The present invention relates to a method and apparatus for calibrating
moisture/yield monitors of the type used in grain harvesters to measure grain moisture
content and crop yield. The invention provides a method and apparatus for calibrating
a moisture/yield monitor on-the-go, that is, without interruption of the harvesting
operation. The calibration may employ an RF attenuator traceable to the National
Institute of Standards and Technology, so that the moisture measurement of grain
is a traceable absolute measurement.
It is known that the moisture content of grain and its bulk density
may be determined by transmitting an RF signal in the low GHz frequency range through
the grain and determining the change in magnitude and phase of the transmitted
signal resulting from its passage through the grain. The change in magnitude (attenuation)
results primarily from the moisture in the kernels of grain, whereas the phase
shift is dependent on the bulk density of the grain and, to a lesser extent, on
its moisture content. Knowing the moisture content and bulk density of the grain,
and by measuring the rate of grain flow through the harvester, the total crop
yield may be determined.
It is obviously important to a grower to know his crop yield. The
moisture content of the grain is also important in that it affects the quality
and market value of the grain and its suitability for storage. Therefore, it is
desirable that measurements of the moisture content and yield be as accurate as
possible. This, in turn, requires accurate calibration of the measurement apparatus.
According to a current method of calibration, microwave apparatus
for measuring grain moisture content is calibrated while the apparatus is empty
to permit nulling out the imbalance or error in the system resulting from contamination
or changes in system components during the measurement interval since the last
calibration. This method has several disadvantages. First, since the sensor must
be empty, the harvesting operation must be interrupted and the accompanying moisture
measurement activity suspended. Secondly, if calibration is lost, there is no
real time record of when the calibration was lost. This means that the accuracy
of any measurement made since the last calibration is suspect. Thirdly, the process
of nulling out is done by a human operator and thus susceptible to error.
Therefore it is an object of the present invention is to provide
a method and apparatus for calibrating a grain moisture measurement system, the
method and apparatus having none of the disadvantages of the prior art.
Another object of the invention is to provide a method and apparatus
for calibrating a grain moisture measurement system during the time the system
is performing its moisture measurement function.
A further object of the invention is to provide a method and apparatus
for calibrating a moisture measurement system in a grain harvester while the grain
harvester is harvesting grain from a field.
Still another object of the invention is to provide a measurement
system which produces a measurement of the moisture content of grain that is a
traceable absolute measurement.
According to one aspect of the present invention there is provided
apparatus for measuring the moisture content of grain comprising:
- a radio frequency signal source for producing a measurement signal;
- a measurement signal path including:
- a transmit antenna responsive to said signal source for transmitting said measurement
signal through the grain;
- a gain control means for controlling the magnitude of said measurement signal
as said measurement signal moves along said measurement signal path;
- a receive antenna responsive to said measurement signal after it has passed
through the grain; and
- receiver circuit means responsive to said receive antenna for producing signals
indicative of the attenuation and phase shift of said measurement signal resulting
from the passage of the measurement signal through the grain; and
- processor means responsive to said produced signals for determining the attenuation
of said measurement signal and the moisture content of the grain.
Said apparatus is characterized in that it further comprises:
- means for selectively introducing into said measuring signal path a resistance
of known value to cause an expected attenuation of said measurement signal;
said processor means including:
- means for determining the difference between the expected and an actual attenuation
of said measurement signal as a result of introduction of said resistance into
said measurement signal path; and
- means for adjusting said gain control means according to said difference.
According to another aspect of the present invention, a method is
provided of calibrating an apparatus for measuring the moisture content of grain
flowing along a grain flow path in a grain harvester, said apparatus having:
- a radio frequency signal source and
- a measurement signal path, including:
said method being characterized in that it comprises the steps
- a transmit antenna responsive to said signal source for transmitting a measurement
signal through the flowing grain;
- a receiver antenna responsive to said measurement signal after it has passed
through the grain;
- receiver circuit means responsive to said receiver antenna for producing signals
indicative of the attenuation and phase shift of said measurement signal resulting
from the passage of the measurement signal through the grain;
- a gain control means for controlling the gain of said measurement signal path;
- processor means responsive to said signals indicative of the attenuation and
phase shift for determining the attenuation of said measurement signal,
- while grain is flowing along said grain flow path, intermittently introducing
into the measurement signal path an attenuator of known value to cause an expected
attenuation of said measurement signal;
- determining the difference between the expected attenuation of said measurement
signal and the actual attenuation of said measurement signal resulting from the
introduction of said attenuator into the measuring signal path; and
- adjusting said gain control means in accordance with said difference.
Preferably, the introduced resistance has a known attenuation traceable
to a standard. The apparatus may include a PIN diode having a calibrated resistance
traceable to a standard, the PIN diode being connected in the measurement signal
path. Herein a control means selectively biases the diode so that it exhibits no
resistance when the apparatus is performing its measurement function but introduces
a known resistance into the path during calibration intervals. Introduction of
the known resistance into the measurement signal path should produce a known or
expected attenuation of the measurement signal if the apparatus is properly calibrated.
The processing means determines the difference between the actual and expected
attenuations and adjusts the gain of the measurement signal path according to the
For reducing the effects of random noise, the gain control means
may be connected in the signal path downstream of the receive antenna. The resistance
preferably is connected in the signal path upstream of the antenna.
An apparatus and a method in accordance with the present invention
will now be described in greater detail, with reference to the following drawings,
- Figure 1 illustrates a first embodiment of an apparatus for measuring the moisture
content of grain, the apparatus including a PIN diode selectively biased by a
processor to introduce a known value of resistance into a measurement signal path
for calibration purposes;
- Figure 2 illustrates an alternative arrangement for introducing a known resistance
into a measurement signal path; and
- Figure 3 illustrates a routine executed by the processor of Figure 1 to control
automatic on-the-go calibration of the apparatus.
Referring to Figure 1, an apparatus for measuring the moisture content
of grain and determining crop yield comprises an RF transmitting means 10, a receiver
circuit means 12 and a controller or processor means 14.
The RF transmitting means 10 comprises an RF source 16 for continuously
producing a measurement signal in the frequency range of about 1-10 GHz, a signal
splitter 18, and a transmit antenna 20. The measurement signal output of RF source
16 is connected to the signal splitter 18 and one output of the signal splitter
is applied to the transmit antenna 20 through a PIN diode 22.
A receive antenna 24 drives the receiver circuit means 12. The antennas
20, 24 are disposed on opposite sides of a closed conveyor chute 26 which comprises
a part of the grain flow path in a grain harvester. The chute 26 is made of a
low dielectric plastic material, or is provided with windows of such a material,
so as to have minimal effect on the RF signals as they pass through it. Preferably,
an interrupted flyte auger (not shown) moves the grain through chute 26. As is
known in the art, the auger flyte is interrupted so that grain may accumulate and
completely fill chute 26 in the region between antennas 20, 24, thereby providing
a more uniform density of the grain through which the signal from transmit antenna
The receiver circuit means 12 includes a quadrature demodulator 28
and, as illustrated in Figure 1, a gain control means such as a controllable amplifier
30 connected between the receive antenna 24 and one input of the demodulator.
The output of signal splitter 18 is connected to a second input of the demodulator
via a lead 32 so as to apply to the demodulator a reference signal which is in
phase with the signal applied to transmit antenna 20.
The demodulator may, for example, be a type MIQ64M5-2 quadrature
demodulator commercially available from Loral Corporation, Eagan MN. The demodulator
produces two time varying output voltages I and Q where I and Q represent the
magnitude of the output signal from amplifier 30 measured at 0° and 90° of the
reference signal on lead 32. The signals I and Q provide an indication of the attenuation
and phase shift of a measurement signal as the signal is propagated over the measurement
signal path extending from splitter 18 to antenna 20, through the grain in chute
26, and from the antenna 24 to demodulator 28. The attenuation A of the measurement
signal is (I2+Q2)1/2 volts and the tangent of
the phase shift &phis; is Q/I.
The controller or signal processor means 14 comprises a processor
34, an operator's control panel 36 and an alphanumeric display 38. The control
panel and display are located near the operator's seat in the harvester so that
the operator may conveniently enter control commands or data into the processor
and observe data and messages generated by the processor.
The processor 34 may be a conventional programmable microprocessor
having ROM, RAM and EEPROM memories as well as analog to digital (ADC) and digital
to analog (DAC) conversion circuits.
In order to determine crop yield, it is necessary to know the speed
of grain flow through the chute 26. The means for sensing grain speed comprises
a doppler transmit/receive antenna 40 connected to a doppler transceiver 42. The
output signal F from transceiver 42 and the output signals I and Q from demodulator
28 are all applied to the ADC circuits of the processor.
As subsequently explained, the processor 34 operates during measurement
intervals to repetitively sample the signals Q, I and F. From signals I, Q and
F the processor determines grain moisture content, bulk density and yield. Intermittently,
and while grain may be flowing through chute 26, the processor executes a calibrate
routine and, if adjustment is necessary, sets the control voltage applied to the
gain circuit 30 over lead 48 during subsequent measurement intervals. For this
purpose, the DAC circuits of processor 34 produce output signals on leads 44,
46 and 48.
The signals on leads 44 and 46 are applied through an RF isolation
circuit 50 to two junctions 52, 54. Processor 34 with its DAC circuits comprise
a means for selectively applying a bias current to the PIN diode 22. The diode
may, for example, be a Hewlett-Packard type 5082-3340 calibrated PIN diode attenuator.
Calibration data with high accuracy, traceable to the National Institute of Standards
and Technology, are available with these devices from the manufacturer. When zero
or reverse bias is applied to the diode it exhibits a very high resistance, introducing
a known attenuation into the measurement signal path. When a forward bias is applied
to the diode its resistance drops to practically zero, actually to a value less
than one ohm.
The RF isolation circuit 50 serves to isolate the RF output of splitter
18 from the processor 34. The inductances L1, L2 serve as
chokes to block the RF signal. Capacitors C1, C2 and C3
act to filter any of the RF signal passing through the inductances to prevent it
from reaching the processor.
Assume that the apparatus of Figure 1 is in a measurement mode, that
is, not in the process of being calibrated. The processor 34 produces signals on
leads 44 and 46 to forward bias diode 22 so that it introduces no resistance or
loss into the measurement signal path. The continuous output signal from RF source
16 is split by splitter 18 and applied to transmit antenna 20 and demodulator
The measurement signal is transmitted through the grain flowing in
chute 26 where it is attenuated and shifted in phase to a degree dependent on the
internal moisture content of the kernels of grain and the density of the grain.
The phase shifted and attenuated measurement signal is amplified by gain control
means 30 and applied to demodulator 28 which in turn produces the time varying
voltage signals I and Q that are applied to the ADC circuits of processor 34.
As will be evident from the following description, the diode 22 and
gain control means 30 do not have to be located in the circuit positions shown
in Figure 1. Either of these elements may be located anywhere in the measurement
signal path downstream of the signal source 16 and upstream of demodulator 28.
Figure 3 illustrates a routine executed by processor 34 to calculate
grain moisture content and yield and, in accordance with the present invention,
calculate a correction factor for calibrating or adjusting the gain of the measurement
signal path. The routine is executed periodically at intervals of 10 ms but this
interval is not critical. During each execution of the routine, steps 100-108
are carried out to calculate moisture content and yield. In addition at intervals
of, for example 1 sec, steps 109-118 are carried out to develop a calibration
factor for adjusting the gain in the measurement signal path, or produce a message
for the operator regarding the condition of the measurement system.
At step 100, the microprocessor gets from the ADC circuits digitized
values corresponding to the magnitudes of the analog signals Q, I and F. Step 101
computes the phase shift &phis; and attenuation A imposed on the measurement signal
as it passes over the measurement signal path and through the grain in chute 26.
The phase shift &phis; is computed according to the equation
&phis; (in degrees) = TAN-1
and the measured attenuation is computed according to the equation
A (in volts) = (I2 + Q 2)1/2
Step 102 converts the attenuation to an equivalent decibel loss dBm.
Step 103 obtains from the processor memory the grain coefficients
required to calculate grain moisture content. In this regard, the kernels of different
grains such as wheat, corn, etc. and different varieties of the same grain, such
as soft and hard red winter wheat, exhibit different physical characteristics which
must be taken into account in determining the moisture content of the grain. A
look-up table is provided in the non-volatile memory in processor 34 and this table
stores constants which are used in computing the moisture content. Prior to initiation
of a harvesting operation, the operator enters through control panel 36 an indication
of the type/variety of grain to be harvested. At step 103, this indication is used
to address the look-up table to obtain the constants to be used in calculating
the moisture content and bulk density.
At steps 104 and 105, the processor 34 calculates the moisture content
(MC) and the bulk density using the measured values of attenuation and phase shift
and the appropriate constants obtained from the look-up table. The prior art discloses
many experimentally developed equations for computing grain moisture content and
bulk density from the measured phase shift and attenuation imparted to an RF signal
as it passes through grain. For example, the calculations of moisture content and
bulk density may be carried out according to the equations of Kraszewski and Nelson
as set forth in Canadian Agricultural Engineering, Vol. 34, No. 4, pgs. 327-335
(1992). The calculations of moisture content and bulk density are not, per se,
a part of the present invention and need not be discussed further.
Step 106 determines the incremental yield (IYIELD). That is, at step
106 the processor determines the crop yield in the interval of time elapsing between
consecutive executions of step 106.
IYIELD = BD * F * K1 * K2
where BD is the bulk density calculated at step 105, F is the velocity of grain
flow through chute 26, K1 is the cross-sectional area of the interior
of chute 26 and K2 is the time between consecutive executions of step
At step 107, IYIELD is added to total yield (TYIELD). The value in
TYIELD represents the total crop yield or total crop processed since the start
of the harvesting operation.
In summary, steps 100-107 sample and convert the analog output signals
Q, I and F from demodulator 28 and transceiver 42 to digital values, determine
the phase shift &phis; and attenuation A of the measurement signal, compute the
moisture content and bulk density of the grain, and from the bulk density determine
the total crop yield.
According to the routine illustrated in Figure 3, calibration is
carried out once each second. A timer (T) tolls the one-second intervals. T is
incremented by 10 ms each time before step 108 is executed and the value in T is
tested to determine if it represents one second. If one second has not passed
the processor bypasses calibration steps 109-118 and exits the routine.
On the other hand, if timer T holds a value representing one second
when it is tested at step 108, the timer is reset to begin tolling another one-second
interval and the processor 34 advances to step 109 to perform a calibration.
The interval of one second is not critical. Other intervals may be
used. It is possible to eliminate step 108 entirely so that the processor performs
a calibration following each measurement. In this case, the step 108 is eliminated
and the routine moves from step 107 directly to step 109
The processor 34 carries out a calibration by reverse biasing PIN
diode 22, measuring the actual attenuation of the measurement signal with the resistance
of the diode in the measurement signal path, and determining the difference between
the actual attenuation and the expected attenuation. Assuming the PIN diode 22
is a 20dB attenuator, the expected attenuation is assumed to be 20dB greater than
the attenuation dBm determined at steps 100-102 immediately prior to
insertion of the diode resistance into the measurement signal path. Although grain
may, or may not be flowing, this assumption is quite accurate because the measurement
of attenuation for calibration purposes always takes place in a matter of microseconds
after steps 100-102 are executed.
At step 109, the processor sets outputs so that the DAC circuits
produce output signals on leads 44, 46 to reverse bias PIN diode 22. The resistance
of the PIN diode increases to a known value, thus introducing into the measurement
signal path a resistance which should produce a known expected attenuation of the
measurement signal if the gain of the measurement signal path is properly adjusted.
To determine the actual attenuation resulting from insertion of the
PIN diode resistance into the measuring path, the processor 34 samples the signals
I and Q (step 110) while the PIN diode is still reverse biased, and from I and
Q computes the actual attenuation Aa (step 111).
Processor 34 next determines the difference between the expected
attenuation and the actual attenuation resulting from insertion of the PIN diode
resistance into the measurement signal path. The actual attenuation value Aa
calculated at step 111 is expressed in volts and this value is converted (step
112) to the equivalent actual loss dBa in decibels. The actual loss
dBa should be 20dB greater than the loss dBm determined at
step 102. Therefore, at step 113 the processor subtracts from the actual attenuation
the sum of 20dB plus the attenuation dBm to obtain the
difference dBe between the actual attenuation and the expected attenuation
resulting from insertion of the diode resistance into the measurement signal path.
The value dBe is used to adjust the gain of the measurement
signal path so that, assuming no further changes in the measurement signal path,
the value dBe will be zero the next time a calibration is carried out.
However, the value dBe cannot be used directly to set the output signal
on lead 48 to control the gain. At the time the signal samples are taken at steps
100 and 110, the processor 34 is already applying a gain adjustment signal to
amplifier 30 as a result of prior calibrations which developed prior difference
values dBe. A correction factor CF, representing the algebraic sum of
these prior difference values is stored in memory. Step 113 adds dBe
to CF and the result is saved for use during the next calibration.
The value CF is used to set the magnitude of the gain adjustment
signal applied to amplifier 30 over lead 48. However, before the gain adjustment
signal is set, the value CF is checked to determine its magnitude. At step 114,
the value of CF is checked to see if it represents an attenuation greater than
60. If so, it is probably a result of a circuit failure. At step 115 the processor
sets an output message which is applied to display 38 to inform the operator.
An audible alarm may also be sounded to call the operator's attention to the error.
If step 114 determines that CF is less than 60, step 116 is executed
to determine if CF is greater than 5. If CF is greater than 5, it is an indication
that some maintenance should be performed such as cleaning dirt/moisture from
the region between the antennas 20, 24. In this case an appropriate message is
sent to display 38 (step 117) and an audible alert is sounded.
If step 116 determines that CF is no greater than 5, the value of
CF is used (step 118) to set an output signal to the amplifier 30 to thereby adjust
the gain of the measurement signal path. CF, which is in terms of decibels, is
used to access a conversion table. The resulting output value from the conversion
table is converted to an analog signal by the ADC circuits and the analog signal
is applied over lead 48 to the amplifier.
The resistance of PIN diode 22 may be removed from the measurement
signal path at any time after I and Q are sampled at step 110 by placing signals
on leads 44, 46 to forward bias the diode.
Although Figure 1 illustrates the invention in a specific moisture/yield
monitoring system, it should be understood the invention is equally suitable for
use in other systems such as, for example, those systems wherein the signal from
the receive antenna 24 is split and applied in parallel to a phase comparator and
an amplitude comparator for comparison with the reference signal.
The invention may also be implemented in different ways as illustrated
in Figure 2. For example an attenuator 60 calibrated to a traceable standard, may
be connected in the measurement signal path and selectively switched into, or
out of, the path by a suitable switching means 62 illustrated as a PIN diode. When
the diode is forward biased, the measurement signal flows through the diode and
the attenuator is removed from the signal path. When the diode is reverse biased
it blocks the measurement signal path and the attenuator is electrically inserted
into the signal path.
Although illustrated as a PIN diode, switching means 62 may be a
pair of electronic switches or even a manual switch with contacts in series with
The bias voltage for biasing the PIN diode 22 or 62 need not come
from the processor 34. A manual switch 64 may be provided for selectively connecting
a source of bias voltage to the diode. Output lead 66 from the switch may be connected
to the processor 34 and, in Figure 3, step 108 is replaced by a step in which actuation
of switch 64 is sensed.
Other modifications and substitutions may be made in the preferred
embodiment without departing from the spirit and scope of the invention as defined
by the appended claims.