RELATED APPLICATION 0001 This Application is a continuation-in-part of both U.S. patent application Ser. No. 09/820,498, filed Mar. 28, 2001, and titled GROUND-PENETRATING IMAGING AND DETECTING RADAR; and, U.S. patent application Ser. No. 10/161,378, filed Jun. 4, 2002, and titled SHUTTLE-IN RECEIVER FOR RADIO-IMAGING UNDERGROUND GEOLOGIC STRUCTURES. The latter of which further claims priority by virtue of U.S. Provisional Patent Applications, serial No. 60/315,149, filed Aug. 27, 2001, and titled RADAR-NAVIGATION TOOL FOR MINING COAL, and serial No. 60/335,520, filed Oct. 31, 2001. Such Applications are incorporated herein by reference.

BACKGROUND OF THE INVENTION 0002 1. Field of the Invention 0003 The invention relates generally to directional drilling of boreholes in coal seams, and more specifically to steering mechanisms, radars and control systems for drilling horizontal boreholes in coal deposits. 0004 2. Description of the Prior Art 0005 The production of coal and methane depends upon the environment of the original coal bed deposit, and any subsequent alterations. During burial of the peat-coal swamp, sedimentation formed the sealing mudstone/shale layer overlying the coal bed. In deltaic deposits, high-energy paleochannels meandered from the main river channel. Oftentimes, the channels scoured through the sealing layer and into the coal seam. 0006 High porosity sandstone channels often fill with water. Under the paleochannel scour cut bank, water flows into the face and butt cleats of the coal bed. Subsequent alterations of the seam by differential compaction cause the dip, called a roll, to occur in the coal bed. Faults are pathways for water flow into the coal bed. 0007 Drilling into the coal bed underlying a paleochannel and subsequent fracking can enable significant flows of water to enter. The current state of the art in horizontal drilling uses gamma sensors in a measurements-while-drilling (MWD) navigation subsystem to determine when the drill approaches a sedimentary boundary rock. But if sandstone is protruding into the coal, such as results from ancient river bed cutting and filling, then the gamma sensor will not help. Sandstone does not have significant gamma emissions, so this type of detection is unreliable. Drilling within the seam cannot be maintained when the seam is not bounded by sealing rock. 0008 Methane diffusion into a de-gas hole improves whenever the drillhole keeps to the vertical center of the coal seam. It also improves when the drillhole is near a dry paleochannel. Current horizontal drilling technology can be improved by geologic sensing and controlling of the drilling horizon in a coal seam. 0009 One present inventor, Larry G. Stolarczyk, has described methods and equipment for imaging coal formations in geologic structures in many United States patents. Some of those patents are listed in Table I, and are incorporated herein by reference. 1TABLE IPatent No.IssuedTitle4577153Mar. 18, 1986Continuous Wave Medium Frequency SignalTransmission Survey Procedure For ImagingStructure In Coal Seams4691166Sep. 01, 1987Electromagnetic Instruments For ImagingStructure In Geologic Formations4742305May 03, 1988Method For Constructing Vertical Images OfAnomalies In Geological Formations4753484Jun. 28, 1988Method For Remote Control Of A Coal Shearer4777652Oct. 11, 1988Radio Communication Systems ForUnderground Mines4879755Nov. 07, 1989Medium Frequency Mine CommunicationSystem4968978Nov. 06, 1990Long Range Multiple Point Wireless ControlAnd Monitoring System4994747Feb 19, 1991Method And Apparatus For DetectingUnderground Electrically Conductive Objects5066917Nov. 19, 1991Long Feature Vertical Or HorizontalElectrical Conductor Detection MethodologyUsing Phase Coherent ElectromagneticInstrumentation5072172Dec. 10, 1991Method And Apparatus For Measuring TheThickness Of A Layer Of Geologic MaterialUsing A Microstrip Antenna5087099Feb. 11, 1992Long Range Multiple Point Wireless ControlAnd Monitoring System5093929Mar. 03, 1992Medium Frequency Mine CommunicationSystem5121971Jun. 16, 1992Method Of Measuring Uncut Coal RibThickness In A Mine5146611Sep. 08, 1992Mine Communication Cable And Method ForUse5181934Jan. 26, 1993Method For Automatically Adjusting TheCutting Drum Position Of A Resource CuttingMachine5188426Feb. 23, 1993Method For Controlling The Thickness Of ALayer Of Material In A Seam5260660Nov. 09, 1993Method For Calibrating A Downhole ReceiverUsed In Electromagnetic Instrumentation ForDetecting An Underground Conductor5268683Dec. 07, 1993Method Of Transmitting Data From ADrillhead5301082Apr. 05, 1994Current Limiter Circuit5408182Apr. 18, 1995Facility And Method For The Detection AndMonitoring Of Plumes Below A WasteContainment Site With RadiowaveTomography Scattering Methods5474261Dec. 12, 1995Ice Detection Apparatus For TransportationSafety5686841Nov. 11, 1997Apparatus And Method For The Detection AndMeasurement Of Liquid Water And Ice LayersOn The Surfaces Of Solid Materials5769503Jun. 23, 1998Method And Apparatus For A Rotating CuttingDrum Or Arm Mounted With Paired OppositeCircular Polarity Antennas And ResonantMicrostrip Patch Transceiver For MeasuringCoal, Trona And Potash Layers Forward, SideAnd Around A Continuous Mining MachineRE032563Dec. 15, 1987Continuous Wave Medium Frequency SignalTransmission Survey Procedure For ImagingStructure In Coal SeamsRE033458Nov. 27, 1990Method For Constructing Vertical Images OfAnomalies In Geological Formations

0010 There are a number of conventional ways directional drills use to steer in a desired direction. One involves placing the drill bit and its downhole motor at a slight offset angle from the main drillstring. The whole drillstring is then rotated to point the offset angle of the drill bit in the direction the operator wants the borehole to head. Another method involves an articulated joint or gimbal behind the drill bit and its downhole motor and using servo motors to angle the joint for the desired direction.

SUMMARY OF THE PRESENT INVENTION 0011 Briefly, a radar-plow drillstring steering embodiment of the present invention comprises a steering plow and a measurements-while-drilling instrument for mounting just behind the drill bit and downhole motor of a drill rod. The instrument includes a radar connected to upward-looking and downward-looking horn antennas and a dielectric-constant sensor. The steering plow includes four pressure pads radially distributed around the outside surface and their associated servo motors. A coordinated control of the pressure pads allows the steering plow to push the drillstring and drill bit up-down-left-right. The antennas and sensor are embedded in respective ones of the pressure pads and are used to electronically and non-invasively probe a coal seam to locate its upper and lower boundary layers. The dielectric-constant sensor provides corrective data for the up and down distance measurements. Such measurements and data are radio communicated to the surface for tomographic processing and user display. The radio communication uses the drillstring as a transmission line and F1/F2 repeaters can be placed along very long runs to maintain good instrument-to-surface communication. A docking mechanism associated with the instrument and its antenna array allows the instrument to be retrieved back inside the drillstring with a tether should the drill head become hopelessly jammed or locked into the earth. 0012 An advantage of the present invention is that a drillstring steering plow is provided for directional drilling. 0013 Another advantage of the present invention is that a drillstring steering plow is provided that keeps radar sensing antennas in intimate contact with the media. 0014 A further advantage of the present invention is a drillstring steering system is provided that can be self-guided and is relatively insensitive to groundwater. 0015 These and other objects and advantages of the present invention will no doubt become obvious to those of ordinary skill in the art after having read the following detailed description of the preferred embodiment which is illustrated in the various drawing figures.

IN THE DRAWINGS 0016 FIG. 1 is a diagram of a coal mine operation in an embodiment of the present invention that uses directional drilling and measurements-while-drilling radar to guide the drilling of horizontal boreholes within the vertical centers of coal deposits; 0017 FIG. 2 is a cross sectional diagram of a coal deposit, similar to FIG. 1, in which a sandstone paleochannel is protruding from the ceiling of the coal deposit, and in which the drilling of a horizontal borehole has dived down below the paleochannel to stay within the middle of the coal; 0018 FIG. 3 is a diagram of a radar-plow drillstring steering system embodiment of the present invention; 0019 FIG. 4 is a functional block diagram of a drillstring radar embodiment of the present invention, and shows the electronic components and their relationships as used in the equipment of FIGS. 1-3; 0020 FIGS. 5A-5D are cross-sectional diagrams of the plow steering part of the equipment illustrated in FIG. 3, and shows a variety of positions of the pressure pads for drilling straight ahead (FIG. 5A), drilling-up (FIG. 5B), drilling-down (FIG. 5C, and drilling-left (FIG. 5D); 0021 FIG. 6 is a schematic diagram of a ground-penetrating radar system embodiment of the present invention, and is one way to implement the radar electronics portion of the equipment illustrated in FIGS. 3 and 4; 0022 FIG. 7 is a graph of the typical driving-point real and imaginary impedance variations seen in the system of FIG. 1 for various thicknesses of a coal layer deposit; and 0023 FIG. 8 is a graph of the theoretical impedance response of a directional-coupler reflection port of FIG. 1.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS 0024 FIG. 1 illustrates a coal mine operation 100 in its earliest stages of development. An overburden 102 of soils covers a coal layer 104. This in turn overlies a deeper substrate layer 106. The coal layer 104 may be interrupted by a fault or scouring that happened over geologic time and was filled with sandstone, e.g., a paleochannel 108. At a minimum, the coal layer will undulate and vary in thickness. 0025 The sandstone in paleochannel 108 is porous and can be saturated with water. If so, flooding of the coal mine can occur from overhead if the coal layer just under the paleochannel is faulted or cut. It is therefore in the economic and safety interests of mining operations to avoid punching through such paleochannel 108 when taking the coal in coal layer 104. 0026 The coal layer 104 will typically contain valuable reserves of methane gas that can be pumped out through a network of horizontally running boreholes 110 and 112. These boreholes are begun as vertical bores 114 and 116 using a directional drilling method. 0027 FIG. 2 illustrates a horizontal drilling operation 200. An overburden 202 sits on top of a coal bed deposit represented by an upper coal deposit 204 and a lower coal deposit 206. These overlie a deeper substrate 208. The coal bed deposit is preferably drilled such that a horizontal borehole 210 is vertically equidistant between upper and lower coal deposits 204 and 206. In some situations, the horizontal borehole 210 may need to be lined with a perforated plastic pipe that lacks perforations in the water saturated areas. 0028 Such horizontal borehole 210 is drilled by a drillstring 212 that has a steerable, directional drillhead 214. A paleochannel 216 filled with sandstone and probably saturated with water represents a common hazard in such mining. FIG. 2 shows that the horizontal borehole 210 has been drilled to stay in the middle of whatever vertical space there is to work within the upper and lower coal deposits 204 and 206. Such guidance while drilling requires a non-invasive probing of the deposits ahead of the drilling that is provided by embodiments of the present invention. 0029 A natural waveguide occurs in layered coal seam sedimentary geology because there is a contrast between the electrical conductivity of the coal and that of shale, mudstone, and/or fire clay. The conductivity of coal is about 0.0005 S/m (2,000 ohm-meters). So dry coal is a pretty good insulator. The typical bounding layers have electrical conductivity in the range of 0.01 to 0.1 Siemens per meter (S/m). Such 20:1 conductivity contrast creates a natural waveguide, and any induced electromagnetic (EM) waves will travel along inside the coal bed layer. 0030 An electric field (EZ) component of the traveling EM wave is vertically polarized, and the magnetic field (Hy) component is horizontally polarized in the seam. The energy in this part of the EM wave travels laterally in the coal seam from a transmitter to a receiver, e.g., a Stolar Horizon Radio Imaging Method (RIM) receiver. A horizontally polarized electric field (EX) component has a zero value near the vertical center of the seam, and is maximum at the sedimentary rock-coal interface. It is this EX component that is responsible for EM wave signal transmission into the boundary rock layer. The energy in this part of the EM wave travels vertically in the coal deposit. 0031 In such waveguide, the coal seam radiowave magnitude diminishes over distance for two reasons. The first is the attenuation rate of the waveguide. The second is the radial spreading of wave energy as it travels away from the transmitter. The cylindrical spread can be mathematically represented by 1

\frac{1}{\sqrt{r}},

0032 where r is the distance from the transmitting to receiving antenna. This compares with a non-waveguide far field spherically spreading factor of 2

\frac{1}{r}

0033 Thus, at hundred meters, the magnitude of the seam EM wave decreases by a factor of only ten in the waveguide, and by a factor of hundred in an unbounded media. 0034 A significant advantage of the seam waveguide is signals can travel farther. Another advantage is that the traveling electromagnetic wave predominantly remains within the coal seam waveguide. 0035 Such coal seam electromagnetic wave is very sensitive to changes in the waveguide geology. The radiowave attenuation rate and the phase shift can be measured and the measurements will reveal structural features of the coal seam. If the waveguide were uniform, then the path would be a straight line. The straight line path is an assumption used in the Algebraic Reconstruction Technique (ART) tomography algorithm. But radiowaves are refracted near significant geologic anomalies causing the travel path of the radiowave to bend and be longer than in the uniform waveguide case. Such bending cannot be accounted for in ART processing and so causes an error. But, by measuring the total path phase shift, the bending effect can be accounted for in Full Wave Inversion Code (FWIC), a tomography reconstruction algorithm. The waveguide attenuation effect reduces the magnitude of the electromagnetic wave along the path. 0036 It has been observed that the coal seam attenuation rate will increase with signal frequency. The wavelength increases as frequency decreases. By lowering the operating frequency, the measurement range increases. But such range still depends on coal seam thickness, the type of bounding sedimentary rock, and the electrical conductivity of the coal. 0037 Intrusions of sandstone sedimentary rock into the coal seam increase the attenuation rate at that point. This is because more of the signal travels vertically into the boundary rock and leaks from the waveguide. If water has invaded the coal, then any wet clay in the coal will cause the electrical conductivity to decrease and the attenuation-rate/phase-shift to increase. Such effects allow tomography to be used to image such geologic anomalies. 0038 The attenuation-rate/phase-shift rapidly increases with decreasing seam height, so coal seam thinning can be easily detected. Higher attenuation-rate areas suggest that the coal seam boundary rock is changing, the seam is rapidly thinning, or/and water has invaded the coal seam. 0039 Faults and dykes in the coal seam will cause signal reflections within the waveguide. Such reflections can appear as excess path loss, but can be differentiated as being caused by faults and dykes in the coal by also taking phase shift measurements. 0040 The preferred operating frequency band extends from a few kilohertz to well over 300 kHz. The few kilohertz lower limit is due to the practical problems in building low frequency antennas with their longer wavelengths, and the high frequency limit is due to the increased coal seam waveguide attenuation rate with frequency. Instrument design and tomographic image processing algorithms can greatly increase coal seam imaging resolution. 0041 Tomographic ART images often have in the direction of the signal wave propagation. The ART algorithm assumes that the coal seam waveguide signal travels on a straight ray path. Severe geologic anomalies can cause signal refraction, in which case the ray-path assumption fails. 0042 Referring again to FIG. 2, the drillstring 212 preferably includes a radar system with both upward and downward looking antennas to non-invasively determine the vertical middle of coal layers 204 and 206. Such further preferably includes a dielectric-constant measuring sensor and an inductive radio for communicating measured data to the surface. 0043 For example, the drillstring 212 includes at its distal end behind drillhead 214 a measure-while-drilling (MWD) instrument with a stepped-frequency radar operating in the 1.7 to 2.5-GHz range. Typically, fifty frequency steps are used to cover the frequency range. At each frequency step, in-phase (I) and quadrature (Q) radar return signal values are measured and transmitted to a user display at the ground surface. The I and Q values for all step frequencies are decoded and processed in a Fast Fourier Transform (FFT) to derive the time domain response from each of the roof and floor to the coal interface. The distances to each interface can then be determined from these measurements by taking into account the relative dielectric constant. 0044 Alternatively, since the object of steering the drillstring 212 is to stay in the vertical middle of the coal seam 204 and 206, the raw measurements from the upward-looking and downward-looking radar antennas can be compared in their raw uncorrected state to see if they are about the same. If so, it can be assumed that the measurements have occurred at the vertical middle, e.g., equidistant to ceiling and floor. The corrective information for each from the dielectric-constant sensor can in such case be dispensed with. 0045 Accurate geologic data can be opportunistically acquired while drilling through such coal seams. But to do this, the radar-acquired distance-data must be corrected for relative dielectric constant (r) to determine the actual physical distance. In such case, the instrumentation must include a local dielectric-constant measuring capability. 0046 The FFT computations needed for real-time MWD is very computer-intensive. With the present state-of-the-art in microprocessors, the measured data must be processed above ground where larger, more capable computers can be operated. Such computers also demand operating power that is too much for battery operation in the drillstring 212. Therefore the measurements are communicated along the drillstring, e.g., using inductively coupled radio communication. 0047 Such upward-looking and downward-looking radar can also determine the type of boundary rock in the roof and floor. Regions of high coal seam where the relative dielectric constant is high, e.g., greater than 6.0, are unattractive because they produce greater amounts of water in the methane drainage system. Any coal deposit set under a paleochannel and having a dielectric constant under 6.0 suggests that the paleochannel is dry and will contain methane. Drilling near dry paleochannels increases methane production. 0048 Conventional horizontal drilling navigation systems cannot directly determine seam thickness or changes in seam orientation, dips and rolls, without trial-and-error exploratory drilling of the floor and the roof in the same region of the panel. Thus, when the seam horizon changes, the drill will impact the roof or floor. A drilling machine operator with such conventional methods detects when the drill is on the roof or floor horizon by evidence of rock in the cuttings. The drilling technician redirects the drill motor to try to keep the drill within the seam. The borehole that results wanders between the roof and floor of the coal seam along its path. Punching through to roof or floor interfaces can invite ground water to inundate the borehole. 0049 FIG. 3 illustrates a radar-plow drillstring steering system embodiment of the present invention, and is referred to herein by the general reference numeral 300. The radar-plow drillstring steering system 300 comprises a drill bit 302 on a distal end, a steering plow 304, an electronics section 306 with retrieval docking, a repeater section 308, and a drillstring section 310. The steering plow 304 includes a set of four radially distributed pressure pads for up-down-left-right drilling control. Three of these pressure pads are visible in FIG. 3, e.g., a top pressure pad 312, a bottom pressure pad 314, and a side pressure pad 316. These are all controlled to assume various states of protrusion or retraction by a plow control 318. A measurements-while-drilling (MWD) instrument 320 processes radar signals to-and-from horn radar antennas and dielectric-constant sensors embedded in respective ones of the four radially distributed pressure pads. The MWD instrument 320 provides estimates in real-time of the distances to the boundary layers in the floor and ceiling of a coal seam. These estimates are used by the plow control to maintain a desired course through the coal seam. A tether 322 is used to retrieve MWD instrument 320 from within the drillstring back to the surface if the drill bit 302 jams and cannot be retracted. A signal repeater 324 assists in long range communication between the MWD instrument 320 and an operator display and tomographic processor on the surface. A second tether 326 is used to retrieve the signal repeater 324. 0050 FIG. 4 illustrates a drillstring radar system embodiment of the present invention, and is referred to herein by the general reference numeral 400. The drillstring radar 400 comprises a downhole drillstring 402 connected to a surface collar 404. A graphical user interface (GUI) 406 is positioned at the surface and processes tomographic images of the measurements-while-drilling on a user display 407. A microcomputer (CPU) 408 does data logging and processes raw data received by a transceiver 409 from the downhole equipment. A signal coupler 410 allows the transceiver 409 to use the collar 404 and drillstring 402 as a common transmission line. A radio repeater 411 repeats and amplifies F1/F2 signals through its couplers 412 and 414. If the drillstring 402 is relatively short, and signal attenuation is not severe, then repeater 411 may not be needed. A measurements-while-drilling (MWD) instrument 416 is connected to the radio signals on the drillstring 402 by a coupler 418. Coupler 418 is operated at about 100-kHz and typical comprises 14-turns of Litz-wire on a 12-inch long coil one inch wide and mounted edge onto the drill rod. 0051 The MWD instrument 416 comprises a radio transceiver 422 that is connected to an antenna 424. Such antenna 424 is operated at about 125-kHz and typical comprises 14-turns of Litz-wire on a 24-inch long ferrite rod one inch in diameter. It communicates with other radar imaging equipment collocated in another parallel borehole. A more complete description of how this antenna 424 is used and how associated radar imaging equipment located in associated parallel boreholes can be used to an advantage in coal deposit imaging, is provided in other United States patent applications of the present inventor, e.g., Ser. No. 10/259,912, filed Sep. 30, 2002, and titled, RADIO-IMAGING OF UNDERGROUND STRUCTURES. Such are incorporated herein by reference. 0052 A processor (PIC) 426 interfaces the raw measurements from a radar electronics system 428 to the transceiver 422. A switching matrix 430 allows the selection of an upward-looking radar horn antenna 432, a resonant microstrip patch antenna (RMPA) 434, and a downward-looking radar horn antenna 436. 0053 Such horn antennas are respectively embedded in the pressure pads associated with the steering plow. Such are preferably operated in the 2.0-2.5 GHz frequency band and have at least 20-dB of back-lobe rejection. A plow controller 438 provides control signals in real-time to guide the direction of drilling according to measurements obtained by the radar electronics 428. 0054 Electrical power for the MWD instrument 416 can be provided by a hydro-generator 440 that taps into the hydraulic fluid flow through the drillstring, or by a rechargeable battery 442, or both. Using both would allow data to continue to be collected and reported even though the hydraulic flow may have been stopped for some reason. 0055 The MWD instrument 416 is preferably retractable should the drillstring 402 become stuck in the drillhole. For example, the MWD instrument is equipped with a self-docking mechanism to interconnect with the antenna array subsection drill rod. A retractable capability permits the MWD instrument and repeater to be hydraulically pumped-in and pulled out of the drillstring. Such a shuttle mechanism is described by the present inventor in U.S. patent application, Ser. No. 10/161,378, filed Jun. 04, 2002, and titled SHUTTLE-IN RECEIVER FOR RADIO-IMAGING UNDERGROUND GEOLOGIC STRUCTURES. Alternatively, a docking mechanism associated with the MWD instrument 416 and its antenna array 420 allows the MWD instrument to be retrieved back inside the drillstring with a tether should the drill head become hopelessly jammed or locked into the earth. 0056 The antenna array 420 is preferably designed to accommodate the radar and dielectric microwave antennas, the radar transmitter antenna, and the data transmission antenna. For economy, the radar transmitter function can use the transmitter section of the data transceiver 422 during periods that data is not being sent to the surface collar. The MWD instrument is placed inside a beryllium-copper drill rod section, e.g., ten feet long. In one instance, the antenna array and MWD instrument was located fifteen feet behind the downhole motor. 0057 In general, embodiments of the present invention include methods for measuring a reflection port signal for a three-port directional coupler with its output port directly connected to the driving-point of a conventional horn antenna. The reflection port is connected through a first balanced mixer. A 0/90 phase shifter rotates the signal phase between a forward coupling port of the three-port directional coupler and a second beat frequency balanced mixer. A first beat frequency balanced mixer injection signal is phase coherent with the reflection signal, thereby canceling the phase shift of the PLL. The impedance of the sensor antenna in close proximity to natural media is made as close as possible to the characteristic impedance of the three-port directional coupler, preferably fifty ohms. The beat frequency occurs with phase-coherent mixing in mixer M4 so as to cancel the phase drift of the crystal oscillator and create in-phase (I) and quadrature (Q) signals derived from the reflection port signal. Such can be used in remote sensing and imaging of non-metallic and metallic landmines. 0058 One method of calibrating the sensor antenna uses sets of polynomial equations wherein the independent variable represents the value to be measured. Alternative embodiments include solving for the value of the independent variable (H) from a set of polynomial equations that is the difference between the calibration polynomial equation and the measured values of I and Q. Alternative embodiments include determining the value of the independent variable (Ho) by independent means and using the difference between the measured and calibration polynomial value to detect and image the object. Other embodiments determine the independent variable as being related to a material thickness, dielectric constant, in-situ stress. 0059 The measurements obtained from embodiments of the present invention can be tomographically processed to construct two and three-dimensional images of the object, e.g., a silhouette image of the object and its associated signal-to-noise ratio. The signal-to-noise ratio and the width of the reconstructed object are generally related to the probability of the object's detection. 0060 FIGS. 5A-5D illustrate a cross section of a drillstring steering plow embodiment of the present invention, and such is referred herein by the general reference numeral 500. The drillstring steering plow 500 comprises a set of orthogonally distributed pressure pads 501-504. An upward-looking radar horn antenna is embedded within pressure pad 501. A downward-looking radar horn antenna is embedded within pressure pad 502. A resonant microstrip patch antenna (RMPA) operated as a dielectric-constant sensor is embedded in either of side pressure pads 503 and 504, or both. A corresponding set of servo motors 505-508 are used to adjust the relative protrudings of pressure pads 501-504. These pressure pads and servo motors can be coupled, for example, by jack screws 509 and 510. Alternatively, hydraulic pumps and pistons can be used to forcefully position each of the pressure pads 501-504. 0061 FIG. 5A shows the pressure pads 501-504 in their neutral position for straight ahead steering of the drillstring. FIG. 5B shows pressure pad 501 retracted and pressure pad 502 extended for upward steering of the drillstring. Such causes the drill bit to bite harder on the roof of the cut. FIG. 5C shows pressure pad 501 extended and pressure pad 502 retracted for downward steering of the drillstring. Such causes the drill bit to bite harder on the floor of the cut. FIG. 5D shows pressure pad 504 extended and pressure pad 503 retracted for steering the drillstring to the left, assuming the view here is from behind. Such causes the drill bit to bite harder on the left side of the cut. In all such instances, the antennas and sensors maintain a very intimate contact with the coal seam and squeeze out water in the gaps that otherwise would interfere with imaging and sensing. 0062 An RMPA antenna sensor is used to determine a relative dielectric constant (C) of any coal surrounding the drillstring rod radar steering plow 500. The dielectric constant of coal is affected by how much moisture is in the coal. Upward and downward-looking radar horns 518 and 520 each provide I and Q signals at each stepped frequency that are interpreted by the microwave radar measurement circuit 514. The distance through the coal layer to a sedimentary interface can be determined from the radar data once the dielectric constant of the natural media has been determined. 0063 FIG. 6 illustrates a ground-penetrating radar system embodiment of the present invention, which is referred to herein by the general reference numeral 600. Such could be used in MWD instrument 320 in FIG. 3 or radar electronics 428 in FIG. 4. The system 600 is used to non-invasively penetrate a coal seam 601 with microwave radio energy to find a boundary rock interface 602. The system 600 includes a local oscillator 603 that produces a reference frequency (FX) with reference phase (1) A first phase-locked loop (PLL) 604 synthesizes a radio frequency F0, an integer harmonic of FX. The radio frequency F0 is passed to a forward-coupling port of a power splitter 606. An output port is connected to a wideband isolation amplifier 608. A three-port directional coupler 610 one-way couples the transmit signal out to a directional radar antenna 612 which illuminates the coal seam 601 and produces reflected waves from the boundary rock interface 602 with the coal seam 601. A reflection port of the three-port directional coupler 610 is used for measurements, e.g., where eReO cos R, the reflected energy is a function of the output energy of antenna 612. 0064 During operation, such directional radar antenna 612 is moved close to a natural media or ground surface. A deliberate impedance mismatch is thus created that will result in higher levels of coupler reflection port output voltage or standing wave ratio (VSWR). Any impedance mismatches appearing at any of the ports on a directional coupler will reduce its directivity and isolation between ports. Wideband isolation amplifiers are used to stop reflected waves from reaching the transmitter stages and causing impedance mismatches. The balanced mixers are in particular susceptible to performance degradations under uncontrolled conditions. 0065 Radar signals radiated from the directional radar antenna 612 enter the natural media or ground and are reflected back attenuated and with a change of phase. A wideband isolation amplifier 614 forwards the reflected-wave sample to a first balanced mixer (M1) 616. 0066 A second PLL 618 synthesizes a coherent frequency F1, e.g., 10.70 MHz. A sample of the transmitted signal is provided by the power splitter 606 and a second wideband isolation amplifier 619 to a selectable 0/90 phase shifter 620. A balanced mixer (M2) 622 provides an intermediate frequency (IF) F5 that is output by a third wideband isolation amplifier 624. The result is a suppressed-carrier signal with upper and lower sidebands offset from the carrier frequency by 10.70 MHz. This in turn mixes with RF in balanced mixer (M1) 616 to produce a first intermediate frequency (IF) F6. A bandpass filter 626 produces an output F6. A third PLL 628 synthesizes another coherent frequency (F2), e.g., 10.720 MHz. This is combined with F6 in a balanced mixer 630 to produce a relatively low-frequency IF signal F7, e.g., 20.0 KHz. 0067 Both the oscillator phase shift 1 and the frequency are multiplied by No to create the output frequency No1. The phase shift No1, is canceled on mixer M1. The coupler reflection port voltage eR dependence on load plane impedance is mathematically represented by Equation (3). The reflection coefficient (F) is a complex number that can be represented as a vector magnitude with phase angle R. Trigonometric identities are used and filter theory is applied to find that,

F8I1k cos R

and

F8O1k sin R(6)

0068 The ratio of these direct current (DC) values and the inverse tangent values solves for the reflection phase angle R. The constant k is determinable after instrument calibration. The magnitude of the reflection coefficient can be found from either of the above two equations. The measured values of and R can be applied in Equation (3) to determine the load plane impedance. 0069 A bandpass filter 632 produces a signal F7. A fourth PLL 633 synthesizes another radio frequency (F3). The radio frequency is applied to the LO part of the fourth balanced mixer (M4) 634. Heterodyning produces a signal output F8. An integrating filter 636 processes and outputs a signal F8. The phase drift (1) in local oscillator 603 is automatically cancelled and does not appear in the output. An analog to digital converter (ADC) 638 outputs a digital signal format, e.g., for further processing by a computer. 0070 A control line 640 selects a 0 or 90 phase-shift through the phase shifter, and this will cause the system 600 to output in-phase or quadrature measurements, e.g., as represented by a DC output voltage of the ADC 638. This control line is typically connected to the same microprocessor that receives the digitized estimates from ADC 638. Such arrangement provides time-multiplexed I and Q amplitude estimates that are indexed to a calibration table, like FIG. 7, to find the depth to the boundary rock interface 602 or the thickness of the coal seam 601 covering the object. The microprocessor is preferably further provided with control signals 641-644 so that the F0-F3 frequencies can be digitally manipulated, e.g., for best transmission penetration and receiver sensitivity on-the-fly as different kinds of media 601 and objects 602 are inspected. 0071 The estimates can be processed to provide landmine detection, guide drilling or excavation operations, control mining equipment, find lost buried objects in the soil, etc. 0072 The three-port directional coupler 610 is preferably located near the RMPA antenna 612 and is connected to it by a short coaxial cable or strip line. The RMPA antenna 612 is preferably placed close to a natural media surface so the driving-point impedance of an antenna can be adjusted to match the characteristic impedance of the coupler. This technique will maximize the sensitivity of the coupler reflection port signal to small changes in antenna driving-point impedance. An antenna driving-point impedance adjustment is therefore preferred, e.g., with a variable slot built into the antenna structure capacitor. 0073 The balanced mixer placed between the coupler reflected port and the first balanced mixer is such that the signals that modulate the coupler reflection output are phase coherent with the transmit signal. Sidebands are thus produced that represent the phase and amplitude information in the reflection signals. The rest of the circuitry demodulates the information from the carrier. 0074 The output signal of the second balanced modulator (M2) is mixed with a coherent sample of the transmit signal in the first balanced mixer. After filtering, such signal will faithfully replicate the reflection port signal. This simplifies the transceiver design and enables accurate signal measurements with low cost synchronous detection circuits. 0075 FIG. 7 charts the relationships that typically develop between the RMPA antenna and the depth to the object and measured values for real (in-phase, I) and imaginary (quadrature-phase, Q). The two vector components of RMPA impedance, real and imaginary, vary differently as a nearby media layer thickness changes. Plotting the imaginary on the Y-axis and the real on the X-axis of a graph yields a calibration curve 700 that spirals to a vanishing point with increasing layer thickness, e.g., from 0.25 to twelve inches. 0076 FIG. 8 illustrates the theoretical impedance response 800, e.g., as seen at the reflected-wave sample port of directional coupler 610. The applied signal (eo) from isolation amplifier 608 is assumed to be constant. Detection sensitivity is best on the steepest parts of the curve. Sensitivity is maximum when the RMPA driving point impedance matches the characteristic impedance of the directional coupler. 0077 A method embodiment of the present invention provides for the calibration of such RMPA sensor antennas. Sets of polynomial equations are constructed with independent variables (H) that allow any given antenna driving point impedance value to be measured, and represented as a calibration function with independent variable (H). The difference between the measured and calibration polynomial value is used to detect and image the coal seam. 0078 Method embodiments of the present invention can further define such independent variables to be thickness, dielectric constant, or in-situ stress of a geologic media or buried object. The information gathered can be used to construct visual images of the coal seam and its composition. 0079 Referring again to FIG. 6, the crystal oscillator phase shift 1 is multiplied by the full multiplication factor No and the frequency of the crystal oscillator is multiplied by No creating the output frequency No1. After filtering, the phase shift No1 is canceled on balanced mixer M1. The coupler reflection port voltage eR dependence on driving point impedance is mathematically represented by Equation (3). The reflection coefficient () is a complex number that can be represented as a vector magnitude with phase angle R. Using trigonometric identities and applying filter theory, it can be shown that,

F8Ik cos R

and

F8Qk sin R(12)

0080 The ratio of these direct current (DC) values can be used to solve for the inverse tangent values. And the reflection phase angle R can be readily determined. Calibration of the instrument will determine the correct constant K. The magnitude of the reflection coefficient can be found from either of the above two equations. The circuits isolate the phase shifter from any blaring reflected signal. The measured values of and R can be applied in Equation (3) to determine the driving point impedance. 0081 The following equations help further describe the necessary working conditions of the system 600 in FIG. 6, and illustrate what each functional element must do to process the signals involved. FR is the reflected-wave output from the reverse port of directional coupler 610. FX is the output of oscillator 603. F0 is the output of PLL 604. F1 is the output of PLL 618. F2 is the output of PLL 628. F3 is the output of PLL 633. F4I and F4Q are the selected in-phase and quadrature outputs of phase shifter 620. F5 is the output of mixer 622. F6 is the output of mixer 616. F6 is the filtered output of bandpass filter 626. F7 is the output of mixer 630. F7 is the filtered output of bandpass filter 632. F8 is the output of mixer 634. F8 is the filtered output of bandpass filter 636.

FXA cos (1t1)

F0B cos (N01tN01)

F1C cos (N11tN11)

F2C cos (N21tN21)

F41D cos (N01tN01)

F4QD sin (N01tN01)

FRE cos (No1tN01R)

N2N11

0082 Placing the quadrature hybrid into the 0 position yields: 3

F5I = F1 F4I = C D \cos (N_{1}_{1} t + N_{1}_{1}) \cos (N_{0}_{1} t + N_{0}_{1}) = \frac{C D}{2} [\cos [(N_{1} + N_{0})_{1} t + (N_{1} + N_{0})_{1}] + \cos [(N_{1} + N_{0})_{1} t + (N_{1} - N_{0})_{1}]]

F6I = F5 F R = \frac{CDE}{2} [\begin{matrix} \cos & [(N_{1} + N_{0})_{1} t + (N_{1} + N_{0})_{1}] \\ \cos & (N_{0}_{1} t + N_{0}_{1} +_{R}) + \\ \cos & [(N_{1} - N_{0})_{1} t + (N_{1} - N_{0})_{1}] \\ \cos & (N_{0}_{1} t + N_{0}_{1} t +_{R}) \end{matrix}] = \frac{CDE}{2} [\begin{matrix} \cos & [(N_{1} + 2 N_{0})_{1} t + (N_{1} + 2 N_{0})_{1} +_{R})] + \\ \cos & (N_{1}_{1} t + N_{1}_{1} -_{R}) + \\ \cos & (N_{1}_{1} t + N_{1}_{1} -_{R}) + \\ \cos & [(N_{1} + 2 N_{0})_{1} t + (N_{1} + 2 N_{0})_{1} +_{R})] \end{matrix}]

0083 After filtering: 4

\begin{matrix} F^{1} 6 I = \frac{CDE}{4} [\cos (N_{1}_{1} t + N_{1}_{1} -_{R}) + \cos (N_{1}_{1} t + N_{1}_{1} +_{R})] \\ = \frac{CDE}{2} \cos (_{R}) \cos (N_{1}_{1} t + N_{1}_{1}) \\ F7I = F2 F^{1} I = \frac{C^{2} DE}{4} \cos (_{R}) [\cos (N_{2}_{1} t + N_{2}_{1}) \cos (N_{1}_{1} t + N_{1}_{1})] \\ = \frac{C^{2} DE}{4} \cos (_{2}) [\begin{matrix} \cos & [(N_{2} + N_{1})_{1} t + (N_{2} + N_{1})_{1}] + \\ \cos & [(N_{2} - N_{1})_{1} t + (N_{2} - N_{1})_{1}] \end{matrix}] \end{matrix}

0084 After filtering: 5

F^{1} 7 I = \frac{C^{2} DE}{4} \cos (_{R}) \cos [(N_{2} - N_{1})_{1} t + (N_{2} - N_{1})_{1}]

0085 where N2N11, 6

\begin{matrix} F^{1} 7 I = \frac{C^{2} DE}{4} \cos (_{R}) \cos (_{1} t +_{1}) \\ F8I = F X F^{1} 7 I = \frac{{AC}^{2} DE}{4} \cos (_{R}) \cos (_{1} t +_{1}) \cos (_{1} t +_{1}) \end{matrix}

0086 and applying filter: 7

F^{1} 8 I = \frac{{AC}^{2} DE}{4} \cos (_{R})

0087 Switching the quadrature hybrid to the 90 position and following similar techniques as shown above, reveals that: 8

F^{1} 8 Q = \frac{{AC}^{2} DE}{4} \sin (_{R})

0088 Let IF18Q and QF18Q and thus,

{square root}{overscore (I2Q2)}

and,

_{R} = \tan^{- 1} \frac{Q}{I}

0089 Although the present invention has been described in terms of the presently preferred embodiments, it is to be understood that the disclosure is not to be interpreted as limiting. Various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure. Accordingly, it is intended that the appended claims be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention.