Author: ADI David Guo, Product Application Engineer, Steven Xie, Product Application Engineer
Precision data acquisition (DAQ) systems are very popular in industrial applications. Some DAQ applications require low power consumption and ultra-low noise. An example is the application of seismic sensors. A large amount of information can be extracted from seismic data. This information can be used in a wide range of applications, such as structural health monitoring, geophysical research, petroleum exploration, and even industrial and household security.
DAQ signal chain requirements
Geophones are electromechanical conversion devices that convert ground vibration signals into electrical signals, and are suitable for high-resolution seismic exploration. They are implanted on the ground along the array to measure the time it takes for seismic waves to bounce off a discontinuous surface (such as a plane), as shown in Figure 1.
figure 1.Seismic source and geophone array
To capture the small output signal of the geophone, a high-sensitivity DAQ signal chain must be constructed for data analysis. The total rms noise should be 1.0 μV rms, the limited flat low-pass bandwidth range is about 300 Hz to 400 Hz, and the signal chain should achieve a THD of about -120 dB. Since the seismic instrument is powered by a battery, the power consumption should be controlled at about 30 mW.
This article introduces two signal chain solutions, the goals and requirements achieved are as follows:
· PGIA gain: 1, 2, 4, 8, 16
· ADC with integrated programmable broadband filter
· When the gain = 1 (the -3 dB bandwidth is 300 Hz to about 400 Hz), the RTI noise is 1.0 μV rms
· THD: -120 dB (when gain = 1)
· CMRR ＞ 100 dB (when gain = 1)
· Power consumption (PGIA plus ADC): 33 mW
· The second channel is used for self-test
DAQ signal chain solution
There is no precision ADC on the ADI website that has all these features and can achieve such low noise and THD, and no PGIA can provide such low noise and power consumption. However, ADI provides excellent precision amplifiers and precision ADCs that can be used to build signal chains to achieve goals.
In order to build low noise, low distortion and low power consumption PGIA, ultra-low noise ADA4084-2 or zero-drift amplifier ADA4522-2 are good choices.
Regarding very high-precision ADCs, the 24-bit Σ-Δ ADC AD7768-1 or the 32-bit SAR ADC LTC2500-32 are the top choices. They provide configurable ODR and integrate flat low-pass FIR filters, suitable for different DAQ applications.
Seismic signal chain solutions: ADA4084-2 PGIA and AD7768-1
Figure 2 shows the entire signal chain. ADA4084-2, ADG658 and 0.1%resistanceLow noise, low THD PGIA can be constructed, providing up to eight different gain options. AD7768-1 is a single channel, low power consumption, -120 dB THD platform. It has a low-ripple programmable FIR, DC to 110. 8 kHz digital filter, and uses LT6657 as a reference voltage source.
figure 2. ADA4084-2 PGIA and AD7768-1 plusMCUFiltered signal chain solution
When AD7768-1 runs with ODR of 1 kSPS, the root mean square noise is 1.76 μV rms; in low power consumption mode, the power consumption is 10 mW. In order to achieve the final 1.0 μV rms noise, it can operate at a higher ODR, such as 16 kSPS in medium-speed mode. When AD7768-1 runs at a higher modulator frequency, it has a lower noise floor (as shown in Figure 3) and higher power consumption. The flat low-pass FIR filter algorithm can be implemented in MCU software to eliminate higher bandwidth noise and reduce the final ODR to 1 kSPS. The final root mean square noise will be about a quarter of 3.55μV, or 0.9μV.
image 3.Use the MCU post filter to balance the ODR of AD7768-1 to achieve the target noise performance
As an example, the MCU software FIR filter can be constructed as shown in Figure 4 to balance performance and group delay.
Seismic signal chain solutions: ADA4084-2 PGIA and LTC2500-32
ADI’s LTC2500-32 is a low-noise, low-power, high-performance 32-bit SAR ADC with integrated configurable digital filter. The 32-bit digital filter’s low noise and low INL output make it particularly suitable for seismology and energy exploration applications.
High impedance sources should be buffered to minimize the settling time during acquisition and optimize the linearity of the switched capacitor input SAR ADC. In order to obtain the best performance, a buffer amplifier should be used to drive the analog input of the LTC2500-32. Must design a discrete PGIACircuitTo drive LTC2500-32 to achieve low noise and low THD (introduced in the PGIA part).
The main specifications of the PGIA circuit include:
· Power supply: 5 V (minimum)
· AD7768-1 has a power consumption of 19.7 mW, so the power consumption of the PGIA circuit should be less than 13. 3 mW, in order to meet the power consumption target of 33 mW
· Noise: when gain = 1, the noise is 0. 178 μV rms, which is about 1/10 of AD7768-1. 78 μV rms
There are three types of PGIA topologies:
· Integrated PGIA
· Discrete PGIA with integrated instrumentation amplifier
· Discrete PGIA with operational amplifier
Table 1 lists ADI’s digital PGIA. The LTC6915 has the lowest IQ. The noise density is 50 nV/√Hz, and the integrated noise in the 430 Hz bandwidth is 1.036 μV rms, which exceeds the target value of 0.178 μV rms. Therefore, integrating PGIA is not a good choice.
Table 2 lists several instrumentation amplifiers, including the AD8422 with 300μA IQ. Its integrated noise in the 430 Hz bandwidth is 1.645 μV rms, so it is not a good choice either.
Figure 4. MCU post FIR filter stage
Figure 5. ADA4084-2 PGIA and LTC2500-32 signal chain solution
Image 6.LTC2500－32 Flat Passband Filter Noise under Different Downsampling Coefficients
Table 1. Digital PGIA
Table 2.Instrumentation amplifier
table 3.Low noise, low power operational amplifier
Figure 7.Discrete PGIA block diagram
Use an operational amplifier to build a discrete PGIA
The article “Programmable Gain Instrumentation Amplifier: Finding the Best Amplifier for You” discusses various integrated PGIAs and provides good guidelines for building discrete PGIAs that meet specific requirements2. Figure 7 shows a block diagram of the discrete PGIA circuit.
You can choose ADG659/ADG658 with low capacitance and 5 V power supply.
For operational amplifiers, IQ (<1 ma per channel) and noise (voltage density <6 nv>
As to the gain resistance, choose 1.2 kΩ/300Ω/75Ω/25Ω resistance to realize 1/4/16/64 gain. The larger the resistance, the noise may increase, and the smaller the resistance, the more power consumption is required. If other gain configurations are required, resistors must be carefully selected to ensure gain accuracy.
The differential input ADC acts as a subtractor. The CMRR of the ADC is greater than 100 dB, which can meet the system requirements.
You can use LTspice? to simulate the noise performance of a discrete PGIA. The integrated noise bandwidth is 430 Hz. Table 4 shows the noise simulation results of two different PGIAs and AD7768-1. The ADA4084 solution has better noise performance, especially at high gains.
Table 4.Noise simulation results
In-loop compensation circuit drives LTC2500－32
AD7768-1 integrates a pre-charge amplifier to reduce drive requirements. For SAR ADCs, such as LTC2500-32, it is generally recommended to use a high-speed amplifier as a driver. In this DAQ application, the bandwidth requirement is very low. In order to drive LTC2500-32, it is recommended to use an in-loop compensation circuit composed of a precision amplifier (ADA4084-2). Figure 8 shows the in-loop compensation PGIA used to drive the LTC2500-32. The PGIA has the following characteristics:
· The key components of R22/C14/R30/C5 and R27/C6/R31/C3 are used to improve the stability of the compensation circuit in the loop.
· Using ADG659, A1/A0 = 00, gain = 1, the feedback path of the upper amplifier is the amplifier output? R22? R30? S1A? DA? R6? AMP — IN.
· Using ADG659, A1/A0 = 11, gain = 64, the feedback path of the upper amplifier is the amplifier output? R22? R8? R10? R12? S4A? DA? R6? AMP — IN.
PGIA is connected to LTC2500-32EVB to verify performance. Experiment with different passive components (R22/C14/R30/C5 and R27/C6/R31/C3) values to achieve better THD and noise performance under different gains (1/4/16/64). The final component values are: R22/R27 = 100 Ω, C14/C6 = 1 nF, R30/R31 = 1.2 kΩ, C3/C5 = 0.22 ?F. When the gain below PGIA is 1, the measured 3 dB bandwidth is about 16 kHz.
Figure 8. PGIA drives LTC2500－32
Test bench evaluation settings
In order to test the performance of noise, THD and CMRR, separate ADA4084-2 PGIA and AD7768-1 boards are made into a complete solution. This solution is compatible with the EVAL-AD7768-1 evaluation board, so it can interface with the control board SDP-H1. Therefore, you can use the EVAL-AD7768FMCZ software GUI to collect and analyze data.
The ADA4084-2 PGIA and LTC2500-32 boards are designed as alternative complete solutions.Circuit boardInterface with SDP-H1 control board and controlled by LTC2500-32FMCZ software GUI.
The PGIA gain of the two boards is designed to be 1/2/4/8/16, which is different from that shown in Figure 8. Table 5 shows the evaluation results of these two boards.
Figure 9. ADA4084-2 PGIA and AD7768-1 evaluation board solution
table 5.Signal chain solution test results
Figure 10. FFT of the ADA4084-2 PGIA and LTC2500-32 boards when the gain is 1
For seismology and energy exploration applications, in order to design a very low-noise and low-power DAQ solution, a discrete PGIA can be designed with a low-noise, low-THD precision amplifier to drive a high-resolution precision ADC. This solution can flexibly balance noise, THD and ODR according to power requirements.
· The low-noise performance of LTC2500-32 combined with the advantages of ADA4084-2 and LTC2500-32 makes the solution exhibit the best noise performance without the need for further filtering by the MCU.
· When PGIA gain = 1, ADA4522-2 and ADA4084-2 have good noise performance. The noise performance is about 0.8 ?V rms.
· ADA4084-2 has better noise performance at high gain. When gain = 16, the noise of ADA4084-2 and LTC2500-32 is 0.19 μV rms, which is better than ADA4522-2’s 0.25 μV rms.
· For AD7768-1, with the help of MCU filtering, the ADA4084-2 and AD7768-1 solutions show similar noise performance as the ADA4084-2 and LTC2500-32 solutions.
The data acquisition solution presented in this article requires low noise and low power consumption, but the bandwidth is limited. Other DAQ applications have different performance requirements. If low power consumption is not necessary, the following operational amplifiers can be used to build PGIA:
· Lowest noise: LT1124 and LT1128 can be considered to obtain the best noise performance.
· Lowest drift: The new zero-drift amplifier ADA4523 has better noise characteristics than ADA4522-2 and LTC2500-32.
· Minimum bias current: If the output resistance of the sensor is high, it is recommended to use ADA4625-1.
· Higher bandwidth: ADA4807, LTC6226 and LTC6228 are good solutions when building high bandwidth, low noise PGIA in high bandwidth DAQ applications.
In DAQ applications where noise and power consumption are not important, but require smaller PCB area and high integration, ADI’s new integrated PGIA ADA4254 and LTC6373 are also good choices. The ADA4254 is a zero-drift, high-voltage, robust PGIA with a gain of 1/16 to ~176, while the LTC6373 is a 25 pA IBIAS, 36 V, 0.25 to ~16 gain, low THD PGIA.
Table 6.Precision Operational Amplifier Selection Table
1 Geophone. ScienceDirect.
2 Jesse Santos, Angelo Nikko Catapang and Erbe D. Reyta. “Understand the basics of seismic signal detection networks”. Analog Dialogue, Volume 53, Issue 4, December 2019.
3 Kristina Fortunado. “Programmable Gain Instrumentation Amplifier: Find the best amplifier for you.” Analog Dialogue, Volume 52, Issue 4, December 2018.
About the Author
David Guo is a product applications engineer in the Linear Products Division of Analog Devices. He joined ADI’s China Application Center in 2007 as an application engineer, and then transferred to the precision amplifier department as an application engineer in June 2011. Since January 2013, David has served as an application engineer in the Linear Products Division of Analog Devices. He is responsible for the technical support of precision amplifiers, instrumentation amplifiers, high-speed amplifiers, current sense amplifiers, multipliers, reference voltage sources and RMS-DC products. David holds a bachelor’s degree and a master’s degree in mechanical and electrical engineering from Beijing Institute of Technology. Contact: david. [email protected] com.
Steven Xie joined ADI Beijing branch in March 2011 as a product application engineer at ADI China Design Center. He is responsible for the technical support of SAR ADC products in the Chinese market. Before that, he worked as a hardware designer in the field of wireless communication base stations for four years. In 2007, Steven graduated from Beijing University of Aeronautics and Astronautics with a master’s degree in communications and information systems.