Data Acquisition (DAQ) Basics

Today, the vast majority of researchers and engineers in laboratory research, test and measurement, and industrial automation use PC-based Data Acquisition Systems with PCI, PXI/CompactPCI, PCMCIA, USB, IEEE1394, ISA, parallel, or serial interfaces. Many applications use plug-in devices to acquire data and transfer the data directly to the computer memory, while in some other applications the Data Acquisition Hardware is separate from the PC and connected to the PC via a parallel or serial interface. Obtaining adequate results from a PC-based Data Acquisition System depends on the components shown in Figure 1:

PC

· sensor

Signal conditioning

· Data collection Hardware

· software

This article details the various components of a data acquisition system and explains the most important principles for each component. This article also defines many common terms used for the components of a PC-based data acquisition system.

Figure 1. Typical PC-based DAQ system

Personal Computer (PC)

The computer used in a data acquisition system greatly affects the maximum speed at which data can be acquired continuously, and today's technology allows the use of Pentium-class and multi-core processors, which can be combined with higher-performance PCI/PCI Express, PXI/CompactPCI, and IEEE1394 (FireWire) buses as well as traditional ISA and USB buses. PCI and USB interfaces are standard equipment on most desktop computers today, while ISA buses are no longer commonly used. The advent of PCMCIA, USB, and IEEE 1394 provides a more flexible bus alternative for desktop PC-based data acquisition systems. For remote data acquisition applications using RS-232 or RS-485 serial communications, the serial communication rate often limits data throughput. When selecting data acquisition devices and bus methods, keep in mind the data transmission methods that your selected devices and buses can support.

The data transfer capabilities of a computer can greatly affect the performance of a data acquisition system. All PCs have programmable I/O and interrupt transfer methods. Most current personal computers can use direct memory access (DMA) transfer methods, which use specialized hardware to transfer data directly to the computer's memory, thereby increasing the data throughput of the system. With this method, the processor does not need to control the transfer of data, so it can be used to handle more complex tasks. In order to use DMA or interrupt transfer methods, your data acquisition device must be able to support these transfer types. For example, PCI and USB devices can support DMA and interrupt transfer methods, while PCMCIA devices can only use interrupt transfer methods. The data transfer method you choose will affect the data throughput of your data acquisition device.

The limiting factor for acquiring large amounts of data is often the hard disk. Disk access time and hard disk partitioning can greatly reduce the maximum rate at which data can be acquired and stored to disk. For systems that require acquisition of high-frequency signals, it is necessary to select a high-speed hard disk for your PC so that there is contiguous (non-partitioned) hard disk space to store the data. In addition, use a dedicated hard disk for acquisition and a separate disk to run the operating system while storing data to disk.

For applications that need to process high-frequency signals in real time, a 32-bit high-speed processor and corresponding coprocessor or dedicated plug-in processor, such as a digital signal processing (DSP) board, are required. However, for application systems that only need to collect or convert data once or twice per second, a low-end PC can meet the requirements.

Determine which operating system and computer platform to choose based on the maximum long-term return on investment while meeting your short-term goals. Factors that may influence your choice include the experience and requirements of developers and end users, other uses of the PC (now and in the future), cost constraints, and the various computer platforms that are available during your system implementation. Traditional platforms include Mac OS with a simple graphical user interface, and Windows 9x. In addition, Windows NT 4.0 and Windows 2000 provide a more stable 32-bit OS and are similar to Windows 9x in use. Windows 2000 is the next generation of Windows NT OS, which combines the advantages of Windows NT and Windows 9x, including inherent plug-and-play and power management features.

Sensors and Signal Conditioning

Sensors sense physical phenomena and generate electrical signals that can be measured by data acquisition systems. For example, thermocouples, resistance temperature detectors (RTDs), thermistors, and IC sensors can convert temperature into analog signals that can be measured by analog-to-digital converters (ADCs). Other examples include strain gauges, flow sensors, and pressure sensors, which can measure strain, flow rate, and pressure, respectively. In all of these cases, sensors can generate electrical signals that are proportional to the physical quantity they sense.