Implementing Project-Based Learning for the Measurement Capable Student With myDAQ and VirtualBench

Steve Watts, Cardiff University

"Because LabVIEW uses a virtual instrument setup, students can design their own methods of integrating the hardware, setting up their test sequences, and manipulating and saving their data, all within one environment."

- Steve Watts, Cardiff University

The Challenge:

We wanted to update a laboratory course to one with a consistent thematic to help year-one students become industry-ready graduates.

The Solution:

We integrated LabVIEW, the myDAQ device, and the VirtualBench instrument to help students learn the design-to-verification chain of a 45 mW discrete transistor power amplifier (PA) by exposure. Students can map taught theory to their measurements, bolstering content and grades in related taught modules that span the academic year.

Author(s):

Steve Watts - Cardiff University
Dr. James Bell - Cardiff University

 

Background

At Cardiff University, we focus on excellence in research that has a measurable impact on society. To achieve greater student participation in research, whether in industry or academia, the electronic and electrical engineering department utilises state-of-the-art research concepts within our teaching in both undergraduate and postgraduate study. This is a challenge for an educator, as often we need to bridge a wide chasm between specialist knowledge and a non-specialist audience. Designing new laboratory content is difficult, especially when university timelines do not afford time to trial new content. We need to cultivate measurement-capable, industry-ready students, whilst also finding ways to pilot new content before the new academic year.

 

Students can use NI hardware coupled with LabVIEW software to develop interesting solutions to these challenges. First, they can easily set up one instrument to take different types of data measurements simultaneously. Because LabVIEW uses a virtual instrument setup, students can design their own methods of integrating the hardware, setting up their test sequences, and manipulating and saving their data, all within one environment. They can then choose to export the raw data or fully analysed graphics depending on preference.   

 

The first-year home student laboratory experiments have traditionally fostered knowledge of instrumentation, circuit prototyping, printed circuit boarding and soldering, good measurement practice and documentation, as well as providing a practical outlet for a range of topics that are fundamental to a first-year student’s electrical and electronic engineering degree. Previous implementations have comprised of a series of experiments and activities with no particular overarching theme, apart from skill development. The autumn semester had a set of small experiments lasting a maximum of 2–3 weeks each. The spring semester had longer experiments and activities lasting 3–4 weeks each.  

 

Our goal was to improve the first-year students’ practical knowledge through project-based learning around a theme of transistor characterisation to system verification to run in the second semester. This implementation would follow a transistor from its final packaging stage into DCIV characterisation for PA design, then into non-linear PA design itself, culminating with both single-tone and then AM signal (multitone) excitation to verify operation and performance.

 

 

International Summer School Pilot

Our first step was starting to test elements of our new course. At Cardiff, we run two-week residential electronic engineering International Summer Schools. We have time after exam marking to create and test laboratory content to pilot in these summer schools. We can use student feedback to improve the new laboratory content. This enhances the student experience, reduces the likelihood of dissatisfied students, and perfects the polish of the content delivery.

 

For this pilot, we used the multifaceted VirtualBench hardware coupled with the LabVIEW environment to drive and measure AM radio transceivers (see Figure 1) in which students can either compose their own MIDI music or plug in their mobile phone music and modulate, transmit, receive, and demodulate to ‘hear’ if their circuit is working correctly.

 

We took advantage of the flexibility of this test scenario to tailor the laboratory content for use as a section of our full undergraduate laboratory course.  

 

 

 

Content Integration Into First-Year Laboratories

After successfully piloting LabVIEW content in our International Summer Schools, we felt confident in our new take on the material and its implementation, which would provide students with a richer context for how to apply their skills in research and industry.

 

 

We grouped four activities out of the existing laboratory module and updated the documentation so that we could implement and coherently express the transistor-related thematic to the students. The broad topics covered by the thematic are: device characterisation, automated test equipment (ATE) measurements, non-linear discrete transistor PA design, and measurement verification. We can split the implementation of these topics into multiple parts:

 

 

 

  • Device Characterisation—A student designs, builds, and uses the myDAQ device to drive a binary count controlled summing operational amplifier (SOP) in which a DC output characterises a discrete transistor. The SOP can transform the static myDAQ ±15 V supply into a variable DC power supply to sweep the transistor (Figure 3). Simultaneously, the myDAQ device drives and records the voltages and currents necessary for plotting the DCIV characteristics.
 
  • Automated Test Measurements—The student follows an in-class tutorial on how to create an automated DCIV routine in LabVIEW using an event-driven state machine as the fundamental architecture (Figure 4). The student automates the DCIV transistor test and real-time plotting of the characteristics, then saves the data to a CSV file. This CSV file is taken into a taught mathematics module for applying statistical measurements and a taught electronics module for transistor modelling, tying three of the student’s modules coherently together.


  • Power Amplifier Design—The student uses (from a taught electronics module) load-line discrete transistor amplifier theory (Figure 5) to create a Class A amplifier that maximises the output power of their transistor delivering 45 mW into a given load.

 


  • Power Amplifier Test (Continuous Wave)—The student measures the output power against a single-tone swept input power and observes the effects of nonlinear behaviour. They characterise amplifier efficiency and total harmonic distortion (THD) and analyse its meaning from a systems perspective (Figure 6). They can use this data to understand a taught analogue communications module content.
 
  • Power Amplifier Test (AM)—The student updates their event-driven state machine to allow for AM generation using the VirtualBench (Figure 7). They experience working with lower level communications between LabVIEW and the VirtualBench and compare and contrast the usefulness of the myDAQ express virtual instruments against writing full LabVIEW code. Furthermore, they create test code that emulates their AM before applying it to their circuit. Upon AM excitation of their PA, the student observes intermodulation distortion and extrapolates the intermodulation distortion IM3 and IM5 points from their results.

 

These activities run as a single thematic throughout the 11-week spring semester and propel students into their second year, which features a more group-oriented sizeable project.

 

 

Assessment

In terms of assessment, a student writes a journal-style technical paper and produces a video explaining the operation of their automated test facility for the DC transistor characterisation. The student must also explain the operation in a mid-term bench demonstration. The course culminates with a full report that requires the student to document results and understanding of the PA and the performance figures of merit. We also ask the student to analyse their LabVIEW code and comment on its effectiveness and document any improvements that they may have thought of or applied.

 

Measuring Impact

We have already observed the impact of implementing this new thematic solution in a few ways.

 

We have seen a qualitative impact based on student feedback. Notable feedback includes:  

  • LabVIEW and NI hardware link easily, whilst still being a challenge for the student  
  • Automated hardware-in-the-loop measurement solutions offer easy access to large data sets that take considerable time if done manually
  • Second semester laboratory module experiments and activities have helped students more thoroughly understand, digest, and own the theoretical content of the taught modules within the semester

 

We have seen a quantitative impact based on student grades. We’ve seen increases in module grades for the following:

  • Laboratory module grades increased by an average of 14 percent  
  • Network Analysis taught module increased by 9 percent
  • Analogue Communications Systems module increased by 8 percent  
  • Electronic Engineering taught module increased by 5 percent
  • Mathematics taught module increased by 2 percent  

 

These are significant increases across the board. This is not a result of better students this year than last. We have not seen such significant increases or variations in any of these module averages over an extended number of years.

 

This clearly demonstrates how project-led learning, in which the project is linked to taught content and the student is encouraged to understand the taught theoretical content using a learn-by-exposure methodology, does work and should therefore be a way forward for any engineering education provider.

 

Summary

Identifying and using a single environment to bring together multiple and different measurements is challenging. Doing this and then in the same environment manipulating and analysing the measurement data is even more of a challenge. Integrating all the above and then plotting and displaying the data in a meaningful and controllable way is near impossible, certainly for a boxed equipment approach. The NI ecosystem of virtual instruments, here the myDAQ and the VirtualBench, coupled with the LabVIEW environment makes it possible to accomplish the above in a comprehensible methodology for the novice through to the expert. It also develops the student LabVIEW experience so they can cross-apply these skills in other laboratory and project scenarios throughout the course.

 

As educators, staff at Cardiff University are in an infinite loop of course updates and reviews. This thematic approach in our first-year laboratory will undoubtedly evolve in the coming years. However, this change to a thematic approach rather than the traditional series of sometimes unrelated stand-alone experiments has had an immediate and positive effect, as shown through student feedback. We can and have used focused feedback from student and staff panels to assess the effectiveness of change.  

 

Cardiff has set up a pipeline of student ambassadors with NI to encourage teaching fundamental LabVIEW paradigms in parallel with the laboratories and bring novice programmers up to Certified LabVIEW Associate Developer (CLAD) level by the end of year one. This addresses student concerns that they do not possess programming skills or the general skill set relevant and required to work in industry.

 

Author Information:

Dr. James Bell
Cardiff University
The Parade, Cardiff, CF24 3AA
United Kingdom
Belljj1@cardiff.ac.uk

 

Author Information:

Steve Watts
Cardiff University
The Parade, Cardiff, CF24 3AA
United Kingdom
WattsS@cardiff.ac.uk

 

 

 

Figure 1. AM Radio Transceivers
Figure 2. myDAQ to ATE and Transistor Design—In One Freshman Semester
Figure 3. ATE Based on myDAQ for Discrete Transistor DCIV Measurements
Figure 4. LabVIEW Event-Driven State Machine Architecture
Figure 5. BJT Load Line Mapped Onto DCIV Plot
Figure 6. PA Output Power, Efficiency, and Harmonic Distortion Versus Input Power
Figure 7. Updated Event-Driven State Machine Architecture