Data acquisition systems have evolved to provide more accurate, fast and repetitive data. With these, you can acquire any signal from anywhere and produce reliable results
SHWETA DHADIWAL BAID
Data acquisition is the most important task for any system or application that involves signals. Acquiring the accurate data and analysing it correctly is critical for applications ranging from healthcare, automation and control to space exploration. You need to acquire data to test a system, make useful analysis and give feedback signal to the system. It is the first and most important step in any project.
All the applications have their own requirements of speed of acquiring, transferring and representing the data. Keeping in sync, there has been a substantial advancement in data acquisition systems (DAQs). “Use of wireless technology, image processing algorithms, and faster and higher-resolution analogue-to-digital converters (ADCs) has increased. New DAQ components have made the systems smaller, faster, cheaper and able to pack in more applications,” shares Shirish Patwardhan, CTO, KPIT Cummins.
“Intelligent DAQs with inbuilt field-programmable gate arrays (FPGAs), developments in bus technologies and virtual instrumentation are some of the latest and fast-moving developments in data acquisition systems,” informs Pradeep Nair, business development manager, National Instruments.
Arif Fahim from AD Instruments explains, “Computer-based data systems reduce a lot of effort put in manual calculations and help in producing error-free analysis.”
Components of data acquisition systems like sensors, transducers, ADCs, interfaces and application software have undergone a revolutionary change, making DAQ systems faster and more reliable.
Sensors go wireless
Sensors are the first point of contact between the system under observation and the DAQ system. Transducers sense physical phenomena like temperature, pressure and strain and convert these into useful electric signals that are actually measured by the DAQ system. These signals undergo signal conditioning depending on the quality of the signal acquired.
As wireless sensors are becoming popular, Nair cites the example of monitoring the flooding Yamuna. He explains, “You can place the sensors along the bank to measure the temperature and water level and push the data to a central server without a wired connection. Earlier, these projects were not touched due to infeasibility and cost involved in laying long cables along the river.”
Fahim feels that wireless sensor technology is very useful for data acquisition in life science research.
“Sensors available today are much more accurate and sensitive and do a fantastic job. There are smart sensors with built-in intelligence, radio frequency identification (RFID) based sensors and micro-electromechanical sensors that can acquire data from small areas,” says Nair.
“Transducer electronics datasheet (TEDS) sensor is another new type becoming increasing popular in DAQ. Essentially, it is a transducer with attached memory that contains information in itself and can configure to any kind of DAQ,” adds Nair.
While acquiring data in certain applications, both the rate at which you need to acquire a sample as well as the resolution at which it gets digitised are important.
“Applications in aerospace and defence industry centre around recording signals either coming from RADAR returns or surveillance systems. Here the sampling requirement is very high and there is a need to have sustained data throughout,” explains Santosh Reddy, application engineer, Agilent Technologies.
“Earlier, the user had to trade off between speed and resolution; for very high speed (in megahertz), you had to compromise on resolution (8-bit). But today you have the hardware that can acquire at very high speed coupled with high-resolution ADC,” explains Nair. There are digitisers capable of sampling at high speeds of mega-samples/second and giga-samples/second. This allows you to have very precise measurement at very high speed.
The resolution can be misguiding, though. Reddy explains, “The resolution mentioned on the datasheet is usually the ADC resolution, but the actual performance is usually much lower due to system noise. To get the true picture of the data acquisition card performance, you need to examine beyond the specification. You must ask for the effective number of bits (ENOB) specification.”
Interfaces for data transfer
With computers entering the test and measurement world, a variety of options for interconnectivity such as Ethernet, USB and high-speed IEEE 1394 have become available for data transfer externally. Legacy systems made use of GPIB/HPIB buses that had severe data throughput limitations. Then came the peripheral component interconnect (PCI) local bus, which was one of the most widely used internal buses for the PCs in 90’s and provided speed of 133 MB/s. In 2004, PCI Express, the successor to PCI, increased the throughput to more than 250 MB/s.
USB connectivity for a data acquisition card brings along a lot of attractive features like low cost, ease of use, plug-and-play functionality and built-in operating system configuration. “USB DAQ cards provide analogue input channels, analogue output channels, digital input/output (I/O) channels and counter/timer channels with a throughput of 480 Mbps,” shares Reddy. However, USB poses the limitation of length as the cable can be extended up to 30 metres only.
Ethernet technology is a standard and now widely used for interconnectivity in standalone systems. It can take the advantage of remote connectivity to new or existing Ethernet network. Today, the most common Ethernet networks are 10BASE-T and 100BASE-TX, which transfer data at 10 Mbps and 100 Mbps, respectively. As PC technology evolves faster than test and measurement, it is important to select a futureproof bus technology.
Despite the advantages of wide availability of USB or existing Ethernet network, there is still a need to connect and convert from one bus technology into another to match the existing set-up. Bridge products allow you to connect the USB/Ethernet to GPIB, as the experts feel that the future of bus technology will comprise mixed I/O connectivity.
“Wireless data transfer using wireless interfaces like Wi-Fi network, Zigbee, Bluetooth and GPRS has now been accommodated in modern DAQ systems. In this, the sensor will acquire the data and put it via GPRS network to your mobile phone. This allows you to push the data anywhere in the world with no limitation as telecom is already a matured network,” explains Nair.
Shrink in size
The components of data acquisition have changed in terms of complexity and density. “Due to advances in semiconductor technology, data acquisition systems have become smaller and pack a much larger number of channels into the same footprint,” states Reddy. “Some modules have integrated signal conditioning elements onto the data acquisition card.”
“DAQ systems with plug-and-play feature have become increasingly popular because of their portability, ease of operation and easy PC connectivity with high-speed USB,” shares Nair. The sensor-USB connectivity allows you to enjoy the flexibility of acquiring signals from any type of sensors.
Advancement in PCs benefits DAQ
With the developments in PC technology, including user-friendly interface, ability to automate the system, and data storage and representation, data acquisition has taken a new form. Today, we have very high-performance machines powered by Pentium IV and PowerPC coupled with new bus architectures. DAQ takes the advantage of higher throughput, improved real-time processing and ability to use complex video graphics.
“Data acquisition can now be integrated with all the existing technologies and take advantage in providing real-time and reliable data,” expresses Nair. The high-speed 32-bit specialised DSPs can now be replaced with today’s PC processor that not only does data acquisition but also takes care of data storage, data transfer and data sharing.
Alok Gupta, chief managing director, AG Measurematics adds “As the use of the Internet has become universal, the DAQ also has become Web-enabled. This allows you to view the behavioural change, say, in the process plant from any location in the world in real time.”
Sensor. It is very important to know your signal and transducer well. “A lot of people make mistake in selecting the sensor, leading to junk acquisition that soils the entire system,” says Nair. You need to choose a sensor and transducer that will give you the most accurate and reliable results.
Signal-conditioning requirement. Every signal that is acquired will show variation, so the signal-conditioning requirement will also vary. Having good knowledge of signal conditioning and getting proper consultation is very important. There are signal-conditioning systems that can be configured as per the requirement.
ADC. Selecting the right type of analogue-to-digital converter (ADC) is important as it will reflect the sampling rate and the data transfer rate. If you use a very high-resolution ADC, the number of codes will be high and it will require more memory to store.
Interface. As the signal acquired gets transferred to the storage drive or PC for further analysis, depending on the requirement of data transfer rate and sampling rate you can select the bus through which you want to communicate.
Drivers. It’s the drivers that make your system futureproof. If the driver is proprietary and you can use it for certain time only, your DAQ system may not be useful later. A good driver allows you to bridge the gap between the application hardware and software.
Application software. “Application software allows the user to set up different channels, display measured values in real time, store data in files on the hard disk, generate test reports and view previously stored values,” explains Gupta from AG Measurematics.
“In the past, we used a paper recorder with styli to trace the brain signals in electroencephalograph. The paper roll ran up to 300 sheets for a 30-minute recording on an average. This also made analysis of the signal difficult as you had to manually view and measure all the signal levels. The floor space requirement was around 50 sq.m,” shares Dr Shantilal Dhadiwal, a renowned psychiatrist. “With digital EEG, the signals are acquired directly into the desktop computer or a laptop and the storage is also digital. The software helps in zooming in and zooming out the anomalies in the signal giving better view for diagnosis. Also, the record, which is digital, can be shared (via CD or the Internet) with any other doctor sitting abroad.”
Intelligent DAQ with FPGA
Data acquisition systems have become intelligent enough to make some of the decisions. “DAQs with built-in FPGAs are used for on-board measurements and control applications,” informs Reddy.
“The intelligent DAQs with FPGA allows you to program the FPGA to take immediate action without transferring the data to the controller. The data acquired is analysed on the local memory with a latency of 1-2 seconds and then you can push the selective useful data to your controller,” says Nair.
He explains, “If there is spike in the system that you are monitoring, you can send an immediate control signal like ‘shut down’ from your programmed FPGA and just send the interesting data, maybe the ‘time’ when the spike occurred, to your PC.”
The intelligent DAQ system delivers high performance, user-configurable time and synchronisation as well as on-board decision-making. It gives you the flexibility to configure the hardware for simulation, bit-error-rate testing, flexible triggering and other applications that require precise timing and control.
Virtual instrumentation and modelling
Application software defines the usefulness of the data that is acquired by your hardware. It allows you to do useful analysis with the data acquired and send back the control signal for your system. The analysis can be done online, where signals are acquired and analysed immediately to generate the control signal, or off-line, where you analyse the data later to study and project trends. Certain DAQ software allow you to import and export data to common databases like Microsoft Excel and Access. Many DAQ hardware providers have their own software modules to present the useful results.
Another part in data acquisition where software plays a major role is modelling and simulation. “Modelling and simulation is very important in technical computing work-flow and acquiring data for computation becomes critical part of it,” says Prashant Rao, technical manager, Mathworks India. “It helps you (the user) to figure out what exactly you would like to do with the data (for example, a feedback signal to a control system) in optimal manner before moving to the real product.”
“Graphical programming has become the most popular feature of the software today. With this, you can click and get the data into the system without having to write and run the lengthy code,” shares Rao.
Companies have developed software applications that allow you to build the entire data acquisition system without using the hardware. “Virtual instrumentation allows you to use highly interactive software along with modular, high-performance hardware to create a powerful computer-based instrumentation solution,” shares Nair.
Beat the challenge
“The biggest challenge for DAQ is the growth in the number of integration points and short cycle time of deployment. Increase in functionalities due to newer regulations in safety, emissions, environment sensibility and drive to reduce the carbon footprint leads to complexity,” informs Patwardhan.
“With varied applications of data acquisition, the requirement and challenges both are different. As we deal with data, purity of data, compression and consolidation are challenging areas,” says Neelam Kumar, executive director, Aplab.
But with technological developments, all these challenges are manageable. Data acquisition systems are able to integrate and benefit from the developments in sensor, high-speed digitisation, bus and PC technologies.
The author is a senior technology journalist at EFY