ASEE Paper on Personal Instrumentation and the ADALM1000

In this paper from the June annual meeting of the ASEE, "Requirements for the Effective Application of Personal Instrumentation in ECE Undergraduate Courses", faculty from a number of universities, including many HBCU schools, report on their findings with regard to the minimum requirements of personal instrumentation necessary for use in undergraduate ECE curriculum.

The full text of the paper in PDF format can be down loaded from here.

The paper attempts to answer this main question: "Can the device chosen be used by students in all activities and all core classes?" Specifically the ADALM1000 is examined to see to what extent it satisfies this question. Since ADI manufactures the M1k we might be slightly biased in its favor.

In a survey conducted by the authors of active users of personal instrumentation from more than a dozen universities (see Figure 1 below from the paper), it was found that over 70% of the core circuits and electronics experiments can be done within the audio frequency range specs of the M1k. Where the limiting factor was more the power supply and signal (0 to 5 volts) range than frequency response, over 80% could be done within the specs of the M1k. The blue "before" bars indicate the percentage of labs that fit ranges used with standard bench instruments and the orange "after" bars indicate the percentage of labs that fit ranges used with USB instruments with specs like M1k and M2k (ADALM2000).

Figure 1, Instructors from 12+ universities surveyed on the required specs Before and After they began using USB instruments in circuits and electronics intensive courses.

Numbers are the percentages of their experiments falling in each range.

It can be plainly seen that given a little ingenuity most labs can or could be adapted in frequency (audio range) and signal range (0 to +5) to use M1k, as we see in the increase from just over 60% to 70% for frequency and just over 60% to 80% for voltage. In the few cases where I have interacted with faculty on adapting labs for M1k, I have generally been able to come up with ways to use the hardware functionality of M1k and/or add new functionality to the software to adapt their labs. So my guess is that with even a little more work the percentages might be well over 90% is not 100% coverage of the key ECE concepts.

ECE is a very broad ranging area of study and there are obviously more fundamental core concepts than can be squeezed into a fixed number of undergraduate courses. Curriculum committees in ECE departments naturally down select what each feels is most important. Given the more than 80 example lab activities developed by ADI for M1k currently available on the Wiki, I would argue that there are enough core Electrical Engineering concept labs (Circuits, Electronics, Signals and Systems) here to fill 100% of an undergraduate ECE curriculum. Ultimately, "what" is important depends on the career path of the student and their eventual employer.

Getting back to the study, essentially all practitioners said that the power supply voltage range limitation could be taken care of by occasionally using 9V batteries. Batteries are a simple and obvious choice but do not supply a constant voltage as they discharge so ADI supplies a linear voltage regulator in the ADALP2000 analog parts kit. By using this 3.3 V LDO (with a 9 V battery and a couple of resistors) hanging below ground or the fixed 2.5 volt supply of the M1k an additional negative supply voltage in the range of -0.8 to -5 V is possible. Stacked on top of the fixed +2.5 and +5V supplies of the M1k an additional positive supply voltage in the range of +5.8 to +10 V is possible.

Using DC-DC converters like the LT1054 or ADM660 to either invert or boost the fixed +5 V of the M1k we can generate -5 and +10 V supplies without the need for extra batteries.

Active users were also asked if they could give an example of an experiment they would like to do, but could not do with current USB instruments. For voltage range, the only response was to provide a power supply that can operate at ±12-15V for more flexible op-amp experiments. My guess is that this response is more from inertia and the tendency to want to stick with the good old tried and true 741 op-amp rather than move to a more modern op-amp. More and more leading edge amplifiers are designed to use a single +5 V supply (or lower) and fewer amplifiers are available even for -5 to +5V supplies. So moving to single supply rail-rail input/output amplifiers only makes sense.

One desirable capability not presently generally available, the ability to source, sink and measure current was mentioned, especially for Thevenin/Norton source studies. Those capabilities are available with M1k which can be a unique plus when designing lab activities.

Moving on to the issue of input voltage range, it should be noted here that sometimes different terms are often used (sometimes incorrectly) when talking about the requirements of the voltage measurement capabilities of an instrument. Dynamic range is more accurately used when talking about the signal to noise ratio or the ratio of the largest and smallest voltages that can be measured simultaneously, rather than just the absolute maximum voltage. Two instruments might have the same maximum input voltage but very different ADC resolutions such as 8, 10, 12, 14 or 16 bits.

Resistive dividers are used to change the maximum absolute voltage. For example the M2k input stage has two resistive dividers built-in. One divider maps the 12 bit ADC full scale to a -3 to +3 V range which results in a 6/4096 or 1.46 mV LSB step. On the other hand the M1k with no input voltage divider maps the 16 bit ADC full scale to a 0 to 5 V range which results in a 5/65536 or 76 uV LSB step. The second M2k divider maps the ADC FS to -30 to +30 V or 14.6 mV steps. If a similar resistor divider were placed ahead of the M1k input such that it maps the ADC FS to 60 V (or +/- 30) the LSB would be 60/65536 or 0.91 mV steps. Which is still smaller than the M2k even with the +/- 3 V divider. So ADC resolution is still very important no matter what the Maximum input voltage might be. Using a low resolution ADC and voltage dividers/gain ranges results in sort of a floating point measurement which does not give you both maximum voltage and minimum voltage measurement capabilities simultaneously.

In various documents on the M1k it has been shown that the (limited) 0 to 5 V analog input range can be increased by the use of resistive voltage dividers and frequency compensated voltage dividers. With the addition of a few passive components the M1k can match the maximum input voltage range of the other instruments and actually exceed them in usable dynamic range.

One final question from the paper:

What is the price point that makes personal instrumentation affordable to a student?

"This is a difficult question to address quantitatively. However, there is a lot of anecdotal evidence in the US and elsewhere that prices over $100 seem to present a barrier to student ownership in developed countries and about half that or less in developing countries. Prices have moved a bit upward once the market has been established because the value proposition of the small, portable learning platform is well understood. The power of the $100 barrier is such that nearly all new products are priced there or below."

The fact that as little as $100 makes a difference when 4 years in an undergraduate ECE program costs many thousands if not tens of thousands of dollars is the real issue here. If such a small fraction (less than 0.1%) of the total expense for this kind of student owned hardware is a show stopper then I would suggest that colleges and universities need to look seriously at why overall education costs are through the roof.

The paper concludes:

"The ADALM1000 (M1k) has been shown to provide solid, easily used data for three reasonably typical experimental configurations. Thus, it is expected that it can provide the kind of hands-on exploration that is too often missing in engineering education at a much lower price point than has typically been the case."

A conclusion we of course very much agree with.

The opinions expressed are my own and as always I welcome comments and suggestions from the user community out there on other fun ways to use the ADALM1000 and the ALICE software tools.



Additional paper links: