Patent Application
20080094057
With reference, first, to
The implementation of transmitter 32 and receiver 33 is specific to a given system with respect to the number and location of axes in transmitter 32 and the number and location of sensor channels in receiver 33. In general, the number of transmit axes in transmitter 32 multiplied by the number of sensor channels in receiver 33 is at least equal to the number of degrees of freedom measured by the system. Individual axes for transmitter 32 are operated sequentially, meaning that a given measurement and compensation interval is completed for a given axis before energizing the next axis. Individual axes for receiver 33 may be measured simultaneously, such that individual axes are processed in parallel during a given transmit axis interval. As such, for clarity, the description of the preferred embodiment will focus on a single transmit axis and a single sensor channel which form the core of the present invention.
Referring to
With further reference to
Sensor 2 is sensitive to the time derivative of a magnetic field and is preferably a coil of wire. The graph 13 in
Integrator 4 is started at time T0 and its output is shown in the graph 14 in
The output of second integrator 5 is divided by the output of transmitter integrator 17 to form a ratio which is dependent only on the coupling between transmitter 1 and sensor 2. That output is independent of effects from eddy field generator 3 provided the interval T1−T0 is long enough to allow the current in eddy field generator 3 to decay to zero. It is also independent of the magnitude or shape of the current pulse through transmitter 1. This simplifies the design and construction of driver 16, as precise control over the current waveform is not needed.
Below, it is shown how a magnetic position measurement system employing the described method can be made immune from eddy current effects.
In a non-ideal system such as that of the preferred embodiment, the value from sensor 2 is amplified by amplifier 18, which for practical devices will output a non-zero output if fed a zero input, the so called offset voltage. This offset voltage, when double integrated by integrator 4 and second integrator 5, is a significant source of error. A method for eliminating this error is presented below.
It is noted that during the compensation interval T2−T1 it is not necessary to energize the transmitter to obtain a useful, eddy-response-free and drift-corrected output from second integrator 5 at time T2. This could simplify the construction of transmit driver 16. The penalty for this simplification is a reduction in signal-to-noise ratio, as there is no signal to measure during the second interval, as only the error terms are being measured and subtracted.
It is also noted that inverting the polarity of sensor 2 during second interval T2−T1, by some means such as analog switches or relays, could be employed instead of inverting the transmitter current. This would allow driver 16 to be of unipolar construction which would simplify its circuitry.
It is further noted that the system can be made to work with T1−T0 not equal to T2−T1. This would require a more complex but readily implemented drift measurement and compensation technique, readily implemented by those skilled in the art.
Also, use of a DC sensitive sensing means for sensor 2, such as a fluxgate magnetometer, magnetoresitive, or Hall effect senor, would allow the output of integrator 4 to be used in place of the output of integrator 5 in the system, with all other aspects of operation remaining the same. Second integrator 5 is then removed from the system.
Additionally, if the output of driver 16 were stable and repeatable, the transmit reference integrator 17 could be removed and a fixed numerical constant representing the known total transmitted flux time integral used instead. In this case, circuitry used to determine the total flux time integral would not be used and the output of second integrator 5 could be used directly to compute the position of sensor 2.
Driver 16 can be a controlled current source, stored charge in a capacitor which is discharged across the coils of transmitter 1, or a simple voltage source and a switch. The exact means by which a charge of electrons is moved through transmitter 1 would be determined by convenience and performance requirements.
In the preferred embodiment, transmitter 1 possesses a linear current to field relationship, such that the simpler current measurement and integration method can be employed accurately. If it were desirable to reduce the weight of transmitter 1, one method would be to employ thinner core material. This has the effect of creating higher flux density and this can result in a non-linear current to field transfer function as the flux density approaches saturation. In this case, current sense resistor 18 and integrator 16 could be replaced with a fixed magnetic sensor either integral with or external to the transmitter. If this sensor were DC sensitive, such as a magnetoresistive sensor, it would be followed by a single integrator. If it were derivative sensitive, such as a coil, it would be followed by two integrators. In either of the two latter cases, it is again possible to measure a value proportional to the total flux integral from transmitter 1.
Furthermore, the system can operate with moving permanent magnets in place of transmitter 1. The time derivative pulse is then created during each interval T1−T0 and T2−T1 by rotating the magnet at the start of each interval. The reference integral in this case could be made by a number of means obvious to those skilled in the art. If the magnet position were accurate enough and the positioning method repeatable enough, the reference integral could be omitted and a known constant employed.
When the distance between transmitter 1 and sensor 2 becomes close, amplifier 18 may saturate. If the gain bandwidth of amplifier 18 is reduced such that it acts as a lowpass filter, it is possible to operate transmitter 1 in a manner such that the peak charge rate and/or duration of each charge pulse is reduced. This effectively spreads the charge over a longer period of time and reduces the peak level, which results in amplifier 18 staying within a linear range. It is also possible to increase or decrease the number of these smaller pulses, which may further benefit the signal to noise ratio as explained earlier.
In order to optimize the signal-to-noise ratio of the system in the presence of conductive metals, a mode of operation is shown in
Graph 53 is of the output of first integrator 4, not really used in the text as it is sort of an intermediate product, serving as the input of second integrator 5 from which the signal of interest is taken.
As such, an invention has been disclosed in terms of a preferred embodiment that fulfills each and every one of the objects of the present invention as set forth hereinabove and provides a new and useful position measurement system employing total transmitted flux quantization of great novelty and utility.
Of course, various changes, modifications and alterations in the teachings of the present invention may be contemplated by those skilled in the art without departing from the intended spirit and scope thereof.
As such, it is intended that the present invention only be limited by the terms of the appended claims.
