Discussion on preamplifier of seismic acquisition system based on ∑-△

0 Introduction

The original front-end acquisition system of the seismograph used in seismic exploration uses sampling/holding circuit + instantaneous floating point amplifier (FPA) + 14-bit successive approximation A/D conversion. Since the flat-top processing process of the sampling/holding circuit is to cooperate with the FPA to achieve the range expansion of A/D conversion, it seriously suppresses the high-frequency seismic reflection signal. Now most of them have been improved to ∑-△ technology to complete A/D conversion. In the systems currently using ∑-△ A/D converters, the front-end preamplifiers are mostly linear amplifiers in signal conditioning. This paper analyzes the time decay of seismic signals and changes the preamplifiers that cooperate with ∑-△ A/D converters into nonlinear amplifier circuits, so as to give full play to the advantages of ∑-△ A/D converters as much as possible, so as to expand their dynamic range and improve the ability to pick up small signals.

1 Analysis of time domain characteristics of seismic signals

The synthetic model obtained by improving the Sinc wavelet simulates the actual earthquake record, as shown in Figure 1.

In Figure 1, it is found that the large signal segment with relatively concentrated energy in the seismic signal accounts for more than 80% of the signal amplitude, while the small signal segment that effectively represents the seismic reflection layer can only account for less than 10%. The literature proposes an intelligent programmable preamplifier, whose gain automatically increases with depth. The depth of the formation ranges from 0.5 to 3.0 s, and the amplifier gain is 0 dB, 18 dB, 24 dB, 30 dB, 36 dB and 42 dB respectively. Obviously, this processing method can better reflect the advantages of the ∑-△ technology to complete A/D conversion and enable the ∑-△A/D converter to have good 24-bit processing capabilities. However, this step-by-step gain adjustment method will destroy the continuity of the time domain signal during the gain adjustment process, and for the ∑-△A/D converter using oversampling technology, it will cause signal distortion and interfere with the data recovery and playback process. Because it uses a fixed period of gain switching, the interference generated by the switching cannot be regarded as noise, so it cannot be suppressed in the digital filtering process of the ∑-△A/D converter.

In this way, the problem becomes how to ensure the continuity of the time domain waveform while enabling the preamplifier to amplify large and small signals at different gains in signal conditioning, and the dividing point between large and small signals can be set automatically and with certainty.

2 Circuit Scheme and Circuit Principle

Based on the above problems, this paper proposes a method of nonlinear pre-signal conditioning. The schematic diagram of the nonlinear amplifier circuit is shown in Figure 2.

From the literature analysis, we know that:

(1) When the input signal Ui satisfies: U0<|EX+1.2|, D1 and D2 are both off, and the circuit gain is: Av1=-16.

(2) When the input signal Ui satisfies: U0>|EX+1.2|, D1 and D2 are both turned on, and the circuit gain is: Av2=Av2'=-1.

(3) The adjustable EX can set different amplification and limiting ranges for large and small signals for Ui, while maintaining a linear relationship between large signal input and output, without losing the effective components of the large signal.

3 Circuit with microcontroller and DAC to realize adjustable EX

The overall Proteus simulation circuit of the system is shown in Figure 3. In the figure, the single-chip microcomputer U1 adopts AT89C51. Different 8-bit data (A1, A2, ..., A8) are set to U2 DAC0808 through the P2 port of AT89C51 to realize the step setting of EX in the aforementioned nonlinear amplifier circuit.

From the DAC0808 parameter manual, we know:

Output by op amp U3. To generate the corresponding -EX, the circuit uses U4 to reversely amplify EX by 1:1. When AT89C51 adjusts EX step by step, -EX changes synchronously. To observe the changes of EX and -EX, a DC voltmeter is set at the output of U3 and U4 to test the corresponding output DC voltage value.

Operational amplifier U5 cooperates with R8, R9, ..., R13 to form a nonlinear amplification circuit unit. In order to complete the simulation test of the circuit under virtual conditions, a simulation signal source is set at the input end of the unit and a simulation oscilloscope is set at the output end (the A channel of the oscilloscope is connected to the output end, and the B channel is connected to the signal source for easy waveform comparison).

The AT89C51 microcontroller U1's P0.0 is connected to an external key switch to change the data that the microcontroller puts into U2DAC0808.

Here, the circuit composed of C5, C6, R17, R20~R23 is used to simulate the large and small signals in the seismic signal, and the pre-differential amplifier circuit is introduced. In order to highlight the processing effect, R9 in the circuit is doubled and changed to 32 kΩ, so that the amplification factor of the small signal is 32 times.

4 Circuit simulation debugging and characteristic testing under Proteus

Here, the microcontroller program is only used to change the EX value, so it is omitted. A sine wave with a 1 V, a duty cycle of 10%, and a frequency of 10 Hz is used to simulate a large signal; a sine wave with a peak value of 10 mV and a frequency of 200 Hz is used to simulate a small signal. The circuit composed of C5, C6, R17, R20~R23 simulates the seismic signal. The waveform measured at EX=0.07 V is shown in Figure 4.

From the waveform in Figure 4, we know that the amplitude ratio of the original large and small signals is about: 0.6/4.8=0.125 (0.6 is the amplitude of the small signal, 4.8 is the amplitude of the large signal). After the nonlinear amplification circuit is processed, the amplitude ratio of the large and small signals becomes: 2/8.1=0.247. It can be seen that the proportion of the amplitude of the small signal is significantly increased, that is, the gain of the small signal is higher than that of the large signal. As shown in Figure 5, the operation of the simulation software Proteus is shown.

5 Conclusion

From the above simulation results, it can be seen that the principle of this paper can be well verified. The experiment found that the preset of EX has a direct impact on the switching level of large and small signals. Since the setting of EX here can determine the switching point of large and small signal gain, and the gain value of large and small signals in the circuit is determined by the circuit, the data obtained by the ∑-△A/D converter can achieve the expansion of the small signal dynamic range after reverse processing during the playback process.

The idea proposed in this paper is to achieve nonlinear gain adjustment of two different signal amplitudes. Can we achieve three or even multiple similar linear gain nonlinear adjustment methods based on this principle, so that the acquisition of similar time domain characteristic signals is more accurate. In this way, the detailed features of seismic profile data are more complete, and oil exploration is promoted to "move towards precise exploration". The research in this paper provides a feasible method, which is also the direction of the next research.