
a switch, under the control of the address line OPTO.sub.3, which connects
a known impedance, resistor 52, to the multiplexer output port Z. Since
the resistor 52 is now in the circuit, the voltage output at port Z is now
proportional to the resistance measurement at terminal B made in parallel
with the resistance of the known impedance 52. This is indicated at the
bottom of FIG. 2b as B+ (terminal B "plus" resistor 52) measured during
time interval t.sub.6 -t.sub.7. Similarly, during time interval t.sub.7
-t.sub.8, X+ is being measured. Resistance values A+ and C+ are then
measured during time intervals t.sub.8 -t.sub.9, and t.sub.9 -t.sub.10,
respectively.
In the preferred embodiment, the value of resistor 52 is 1,000,000 ohms.
This value is much higher than the normal operating range of resistance
values of the RTD, nominally 100 ohms, while it is much lower than the
potentially infinite resistance value seen at the output of the
multiplexer 44 if a wire is broken on the selected input port. In other
words, in the Test Mode, the presence of the known impedance 52 would not
significantly affect the voltage level at the output port Z when all the
wires are connected to a relatively low-impedance RTD. However, if one of
the wires is broken, the presence of the known impedance 52 will cause the
output voltage level to drastically change, or rise in this case, if one
of the RTD input connection wires is broken. Hence, if one of the
connection wires 25 is broken such that one of the multiplexer inputs is
open, the measured resistance value for that selected terminal will
approximate that of resistor 52. Accordingly, if a multiplexer input is
open, the frequency output of the V-to-F converter 32 will not be within a
nominal RTD range when that particular multiplexer channel is addressed.
Since the microcontroller 22 is controlling the address lines OPTO.sub.1,
OPTO.sub.2, and OPTO.sub.3, the microcontroller knows exactly which
multiplexer input terminal X, A, B, or C is being addressed. Hence, the
microcontroller can determine exactly which wire is broken, and indicate
this information to the user via the display 36.
FIG. 3a is an alternate embodiment of the multiplexing circuitry of FIG.
2a, which has been modified to utilize only two address lines. In the
multiplexing circuit 54 of FIG. 3a, the third address line OPTO.sub.3 has
been eliminated through the addition of a D-type flip-flop 56. Address
line OPTO.sub.1 is connected to the clock input, and address line
OPTO.sub.2 is connected to the D input, respectively, of the D-flip-flop
56. The flip-flop output Q is connected to the resistor 52. If the D input
is high, the Q output will go high on the next rising edge of the input
clock waveform. Hence, the phase relationship between address lines
OPTO.sub.1 and OPTO.sub.2 are used to determine the state of the
D-flip-flop output Q, which now functions as the third address line
OPTO.sub.3. Referring now to FIG. 3b, the operation of the multiplexing
circuitry 54 of FIG. 3a will be described. During the Measurement Mode,
the multiplexer address line OPTO.sub.1 always rises before the address
line OPTO.sub.2 such that the Q output of the flip-flop 56 remains low.
With a low output Q, the diode 50 remains reversed biased, such that the
known impedance 52 has no effect on the measurement of the voltage at
output port Z. Hence, as shown at the bottom of FIG. 3b, the output port Z
represents the resistance at terminals B, X, A and C, respectively.
However, note that the address line OPTO.sub.2 remains high at time
t.sub.6, such that the Q output goes high upon the rising edge of
OPTO.sub.1. Therefore, during time interval t.sub.6 -t.sub.7, the parallel
combination of the resistance seen at input terminal X plus resistor 52,
or X+, is being measured. Similarly, during time intervals t.sub.7
-t.sub.8, t.sub.8 -t.sub.9, and t.sub.9 -t.sub.10, resistance values B+,
A+, and C+, respectively, are measured in the Test Mode.
FIG. 4a is another simplified schematic diagram for the multiplexing
circuit 28 of FIG. 1, wherein an additional multiplexed input, terminal D,
is utilized to more accurately determine the value of the reference
resistor R.sub.REF. Note that the input terminal D is internal to the
transmitter unit. Also note that an additional address line is required to
select the additional input terminal D.
In multiplexing circuit 58, the reference resistor 62 is connected between
input terminals C and D of the multiplexer 60, and is not directly
connected to ground as before. Instead, a current return resistor
R.sub.RET 64 is connected between input terminal D and ground as shown. In
this way, the current from the current source 46 flows through the RTD 24,
the reference resistor 62, and the return resistor 64, to ground. Instead
of measuring the voltage V.sub.C to ground in order to determine the value
of the reference resistor R.sub.REF, two voltage measurements, V.sub.C to
ground and V.sub.D to ground, are made such that the value of the
reference resistor R.sub.REF is proportional to V.sub.C -V.sub.D. The use
of this fifth internal terminal D allows the measurement of the reference
resistor R.sub.REF to be completely differential, i.e., V.sub.C -V.sub.D,
such that voltage offsets no longer affect the accuracy of the
measurement. Hence, operational amplifiers may be used in the current
source 46 which do not have a necessarily low offset differential
specification. Since input bias and offset currents are no longer at
issue, the circuit has substantially no Zero or Span error, except for the
temperature coefficient of the reference resistor 62. The effect of any
noise on the ground lines is also significantly reduced. In other words,
in using this fifth input terminal configuration, a much more accurate
determination of the reference resistor R.sub.REF can be achieved.
In order to measure the voltage V.sub.D at the additional input terminal D,
an additional data port is required on the multiplexer. As shown in FIG.
4a, a five-channel analog multiplexer 60 is controlled by three address
lines a.sub.0, a.sub.1, and a.sub.2, which are connected to OPTO.sub.1,
OPTO.sub.2, and OPTO.sub.3, respectively. A fourth address line OPTO.sub.4
is connected to the known impedance 52 as shown.
Referring now to FIG. 4b, representative addressing waveforms for the
operation of the multiplexing circuit 58 of FIG. 4a are shown. In the
Measurement Mode, the fourth address line OPTO.sub.4 remains low, such
that the other three address lines OPTO.sub.1, OPTO.sub.2, and OPTO.sub.3
control the selection of the input terminal voltage which is applied to
the output port Z. For example, during time interval t.sub.1 -t.sub.2, a
binary `001` is applied to the multiplexer 60, such that data port d.sub.1
is selected, whereby the voltage at input terminal C is connected to
output port Z. Again, this is shown at the bottom of FIG. 4b as Z=C.
During time interval t.sub.2 -t.sub.3, a binary `011` is used to select
multiplexer data port d.sub.3, such that Z=A. Similarly, input terminals
B, D, and X are selected in accordance with the waveforms shown. At time
t.sub.7, the fourth address line OPTO.sub.4 goes high in the Test Mode
such that the known impedance 52 is switched into the circuit. Again, a
binary `001` address is sent by the microcontroller to select input
terminal C, such that Z=C+. A similar addressing scheme is used to select
A+, B+, D+, and X+ as shown.
FIG. 5 is a detailed schematic diagram of the preferred embodiment of the
MUX 28 of FIG. 1, wherein five multiplexed inputs are controlled by only
two address lines. In multiplexing circuit 70, the four external input
terminals X, A, B and C, have been reversed from the previous figures to
more accurately illustrate the operation of the current controller
circuitry and the isolated power supply circuitry. As before, the two
address lines OPTO.sub.1 and OPTO.sub.2 from the microcontroller 22 serve
to control the four-channel analog multiplexer 72 via the optoisolator 30,
and this circuitry operates substantially as explained above. However, an
additional two-channel analog multiplexer 74 is used to multiplex the
internal input terminal D with the output port Z.sub.1 of the multiplexer
72, and thereby provide the output port Z.sub.3 as the input to the
voltage-to-frequency converter 32 of FIG. 1. Another difference in FIG. 5
from the previous circuit is that another two-channel analog multiplexer
76 is used as an electronic switch to perform the function of the diode
50, i.e., to control the switching of the known impedance in the Test
Mode. Finally, note that a D-type flip-flop 80 is again used to eliminate
the need for the third address line, and a diode-OR circuit is used to
eliminate the need for the fourth address line.
The circuitry in the upper-left portion of FIG. 5 performs the function of
the current controller 26 of FIG. 1. An operational amplifier 82, powered
from a split voltage supply V+/V-, is used to sink current returning into
the tr | | |
