Start • Hardware Overview • Embedded Code • User Interface • Functional Code • System Development • Sources and PDF

USER INTERFACE

I developed the user interface with National Instruments’s LabVIEW (see Photo 2). LabVIEW is a graphical programming language meant for engineers to easily acquire and display data. Unlike a text-based language, code is created using graphical virtual instruments (VIs), which are the equivalent of functions. You can create a function and then save it as an icon. Then multiple functions can be used by a larger function, again represented as an icon. Hieratical code is easy to build, and it’s easy to reuse functions.

LabVIEW is a data flow language, which means the flow of data has full control over the order of events. Each of the inputs and outputs of a VI are represented as terminals on the icon. To pass data from VI to VI, you must draw wires from the output terminal of one to the input terminal of another. Built-in functions are found on a graphical palette rather than within .h files. Functions are logically grouped and are as easy as drag and drop. Code can be executed immediately. If there is something that will prevent it from running, the environment is very clear as to where the problem lies. 

The view used to create the program code is called the block diagram. The view used to create the user interface is called the front panel. To create the user interface, all of the controls, indicators, and graphs are also logically grouped in their own palette so you can drag and drop them. When you drop a control on the user interface, it gets a representative icon on the block diagram as a source of data. When you drop an indicator on the user interface, it gets a representative icon on the block diagram as a sink for data. 

In the block diagram, data is represented by wires similar to those you’d see in a schematic capture package (see Photo 3). Arrays of data appear as thicker wires. As dimensions are added to the array, the wires appear even thicker. Also, like in the C language where structures are used to create custom data types that contain multiple types of data, LabVIEW uses something called a cluster. For loops and while loops are just drawn around the code that you want to iterate on. Local data is sent around the loop via shift registers. LabVIEW handles alternate code execution by way of case structures. For every anticipated case (and a default), you draw code within the structure. Debugging is also very easy. LabVIEW lets you watch data flow through the wires in slow motion. You can also set break points and create probes (watch windows).

The other thing that LabVIEW is famous for is its connectivity to hardware and other applications. National Instruments sells hardware such as data acquisition devices with prebuilt VIs to access the functionality with little or no programming. Many equipment vendors such as Agilent provide LabVIEW drivers for their products.  LabVIEW can also handle ActiveX and .NET code, Internet protocols, and much more with ease.

The hardware that interested me was my RS-232 serial port. I needed only four functions to handle the serial port in Windows XP: Open, Read, Write, and Close. It just doesn’t get any easier than that. LabVIEW also produces a multithreaded application by simply drawing separate sections of code on the same block diagram. It can also handle some event-driven programming that responds to button clicks on the front panel. 

The user interface was created to look like an oscilloscope screen. The drag-and-drop graph is set up to display 500 points at a time. Clicking in a display list sets the timebase for the oscilloscope. The analog inputs and outputs are enabled or disabled by clicking on appropriately labeled buttons. The amplitude of the analog output is set by a slider control, while the shape of the wave is selected from a dropdown box. The PWM output is set by clicking in a list that displays all of the available times bases on my analog oscilloscope.