GB2288024A - Diode-type NMR RF transmitter decoupling circuit - Google Patents

Abstract

The junction capacitance of inverse parallel decoupling diodes 17, 18 in an NMR RF transmitter circuit 21, 16 is reduced by applying reverse bias by means of batteries (19, 20 in figure 3) or control potentials 44, 45. The diodes may be switched by FET 24, 25 or bipolar transistors. At the end of a transmitted pulse, coil energy is rapidly dissipated in the transistors under the control of clamping Zener diodes 26, 27. Resistors may be placed in parallel with the transistors to assist in coil energy dissipation. The circuit may be used in an NMR well-logging tool. <IMAGE>

Description

OSCILLATING MAGNETIC FIELD GENERATING ASSEMBLY This invention relates to an oscillating magnetic field generating assembly.

Magnet assemblies with which the oscillating magnetic field generating assembly may be used have been developed in recent years for inspecting bore holes for oil bearing rock. A typical example of such an assembly is described in US-A-4350955 in which a pair of permanent magnets are provided spaced apart with their north-south axes aligned and with like poles facing one another. This arrangement has the advantage that the oscillating magnetic field generating assembly which has to be provided can be in the form of a RF coil positioned in the space between the magnets. More recently, a modified form of this arrangement has been described in W092/07279. In both examples the magnet assembly generates a magnetic field in a working region externally of the magnet assembly having a uniformity suitable for performing an NMR experiment.

One of the problems with these arrangements which has been discovered in use is that there is a significant variation in the magnetic field generated by the oscillating magnetic field generating assembly (conventionally known as the B1-field) over the working region. This becomes particularly significant when the radial extent of the working region is increased. This has two effects on the NMR measurement.

The first effect is the non-uniform excitation of the spins. This means that in a typical spin-echo measurement 900, 1800 etc pulses are not defined, and so optimum excitation cannot be obtained over the whole volume. This both reduces the signal strength and confuses the interpretation.

The second effect is that the signal which is received by the RF coil is weighted in favour of the nearer parts of the sensitive volume. This again confuses the interpretation of the data.

Both of these untoward effects can in principle be overcome by complex pulse sequences, but these are difficult to design and implement and may restrict the range of measurements which can be made. The solution to these problems is to design the assembly so that the oscillating magnetic field generated is sufficiently uniform within the working volume. This might involve having separate transmitter and receiver coils, the former being designed to produce a uniform B1 field over the volume of interest and the latter being optimised for sensitivity to the NMR signal.

However, a further problem is encountered where the receiver (Rx) and transmitter (Tx) coils of the assembly couple strongly, preventing the assembly from working. The solution to coupling of the transmitter with the receiver is that a second receiver coil is wound and connected in opposition to the main receiver coil, via crossed diodes as described in our copending, unpublished application no.

PCT/GB93/02370.

For large signals (i.e. those arising from the Tx coil) the receiver coil appears non-inductive because current flows in the second receiver coil. For small signals (viz NMR) only the first of the pair of coils in the receiver is effective and the Rx coil behaves as normal. This therefore prevents the transmitter coupling with the receiver.

To prevent the receiver coupling with the transmitter a pair of crossed diodes are provided in series with the transmitter coil. These solutions are dealt with in our copending, unpublished application No. PCT/GB93/02370.

However, in practice the crossed diode arrangement in the transmitter circuit only works when relatively small diodes (rated for a few amps of current) are used in the transmitter coil circuit. Large diodes, which are required to cope with the currents produced in the transmitter coil circuit by a transmitted pulse, have sufficient capacitance to behave as a short circuit to the RF signal induced from the receiver coil thereby negating the effect of the diodes in preventing coupling of the receiver with the transmitter.

In accordance with the present invention, an oscillating magnetic field generating assembly comprises a transmitter coil connected to an oscillating electrical source, wherein the magnetic field generated by the oscillating magnetic field generating assembly is sufficiently uniform within a working region for use in an NMR experiment; the assembly further comprising a pair of diodes connected in parallel in opposite senses between a capacitor and the transmitter coil; and biassing means connected to each diode operable to reverse bias each of the pair of diodes.

Although the diodes themselves have too much junction capacitance to operate in the assembly, by reverse biassing them the capacitance is reduced to an acceptable value at which the diodes do not act as a short circuit to an RF signal induced from the second coil and also substantially prevent the receiver coil from coupling with the transmitter coil.

Preferably, the assembly further comprises control means for selectively controlling the biassing means.

In one example, the biassing means comprises one or more batteries, such as dry or wet cells, or other forms of battery.

The biassing means of the assembly are not limited to use in the NMR application described. The use of reverse bias in this way to reduce diode capacitance is equally applicable to any magnet, RF or other application in which a high-current switch is required to have a low capacitance in the off state, or any NMR application where separate transmitter and receiver coils are used, which might couple to some extent.

A further problem encountered in a conventional circuit is that there is a delay between the end of a transmit pulse and a time when the receiver may be used because energy stored in the transmitter coil continues to circulate in the circuit for some time. The larger the delay, the more limited the materials which can be measured since the delay should be less than the relaxation time constants, T1 and T2, of the materials for measurements to be taken.

Preferably, therefore, the biassing means comprises a semiconductor switch in series with each one of the pair of diodes; and, wherein the assembly further comprises respective voltage limiters for limiting voltage such that energy in the transmitter coil is dissipated in each of the voltage limiters. Alternatively, energy in the transmitter coil may be dissipated in the semiconductor switch.

This arrangement reduces the delay before a receiver coil can be used following signal transmission by the transmitter coil as well as removing the requirement to reverse bias the diodes with batteries, which might become discharged with time.

Typically, the voltage limiters comprise a pair of reverse connected zener diodes. In one example, the energy in the transmitter coil is dissipated in both the semiconductor switch and the zener diodes.

In accordance with a second aspect of the present invention, a magnet assembly which generates a magnetic field in a working region externally of the magnet assembly having a uniformity suitable for performing an NMR experiment incorporates an oscillating magnetic field generating assembly according to the first aspect of the invention.

In accordance with a third aspect of the present invention, we provide a method of operating a magnet assembly according to the second aspect of the invention incorporating an assembly according to the first aspect of the invention.

Examples of an oscillating magnetic field generating assembly in accordance with the present invention will now be described with reference to the accompanying drawings, and compared with a conventional assembly in which: Figure 1 illustrates a conventional transmitter coil circuit adapted to reduce coupling between transmitter and receiver coils; Figure 2 illustrates a conventional receiver coil circuit adapted to reduce coupling between transmitter and receiver coils; Figure 3 illustrates an example of part of an oscillating magnetic field generating assembly according to the present invention; Figure 4 illustrates an example of an alternative arrangement of a part of an oscillating magnetic field generating assembly according to the present invention; and, Figure 5 is a schematic block diagram of a magnet assembly, in which the present invention may be included, inserted in a bore hole.

Figure 1 illustrates an electrical resonant circuit including a transmitter coil 1 and a capacitor 5. The transmitter coil 1 is connected in series with a control device 6 consisting of a pair of diodes 7,8 connected in parallel and in opposite senses.

The characteristic of the control device 6 is such that current will only flow in the resonant circuit if the voltage across the control device 6 exceeds about 0.5 volts. This prevents noise voltages induced or arising within the transmitter coil 1 from causing a circulating current which would couple with the receiver coil by electromagnetic induction. The resonant circuit is connected to a conventional RF source (not shown) via a matching capacitor 9.

In this example, a separate receiver coil is used although the disadvantage of this is that some way has to be found of decoupling it from the transmitter coil. The classical method of physically orthogonal coils is not available for cylindrical symmetry. Therefore, a double receiver coil is used, made up of coils 10,12 connected in series opposition via crossed diodes as shown in Figure 2.

The coil 12 is a conventional receiver coil which is connected in a resonant circuit with a variable capacitor 11, the resonant circuit being connected via a matching capacitor 15 to conventional receiving electronics (not shown). In addition, the resonant circuit is connected to the second receiver coil 10 having substantially the same dimensions and number of turns as the receiver coil 12 but electrically connected in opposition and via two pairs of parallel Schottky diodes 13,14. With this arrangement, at low signal levels (received) no current can flow in the coil 10 and the coil 12 acts as a conventional receiver coil, but at high levels (for example during a transmit sequence) current can flow in the coil 10 and the circuit is essentially non-inductive, so the net coupling to the transmitter coil is almost zero.

In one example, each coil 10,12 comprises twelve turns in two layers (six turns per layer), the two coils being interleaved. The conductor would be 2mm x lmm rectangular copper wire, with a 3mm winding pitch and lmm inter-layer gap.

Typical values for the circuits of Figures 1 and 2 are Tx coil inductance 22.6H; maximum Tx coil current to produce 3 gauss B1 at sensitive volume is 60A peak, 40A rms. The duty cycle has a maximum value of , short times only, and is typically 1/20 for example with a pulse length of 60 Asec and frequency of 440 kHz. The Tx coil Q is 80 when tuned with 5500pF and matched to 50 ohms. The Rx coil inductance is ll H and the Rx coil Q is 70 - 80 tuned with 12800pF and matched to 50 ohms.

Figure 3 illustrates a first example of the present invention. A transmitter coil 16 is connected in series with a parallel circuit arrangement 23. The parallel circuit arrangement consists of a pair of diodes 17,18, each connected in opposite senses in series with respective dc sources 19,20 also connected in opposite senses.

Additional means 24,25 are provided for dissipating coil energy and/or preventing overcharging of the batteries. A capacitor 21 is connected in series between the coil 16 and one end of the parallel arrangement 23, and a resistor 28 is connected to ground between the other end of the parallel arrangement 23 and the coil 16. A typical value for the resistor 28 is 2K2n with the diodes rated at 40 amps. The diodes shown in Figure 3 are reverse biassed by their respective dc sources 19,20. The coil is connected to an oscillating source (not shown) via junction 43, typically through a coupling capacitor.

Although 40A rated diodes by themselves have too much junction capacitance to operate properly, reverse biassing them in the circuit shown in Figure 3 reduces the capacitance to an acceptable value. This is because the capacitance of the depletion layer in a reverse biassed diode varies according to:

Where Vf is the forward voltage of the diode (typically 0.5V) Vr is the applied reverse bias, and CO is the capacitance when no voltage is applied (see for example "Semiconductors and electronic devices", Bar-Lev, Prentice Hall 1984, pp144 et seq). By applying a reverse bias of 50V, for example, the capacitance can be reduced by a factor of 10, from say 1000pF to 100pF for a 40A rated diode.

After the end of a transmit pulse, energy stored in the transmitter coil 16 will cause an alternating current to continue to circulate in the resonant circuit for some time. For the values given in the example above, a time constant at which this energy is dissipated is 570use. In order to use a receiver coil (not shown), a transmitter coil voltage needs to have fallen to < 0.5 V, which would take about 9 time constants. The earliest time at which the receiver coil could be used, without being affected by the transmitter circuit would therefore be about Smsec after the transmitter pulse. This limits the use of an NMR instrument, and would prevent measurements being made on materials whose relaxation time constant (T1) was less than about lomsec. It is desirable that the delay before the receiver coil can be used is reduced, as well as preventing coupling effectively between the receiver and transmitter coils.

Thus, in another example of the present invention, illustrated in Figure 4 a solution to this delay is explained. The coil 16 is connected to a parallel arrangement 2 of diodes 17,18 connected in opposite senses.

Each diode is connected to a circuit including a pair of opposed zener diodes 26,27 and a field effect transistor 24,25 with suitable gate decoupling resistors 29,30. Diode 17 is connected to n-type transistor 24 and diode 18 is connected to p-type transistor 25. Alternatively, instead of field-effect transistors n-type and p-type bipolar transistors could be used respectively.

In the circuit shown in Figure 4, the coil 16 is in transmit mode when both transistors 24,25 are turned on, by means of applying -10V to a gate of the p-type transistor 25, and +10V to a gate of the n-type transistor 24. The transistors 24,25 then appear as very low resistances, and one or other of the diodes 17,18 is biassed into conduction by a transmitter signal, depending on the part of the cycle. In receive mode, both transistors 24,25 are turned off (OV applied to their gates) and the diodes 17,18 are biassed off by a fixed voltage 44,45 of the appropriate sign. The diodes then appear as a moderately low capacitance in series with a somewhat larger capacitance of the transistors. The coil 16 is then effectively open circuit.

Normally, attempting to interrupt a current flowing in an inductance results in a high voltage. This voltage is limited by reverse connected zener diodes 26,27 which cause the transistors to conduct with a voltage drop of e.g. 33V, chosen to be slightly greater than the bias voltage 44 so that this bias voltage does not cause the zener diodes to conduct. The energy in the transmitter coil 16 is thereby dissipated in the transistors 24,25, with the current decaying at a rate given by -V/L, where V is voltage drop across the coil and L the inductance of the coil. This results in energy being removed in about 40sec enabling the receiver circuit to be operated more quickly after the transmitter pulse. Where the duty cycle is so great that the energy could not be dissipated in the transistors without damaging them, or in other situations where the mean power is very high, additional dissipative devices (such as well-cooled resistors) could be connected in parallel with the transistors. In a higher power version a separate dissipative device may prove more cost effective.

An example of apparatus in which the present invention may be used is shown in Figure 5. This is based on apparatus described in W092/07279 and will only be briefly described here. A pair of main, first magnets 31,32 are positioned coaxially with their north poles facing one another. The magnets 31,32 are mounted by means not shown to a support 3. Axially inwardly of the magnets 31,32 are mounted a pair of second, auxiliary magnets 33,34 with their north to south axes coaxial with the axis defined by the magnets 31,32. The magnets 31-34 are generally symmetrically positioned about a mid-plane orthogonal to the axis. Four RF transmitter coils 35-38 in series are positioned coaxial with the central axis in the space between the magnets 33,34. These are equivalent to the transmitter coil 16 of the invention described above and the coils 35-38 are connected to the parallel circuit arrangements of Figures 4 or 5. A pair of receiver coils 39,40 are centrally positioned between the transmitter coils 36,37 coaxial with the central axis.

In use, the support 3 is coupled to a mechanism (not shown) which lowers the assembly down through a bore hole 41. The magnets 31-34 are positioned and have field strengths such that a toroidal working volume 42 is generated within the rock strata surrounding the bore hole 41, the magnetic field within the working volume 42 being sufficiently uniform to perform an NMR experiment. This is described in more detail in W092/07279 incorporated herein by reference.

Once the assembly has been correctly located, the transmitter coils 35-38 are suitably actuated to excite nuclear spins within the working volume and as the spins relax and generate corresponding NMR signals, these are received by the receiver coils 39,40 and fed to suitable detection circuitry (not shown) but which is of a conventional form. By using the transmitter coil circuits of the present invention improved performance is achieved over conventional magnet assemblies.

Claims (8)

1. An oscillating magnetic field generating assembly comprising a transmitter coil connected to an oscillating electrical source, wherein the magnetic field generated by the oscillating magnetic field generating assembly is sufficiently uniform within a working region for use in an NMR experiment; the assembly further comprising a pair of diodes connected in parallel in opposite senses between a capacitor and the transmitter coil; and biassing means connected to each diode operable to reverse bias each of the pair of diodes.

2. An assembly according to claim 1, further comprising control means for selectively controlling the biassing means.

3. An assembly according to claim 1 or claim 2, wherein the biassing means comprises one or more batteries.

4. An assembly according to claim 1 or claim 2, wherein the biassing means comprises a semiconductor switch in series with each one of the pair of diodes; and, wherein the assembly further comprises respective voltage limiters for limiting voltage such that energy in the transmitter coil is dissipated in the voltage limiters.

5. An assembly according to claim 4, wherein the voltage limiters comprise a pair of reverse connected zener diodes.

6. An oscillating magnetic field generating assembly as hereinbefore described with reference to the accompanying drawings.

7. A magnet assembly which generates a magnetic field in a working region externally of the magnet assembly having a uniformity suitable for performing an NMR experiment, the assembly incorporating an oscillating magnetic field generating assembly according to any preceding claim.

8. A method of operating a magnet assembly according to claim 7.