In Part 1 of this blog, we started looking at reducing noise, increasing dynamic range, and increasing the ENOB (effective number of bits) with a SAR (successive approximation register) ADC. The method is based on oversampling — commonly used on low-speed, high resolution delta-sigma ADCs — less commonly used elsewhere. We continue by looking at some test results with a SAR ADC using an Eval board and its software.
The oversampling capability is implemented in the AD7960/61 evaluation software using a simple averaging of the ADC output samples, meaning, and summing the number of ADC samples and dividing it by the oversampling ratio to get the increased dynamic range. This software allows the user to select the “Oversampling Ratio” up to 256 from the drop-down menu (red box) under the Configure tab as shown in Figure 1. The maximum dynamic range achieved is limited by the low frequency 1/f noise of the system, which starts to dominate at lower output data rates below 20kSPS.
A spectrum of the signal and the flat noise from DC to fs/2 in Figures 2 and 3 shows that the noise can be filtered to fs/(2•OSR) to improve the dynamic range and SNR. In this case, the oversampled dynamic range is the ratio of the peak signal power to the noise power measured in the ADC output FFT from dc up to fs/(2•OSR), where fs is the ADC sample rate.
The AD7960 and AD7961 achieve a typical dynamic range of 100dB and 96dB respectively, using a 5V reference as specified in the datasheet, so in theory, we should see a 24dB increment in the dynamic range due to oversampling by 256.
In reality, the measured oversampled dynamic ranges of these parts are 122dB and 119dB, respectively, with no input signal when oversampled by 256X at output data rate of 19.53kSPS, which shows a 1 to 2dB degradation in the dynamic range from the theoretical calculation. This is limited by the low frequency noise coming from signal chain components, input source, and printed circuit board. With a 1kHz full-scale sine wave input signal, these parts achieve oversampled SNR of roughly 111dB and 110dB, respectively. Figure 4 shows how the AD7960 achieves the increased dynamic range as the oversampling ratio is increased and output data rate is decreased.
MRI system operates in the 1MHz to 100MHz RF frequency band, whereas computed tomography (CT) and Digital X-ray operate in the 1016 Hz to 1018 Hz frequency range, subjecting patients to ionizing radiation that can damage living tissue. MRI gradient control systems demand a very high dynamic range, tight linearity, and fast response time from DC to tens of kilohertz, and its gradient be precisely controlled within around 1mA (1ppm) in either analog or digital domain for enhanced image quality.
Using an oversampled SAR ADC with good specs (such as the AD7960) would allow design engineers to achieve the high dynamic range and meet the key requirements for MRI systems. Such systems need measurement repeatability and stability over long periods of time in a hospital or doctor's office environment. The additional requirements that the design engineer should look for are high resolution, accuracy, low noise, fast refresh rates, and very low output drift.
Have you worked with any ADC (delta-sigma or SAR) and run it in an oversampling mode? Did you get the expected results? Did you encounter any problems?
- Increase Dynamic Range With SAR ADCs Using Oversampling, Part 1
- Which Is Better: SAR or Delta-Sigma ADCs?
- ADC Noise: Where Does It Come From?
- Interleaving Spurs: The Mathmatics of Timing Mismatch
- ADC Basics, Part 9: PGA Embedded in an 8-Channel, 12-Bit SAR
- Signal Chain Basics #80: Optimizing Power vs. Performance for a SAR-ADC Drive Amplifier
- ADC Guide, Part 13: Input Impedance
- Data Converters in Massively Parallel Analog Systems
- ADC Basics, Part 4: Using Delta-Sigma ADCs in Your Design
- ADC Basics, Part 3: Using Successive-Approximation Register ADC in Designs
- ADC Basics, Part 2: SAR & Delta-Sigma ADC Signal Path