Temperature responsive transmitter.

Technical Field

The invention relates to an implantable or ingestible temperature indicating transmitter whose output frequency varies linearly with temperature.

Background Art

Accurate temperature measurement is particularly important in the medical and veterinary fields, where internal temperature measurements indicate the health of a patient, or how the patient is reacting to various conditions. Measuring the internal temperature of a part of a patient's body may be accomplished by implanting, or ingesting, a temperature responsive transmitter.

Temperature responsive transmitters generally utilize a temperature sensitive component which varies in some physical characteristic, such as resistance, inductance, or capacitance, as temperature changes. This variation, in turn. is employed to change the frequency of the transmitter. Factors of interest which affect the design of a temperature responsive transmitter include linearity with temperature, frequency stability, frequency range, power consumption, output power, temperature range, accuracy, physical size, and cost.

Typical temperature sensitive components used include thermistors, diodes, and piezoelectric crystals. Where a thermistor is used to vary the frequency of a transmitter, linearity and stability have proved to be unsatisfactory. Where a diode is used, the voltage across the diode varies with temperature, and this voltage change must be converted to a change in frequency. This requires complex circuitry, and the linearity of conversion is generally not satisfactory. The complex circuitry also increases the size of the transmitter which makes it more suitable for external rather than internal use. When a piezoelectric crystal is used as the frequency determining element of a transmitter, linearity and frequency stability are satisfactory if the crystal is specially cut to have a large linear temperature coefficient. Previously, these specially cut crystals were relatively expensive, and this limited their use to situations in which the probability of loss of the temperature responsive transmitter was low.

According to the present invention, these drawbacks are overcome by providing two identically biased transistors connected together in parallel. A capacitor is connected to the common emitters such that its capacitance is reflected to the common bases to generate negative resistance. The negative resistance and the capacitor are effectively in parallel, and oscillations will occur if an inductor, such as a crystal, is placed effectively in parallel with them and if the positive resistive impedance of the capacitor and the inductor is less than the negative resistance. The crystal is selected to have a linear temperature coefficient and a resonant frequency which is substantially less than the gain-bandwidth product of the transistors. A high-Q tank circuit having an inductor and a capacitor is connected to the common collectors.

Brief Description of the Drawings

<ul id="ul0001" list-style="none"><li>Figure 1 is a schematic diagram of a temperature responsive transmitter of this invention.</li><li>Figure 2 is a schematic diagram of a temperature responsive oscillator of this invention.</li></ul>

Detailed Description of the Invention

Referring now to the drawings, Fig. 1 shows a temperature responsive transmitter 10 including two transistors T₁, T_Z connected in parallel. Two transistors are advantageously used to double the output power of the transmitter while simultaneously keeping the relationship between frequency and the gain-bandwidth product of the transmitter constant over its entire frequency range. To triple the output power of the transmitter 10, three transistors connected in parallel may be used. In the illustrated embodiment, the two transistors may either be discrete or dual bipolar npn transistors, although other active devices may also be used. A pair of MAT-02 bipolar npn transistors manufactured by Precision Monolithics Inc. of Santa Clara, California have been found to perform satisfactorily.

The quiescent point of the transistors T₁ and T₂ is set by a voltage divider network connected between a source of B+ voltage and ground, and by a resistor R_E connected between the common emitter electrodes and ground. The voltage divider network is formed of serially connected resistors R₁ and R₂ connected at their juncture 20 to the common base electrodes of the transistors, thereby providing the bases with an intermediate voltage. In the preferred embodiment the value of the B+ voltage is +3 volts, which may be supplied by a battery. Lithium batteries, which are very small in size and which are rated at 35 to 50 milliampere-hours to last at least 100 hours at a constant current drain of 0.1 milliampere, are preferred where the transmitter 10 is to be implanted or ingested in a patient for a prolonged period of time. Resistors _R1 and R₂ are both 1MΩ which provides the bases with 1.5 volts at the juncture 20. Resistor R_E is between 20 and 40 KΩ. A capacitor C₁ is connected in parallel with resistor R_E so that resistor R_E will affect only the direct current operation of the circuit. It also serves other purposes which will be discussed hereinafter. The value of C1 in the preferred embodiment is 100 pf.

A piezoelectric crystal 30 is connected in parallel with resistor R₂. The resonant frequency of the crystal 30 primarily determines the frequency of the transmitter 10. One factor that plays a part in determining the resonant frequency of the crystal 30 is its temperature coefficient which, in turn, depends upon the mechanical dimensions and the elastic constants of the crystal, and the mode of vibration involved. If a crystal which is specially designed and cut to have a large linear temperature coefficient is used, the frequency of the transmitter 10 will vary linearly with temperature over the range of the crystal. In the preferred embodiment, a 262.144 kHz quartz crystal having a large positive linear temperature coefficient was employed. An example of such a crystal is quartz crystal TS-2 manufactured by Statek Corporation of Orange, California.

The piezoelectric crystal 30 may be represented by an equivalent circuit having a capacitance shunting series resonant circuit. The capacitance is due to the crystal electrodes, and its value is very small, typically only a few picofarads. The series resonant circuit is comprised of an inductance, a capacitance, and a resistance. Typically, the value of the inductance is a few henries, the value of the capacitance is a few hundredths of a picofarad, and the value of the resistance is less than one hundred ohms. Crystal 30, therefore, is primarily an inductive impedance.

The inductive impedance of the crystal 30, along with the capacitance of the capacitor C₂, are effectively combined in parallel in the circuit of the transmitter 10 to create a two terminal oscillator. A two terminal oscillator can be defined as a negative resistance in parallel with a LC tank circuit. The inductive impedance of the crystal 30 provides the L required for the LC tank circuit. The capacitance of C₁, which is reflected into the commonly connected bases, provides the C required for the LC tank circuit. The negative resistance is provided by the action of the reflected capacitance from the capacitor C₁on the circuit of the transmitter 10, and it is effectively in parallel with the LC tank circuit. Oscillations will occur in the two terminal oscillator if the magnitude of the absolute value of the negative resistance is less than the positive resistive impedance of the LC tank circuit. In the embodiment illustrated, oscillations will take place if the magnitude of the absolute value of the negative resistance is larger than the positive resistance seen in the commonly connected bases of the transistors T₁ and ^T₂^due to the resistors ^R₁, R₂, and R_E. The two input terminals of the oscillator are the commonly connected bases and the commonly connected emitters of the transistors T₁ and T₂.

The value of the capacitance which is reflected into the joined bases of the transistors T₁ and T₂, and the value of the negative resistance generated by the capacitance of C₁, are both functions of the value of the capacitor C₁, the gain-bandwidth product of the transmitter 10, and the frequency of oscillation. The value of the reflected capacitance, Cref may be expressed as:<maths id="math0001" num=""><img file="EP0243573A2_D0001.tif" /></maths>where w_G is the gain-bandwidth product of the transmitter 10, and w is the angular frequency of oscillation. The value of the negative resistance, -R, may be expressed as:<maths id="math0002" num=""><img file="EP0243573A2_D0002.tif" /></maths>

In equations (1) and (2), the value of the gain-bandwidth product and the frequency of oscillation serve to limit the values of the reflected capacitance and the negative resistance. The gain-bandwidth product and the frequency of oscillation, which vary with each other in an inverse square relationship, also determine whether the crystal 30 primarily determines the frequency of oscillation. If the gain-bandwidth product is approximately four times the frequency of the crystal 30, then the crystal will primarily determine the frequency of oscillation. In the preferred embodiment, the gain-bandwidth product is _{1 M}Hz, which is approximately four times the value of the 262.144 KHz crystal 30. Accordingly, the crystal 30 will primarily determine the frequency of oscillation.

The two terminal oscillator present in the circuit of the transmitter 10 is extremely stable. This is partly due to the way the circuit is biased. The resistors R₁ and R₂ are chosen so that the bias voltage at the commonly connected bases is equal to one half of the voltage of the source of B+ voltage to assure symmetry of the oscillator signal. This biasing arrangement, along with the circuit configuration of the transmitter 10, causes the gain of the circuit to be below the level at which hard limiting or clipping of the oscillator signal occurs. If the oscillator signal, which in the preferred embodiment is sinusoidal, is hard limited or clipped, then excessive noise is produced, which in turn produces frequency instability. Furthermore, the low gain level, which prevents hard limiting or clipping, allows the transmitter 10 to operate at full power throughout each portion of each cycle.

The high stability of the two terminal oscillator present in the circuit of the transmitter 10 is also partly due to the capacitor C₁ and the reflected capacitance generated thereby. The voltage across the capacitor C₁, which controls the positive feedback, is a pproximately 180° out of phase with the voltage at the common bases of the transistors T₁ and T₂. The phase difference drives the common bases positive while simultaneously driving the common emitters negative in a regenerative process. This regenerative process is positive feedback, or equivalently, negative resistance. The difference between the voltage across the capacitor C₁ and the voltage at the common bases cancel, and this causes the quiescent point of the transistors T₁ and T₂ to remain constant and thereby eliminate frequency drift.

The output from the two terminal oscillator is derived from the commonly connected collector electrodes, whose current is multiplied by the Q of a parallel resonant LC network comprised of an inductor L₂, which functions as the radiating element for the transmitter 10, and a capacitor C₂. The Q of the LC network multiplies the current present in the common collectors to enhance the inductive current present in inductor L₁, which in turn increases the transmission distance. Inductor L₁ and capacitor C₂ are connected between the source of B+ voltage and the common collectors. In the preferred embodiment, the value of the inductor L₁ is 0.6µh, and the value of the capacitor C₂ is 300 pf. The inductor L₁ may be wound in miniature size so as to be usable in an ingestible pill. For example, it may have 100 to 200 turns, a diameter of 3/16 of an inch, and a length of 1/2 of an inch.

In some applications it may be desirable to use the circuit of transmitter 10 as only an oscillator, and not as a transmitter. To accomplish this, the parallel LC network of Figure 1 is replaced by a resistor R, as illustrated by the oscillator circuit 40 in Figure 2. The preferred value of the resistor Rcis 200Ω. The output from this circuit is preferably derived from resistor R to maintain the symmetry of the circuit. The oscillatory frequency may also be determined by a frequency absorption meter or by a dip meter placed in close proximity to the oscillator.

Referring again to Figure 1, the temperature measuring transmitter 10 will produce a signal level of several microvolts in the antenna circuit of a receiver, not illustrated, positioned approximately a third of a meter away from the transmitter. Any receiver which can indicate the frequency of the transmitted signal can be used because the transmitter 10 only generates a range of radio frequencies, which in the preferred embodiment is from 262.094 KHz to 262.194 KHz. The transmitter 10 will have a maximum transmission distance of one meter, but the most effective distance is approximately a third of a meter. To achieve these distances where the transmitter 10 is to be implanted or ingested in a patient, a frequency range is selected which will provide maximum transmissibility through living tissue, and which will not be subject to interference from other users of the radio frequency spectrum. The frequency range selected for the preferred embodiment is centered on 262.144 KHz, which is the frequency of the quartz crystal 30.

The range of temperatures which may be measured is from -10°C or less to +140°C or more, with an accuracy of ±0.01°C. Where the transmitter 10 is to be used to measure the internal temperature of a human body, the temperature range is 30-40°C, and an accuracy of 0.1°C is required. With a temperature range of 3.0-40°C, the frequency range of the preferred embodiment is 262.144 KHz ± 50 Hz, and the frequency stability 0.1 Hz. This translates to 9 Hz/°C, or 0.9 Hz/0.1°C, with an accuracy of at least 0.1°C.