Can you describe the difference between bench instruments and virtual instruments? If not, you're not alone! It gets more and more difficult to distinguish between the two. Virtual instruments, born in the 1970s, let you add functions as you need them, they provide sophisticated tools so users can program complex tasks, and they analyze data. Bench instruments, around much longer, are able to do many of the same things. Sophisticated instruments come with built-in PCs, run standard software, and let users quickly configure tests. And both types of instruments provide Ethernet connections that simplify communications with remote PCs.
Over the years, engineers have found that both bench instruments and virtual instruments play useful roles in product development, all the way from R&D labs to production test. So when should you use which?
Some engineers still think bench instruments are easier and faster to set up and use. But, the software and hardware in today's virtual instruments are burying the reputation this class of tool has had for being difficult to use. Now, virtual instruments are about as easy to use as bench instruments--and they make comparable measurements.
Virtual instruments serve particularly well when engineers must measure signals from many sensors, and when they need to change instrument configurations to accommodate more analog inputs or digital outputs, for example, and when they need to run complicated sequences of tests. Lots of extra channels can't be added to a bench oscilloscope or signal generator. Usually a bench instrument performs one task very well, but in most cases, it can't take on added tasks such as digital I/O and motion control the way virtual-instrument cards can.
Virtual instruments prove valuable when an application requires the acquisition of a lot of data for later analysis, such as a mechanical assembly that's undergoing lengthy stress tests that involve measuring and analyzing signals from hundreds of strain gauges. It's unlikely a stand-alone instrument can take all these readings, store them, and then run a program to produce a test report. But virtual instruments handle that sort of application very well. Need more inputs? Add a card. Update the test sequence? Engineers need to simply change the test software.
Speedy Shortcut: when engineers don't
have a complete system available for testing, they simulate missing parts
with hardware and software. This sort of "hardware in the loop"
configuration can save time and money, although it requires a careful
analysis of the characteristics.
On the other hand, if engineers must measure only a few signals, say during development of a new motor controller, they'll make many tests to quickly home in on signals of interest, and then move on to test other parts of the controller. In this type of situation, they may prefer a bench oscilloscope with manual front-panel controls.
After developing a circuit or assembly, engineers must validate that the design will work under actual operating conditions. But during validation testing, some parts of the overall system that relies on the circuit may not yet exist. In such a case, the engineers may design other equipment to simulate the missing pieces. During development of a navigation controller, for example, it's unlikely a company will actually supply a complete aircraft on which to test the controller's operation.
Instead, the design engineers must simulate actuator motors and use instruments to produce temperature, speed, location, and other data for the controller's inputs. In most cases, virtual instruments offer the only practical means to simulate these signal sources and make a controller "think" it has a real aircraft under its control. Engineers call this type of setup "hard-ware in the loop" because electronic instruments exist in the control or feedback loop to simulate actual motor and sensor operations. Of course, someone on the development team must program the virtual instruments to properly simulate the aircraft's characteristics.
Virtual instruments also excel on production lines. For the most part, production engineers want simple test systems they can set up and almost forget about. Such systems must provide a simple user interface–not a sea of knobs, switches, and displays–and a simple pass/fail indication for production workers. And engineers don't want to set up or configure dozens of individual instruments. Virtual instruments, combined with software provide this type of performance.
Virtual instruments let test engineers choose the capabilities they need from a variety of manufacturers, usually with the assurance that equipment from different sources will "plug and play." The sophisticated software that supports virtual instruments also lets production managers adapt test systems to changing needs, develop several test sequences for different products, and analyze data to produce quality-control reports. And because virtual instruments provide a great deal of flexibility, production engineers can quickly expand test systems to include provisions for motion control, machine vision, and feedback of measurement data to production equipment.
Bottom line: Engineers can endlessly debate the virtues of each instrument type, but plenty of applications exist for both instrument types. Ultimately, decisions come down to choosing the best instrument for a job and ensuring a long life for today's investment.