Follow Us The principal factors that affect data acquisition systems are: speed, accuracy, power dissipation, package size, and component cost, with varying factors becoming critical depending upon the application. This article shows how a single op amp can be used to drive the ADC in an 8-channel data-acquisition system, reducing the cost and size of the overall system.
The AD7329 eight-channel, 12-bit plus sign, 1-MSPS ADC has true bipolar inputs with four independently programmable software-selectable input ranges: ±4×VREF, ±2×VREF, ±VREF, and 0-to-4×VREF. Boasting a flexible design, it can be configured to suit a wide variety of application requirements. As shown in Figure 1, the AD7329 comprises an 8-channel multiplexer followed by a track-and-hold and successive-approximation ADC, a channel sequencer, a 2.5-V reference, and an SPI-compatible interface.
The analog input channels are routed through the multiplexer to the MUXOUT+ and MUXOUT– pins. The ADCIN+ and ADCIN– pins connect to the track-and-hold input switch (R1) and sampling capacitor (C2), as shown in Figure 2. Note that the input source must provide the current required to drive the ADC input, settling to the required accuracy within the ADC’s 300-ns acquisition time. When the track-and-hold switch goes from hold to track, the transient kick back from the ADC can affect the input source. Applications operating at the maximum sampling rate may require an input buffer amplifier to drive the ADC, isolating the source from the track-and-hold switch.
The flexible design on the AD7329 allows an op amp to be placed between the MUXOUT+ and ADCIN+ pins. In Figure 3, the AD797 ultralow-noise, ultralow-distortion op-amp, isolates the input source from the AD7329’s input structure, increasing the input impedance and reducing the current required to drive the ADC. This configuration also allows a single op-amp to drive eight analog input channels at the maximum sampling frequency, leading to reduced component count, board area, and system cost.
Configuring the op-amp for gain, as shown in Figure 4, allows the AD7329 to accommodate signals in the millivolt range while maintaining high performance. Small signals are amplified by the AD797 with the amplified signal applied to ADCIN+. To maximize performance, the gain is chosen such that the full-scale input signal uses the full dynamic range of the ADC.
Table 1 shows the performance vs. gain achieved on the ±10-V range with a 10-kHz input and a 1-MSPS sampling rate. Note that the converter can still achieve greater than 11 effective bits (ENOB) at a gain of 1000, thus achieving a dynamic range equivalent to that of a 21-bit ADC. In addition, all of the advantages offered by the configuration presented in Figure 3 also apply to this application.
Table 1. AC Performance vs. Gain| Gain (V/V) |
SNR (dB) |
THD (dB) |
ENOB (bits) |
| 1 | 73.57 | −80.80 |
11.93 |
| 20 | 73.00 | −79.91 |
11.71 |
| 200 | 71.66 | −78.99 |
11.61 |
| 500 | 71.48 | −78.46 |
11.58 |
| 1000 | 69.94 | −75.38 |
11.32 |
Some applications require the gain to be changed to accommodate input channels with different signal amplitudes. In these cases, a programmable-gain instrumentation amplifier (PGIA), such as the AD8250, AD8251, or AD8253, can be used in place of the op-amp.
References
AN-0972, How the AD7329 Helps Reduce Costs
