User interface design for the HP 8146A OTDR - optical time-domain reflectometer - Technical

Hewlett-Packard Journal, Feb, 1993 by Robert Jahn, Harald Seeger

Based on a multiprocessing operating system, the HP 8146A OTDR software can handle simultaneous execution of instrument operations, hide the complexity of instrument operations from the user, and provide a range of user-friendly features.

Although many of the tasks performed by an instrument such as an OTDR may involve many complex operations, the user interface must provide the functionality to hide this complexity from the user. Trained operators must be able to do difficult measurements for characterization of optical links after installation. If a link is broken, the location of the break must be found quickly and precisely. In this situation ease of use is very important because the operator has no time to read a manual or become familiar with a complicated instrument. The following are some of the user-accessible features and capabilities provided by the HP 8146A OTDR:

* Comprehensive documentation and storage of measured traces

* Automatic trace analysis and fault detection

* Help screens and native language support

Automatic ghost detection and removal Remote operation.

OTDR Software

The binary code for revision 1.0 of the OTDR software is almost 500K bytes. The software runs on a Motorola 68010 microprocessor and is based on the real-time multiprocessing operating system pSOS-68k, which can coordinate multiple asynchronous activities. Thus, we have six independent processes running in parallel. Each process has its own data area and stack. A process communicates with others by sending signals or messages via the operating system. Information common to multiple processes is stored in databases, which are accessible only to driver routines. The processes are created after power-on by a special process called the root process. The root process spawns the other processes, creates their mailboxes, and then activates them.

After all six processes are ready to run, the root process eliminates itself.

The display process handles most of the human interface activities such as accepting all input in local mode via the RPG and keystrokes, starting the requested actions, doing some analyses like calculating loss, attenuation, splice loss, and return loss, and drawing CRT text and graphics. Because of complexity, some tasks cannot be handled by the display process and so it passes requests to other processes to do the work.

The other five processes (or message exchanges) are: the measurement process, the dump process, the memo process, the scan process, and the remote control process. The interactions and data flows between these processes are illustrated in Fig. 1.

Measurement Process. To start and stop a measurement, messages are sent to the measurement process, which returns a positive or negative acknowledgment message. In the HP 8146A averaging mode, the measurement process transfers results from the digital signal processing system to the database once every second. After every transfer a message is sent to the display process informing it that new data is available for display.

Dump Process. All communication with output devices such as the internal and external printers and plotter is done by the dump process. The dump process also configures the assigned interface (HP-IB or RS-232).

Memo Process. Every access to registers in battery-buffered RAM, flexible disk, or the memory card is executed by the memo process. When the memo process receives a request such as to store a result to disk, it will acknowledge with a request done message if the operation is successful; otherwise, it reports an error to the display process.

Scan Process. A complete automated analysis of fiber link measurement data is done by the scan process. Results of this scanning are stored in the register database.

Remote Control Process. Remote control of the OTDR is done via the remote control process. The remote control process configures the assigned interface and manages all I/O requests.

The benefit of using independent processes instead of simple function calls is parallelism. The dump process, for example, has a buffer large enough to hold one complete measurement result including parameter settings and results of the automated fiber scanning. The display process accepts the user request for priming the trace and then sends a print message to the dump process and waits for an acknowledgment message. After the dump process receives the print message, it tests the connection to the printer and the printer's status and then copies the trace from the display process's memory to its buffer. This takes some hundreds of milliseconds. Once the dump process is done copying the display data to its buffer, it sends an acknowledgment to the display process so that it can resume accepting user inputs. While running at a low priority in the background, the dump process sends the text and graphic data to the printer. Some OTDRs block all activity while printing or plotting, but the HP 8146A does not.

Help Text

To help users understand and use the instrument effectively we implemented an online help facility. This kind of shortform manual must be available in several languages. Therefore, we offer the capability to provide native language help text. In the field offices where the instrument is sold, engineers are provided with the tools to translate English text into the corresponding local language text and then link this translated text into the OTDR software. Currently we offer help text for English, French, German, Polish, Spanish, Czechoslovakian, Portuguese, and Italian (see Fig. 2).