Shifting Gears with Test System Automation

Author : Andrea Vinci, Senior Technical Marketing Manager, Tektronix

09 November 2023

Automation for test and measurement improves effectiveness, performance and time-to-market. Here, an overview of the most recent advancements and strategic initiatives towards automation-optimised test equipment is given.

Automation seems the only way to go in order to sustain the ever-shrinking delivery time constraints we attempt to respect. As humans, we strive for accuracy and efficiency, yet are prone to struggle and often fall short. But an industrial process, a design lab, a manufacturing unit has to be made so automation comes to the rescue.

Test and measurement processes are no different - engineers try to improve accuracy by increasing precision via more modern tools and instruments. However, they know that an instrument will only be as precise as the person who uses it.

Automated testing processes can provide new insights into the way we operate. They can enable consistent test methodologies, and reduce the risk of accidental human errors. This ensures repeatability of results, as well as saving precious time. 

Instrumentation is generally equipped with multiple physical interfaces (ports) to be controlled - Ethernet, USB, proprietary links (such as TSP-Link), plus older interfaces like RS-232 and GPIB (which are gradually disappearing). 

Through this physical access to the unit, the software can take control and automate in several different ways - with instrument web interfaces clearly being the easiest and most common. Unfortunately, they often offer little benefit when they just replicate the same human machine interface (HMI) that the physical hardware presents, thus leaving users to look for advantages elsewhere.

LabVIEW is a very common tool offering instruments a path to automation and visualisation, though the instrument manufacturer needs to offer structured instrument driver APIs to ensure the customer’s virtual instrument (VI) contains all the necessary functionality. Similarly, driver APIs in programming languages specific to a user’s development environment of choice - be it C#, Python, VB.Net, or C++ - are likely available from the instrument manufacturer or 3rd party. In all cases, the driver API removes the need to work with the native SCPI command set, while providing immediate context-sensitive help and command autocompletion to aid in understanding instrument operation and expediting the code development process.

Figure 1: Automate test procedures is now becoming essential
Figure 1: Automate test procedures is now becoming essential

Test libraries for specific automated test procedures
For those use cases when data retrieval speed is essential and the need for versatility with application analysis is critical, measurement engineers are choosing Python for implementing test automation. Test engineers can release and share code through a platform like GitHub, where a variety of code repositories are hosted. Various active online communities can cooperate to develop and share tools for use in testing processes. Specific Python drivers are provided to improve and speed up measurement analysis - for instance, in the area of signal integrity analysis, where timing, jitter and noise statistical evaluation on a large number of samples is essential. Test and measurement vendors offer the possibility of executing Python commands directly from Python drivers, rather than recalling old-fashioned SCPI commands. This is quite a huge step forward (especially for those who started their first test and measurement experiences, when only manual measurements were available). Tektronix today invites engineers who are also code developers to experiment with the power of native Python commands for controlling oscilloscopes, digital multimeters, source measure units, waveform generators, etc. A user-friendly Python interface is offered, eliminating tedious manual parsing towards a more streamlined workflow.

Test automation workflow in a nutshell
The workflow for communicating with measurement instruments is generally quite simple: any action we request to a test unit corresponds to a control command; sequencing multiple commands (in the correct order) corresponds to what is generally considered to be an automated test script procedure. Imagine a command for initialising/configuring the unit parameters, followed by a command for initiating the acquisition, and finally one to have the acquired data returned, once the intended measurement window concludes. Optimisation is mostly around this, making sure the right chunks of data are transferred, and that test units respond promptly to acquisition requests. 

There are several ways to improve timing in this workflow. One approach is to ensure the test instrument is capable of parallel handling the acquisition and the data processing/transferring. Another approach is to have a local processor carrying out computation, only returning the results if they indicate some critical test condition requires the unit to suspend taking measurements.

Now, keeping in mind our objectives of speeding up testing, reducing inefficiencies, improving accuracy and making the measurement workflow easier for operators, let’s take a look at what the test industry is doing to accomplish that.

For example, measurements related to the validation of specific communication standards must follow a precise and standardised workflow. Measurement equipment today applies some intelligence to the test methodologies, guiding the operator through the measurement. Smart Wizards also guide operators so that they can follow a precise checklist as per standard requirements. In the most complex measurement scenarios, instruments can correlate results to previous acquisitions to identify potential patterns and suggest troubleshooting of root causes. Validation of high speed serial standards is the test application where these efforts are adding most value.

The on-instrument scripting approach
Instruments with advanced processing capabilities, such as advanced graphical SMUs and DMMs, and DAQs, can be programmed with self-developed scripts appearing as Apps, similar to what is available for a common smartphone. These can be referred to as test script processor (TSP) applications. 

Figure 2: The Kickstart ecosystem of instruments
Figure 2: The Kickstart ecosystem of instruments

On-instrument scripting with TSP allows direct, automatic control and processing without the need for an external computer. Moving logic control to the instrument itself essentially reduces communication overhead and gives the instrument the power to make decisions on the fly.

Tektronix offers users the ability to generate custom scripts that can apply mathematical functionality to the acquired set of data - and display or return a calculated value. TSP scripts can be written for a variety of application needs. These go from setting up buffers and triggers with code, through to dynamically applying current and/or voltage limits for sourcing and suchlike. 

Controlling equipment without the complexity of programming 
As stated earlier, Python, C# and industry standard application environments (like LabVIEW) are all extremely powerful ways to take control of equipment. With the necessary connectivity, an instrument can respond to commands, executing functions in any complex developed libraries imaginable. But is that always necessary? According to test operators, often the instrument is used for a very specific and well-defined type of measurement, which is repeated again and again. So why not leave the programming burden to the test solution suppliers and use the various apps they release to accomplish specific jobs? 

That is exactly the case with Kickstart from Tektronix. This proprietary software, once installed on the user’s laptop, enables full control of instrument configuration, test execution and data retrieval from the test device within a period of just a few minutes. For those looking to characterise semiconductor devices and trace the same curves for several days, time-and-again, just changing environmental conditions for each additional test run, there is an app for it. Where cycling batteries and tracing their characteristics over time to understand the capacity degradation behaviour, there is likewise an app for that too. If proving the level of insulation for a specific material is required, while not wanting to get lost in the complexity of using a sophisticated electrometer, this is covered as well.

Dedicated Apps have a relatively low cost and are designed around a specific instrument’s capabilities. Test solution providers regularly maintain them, so that they continue to address the latest updates to standards, and also support the latest equipment models. Customers can benefit from the fact they are solid, robust software options whose maintenance licensing costs are low and can be planned/budgeted for with ease. Tektronix also offers several types of bench software, including TekScope and SignalVu PC. These support instrument automation with point-and-click simplicity.


Contact Details and Archive...

Print this page | E-mail this page