Pico Technology

By Alan Tong and Mike Green

The majority of today's traditional test and measurement instruments and expensive data logging equipment are destined to become museum pieces as PC-based instruments (or virtual instruments - VIs) make their way into R&D, engineering, industrial applications and ultimately the home.

A few years ago conventional instruments, such as oscilloscopes and spectrum analysers, were available only as expensive, bulky bench-top products. Similarly, data acquisition and logging required bespoke hardware and software solutions, and was often prohibitively expensive for low-end applications.

Then, in the early 1990s, PC-based instrumentation appeared on the scene and test and measurement and data logging changed forever. Data acquisition could now ride on the back of the PC industry, reaping the benefits of low cost and standardised user interfaces. In addition, we saw the dawn of virtual instruments (VIs): comprising software and hardware elements which enable conventional PCs to be used as data loggers, oscilloscopes, spectrum analysers and/or multi-meters.
In it’s infancy, however, VI depended on plug-in cards and skilled IT people to configure the systems, but as the industry went ‘plug and play’ the hardware for VIs became a plug-on module outside the PC and the software became far more user-friendly.

Most of the benefits of PC-based instruments are well known and include portability, large colour displays, fast processors, disk drives and printers. True, all are available on traditional bench-top instruments but only on the high-end (and high price) models or as expensive accessories. Another of VI’s key advantages is that the PC can utilise common hardware for different software tools - in other words it can be more than one instrument at any one time.

Other benefits include the fact that data can be exported to spreadsheets and that data can be ‘pasted’ straight into reports. In addition, with the growth of e-mail and the power of the Internet, engineers can transmit waveforms, traces and readings from site to site, via their portable PC, for further analysis. Indeed, many companies are employing PC-based instrumentation to build up libraries of traces showing typical waveforms for calibration and faultfinding purposes.

The early perception that PC-based instrumentation could not compete with the traditional T&M and data logging methods did not hold water and was soon dismissed as engineers soon realised they were purchasing functionality and not form. Capturing a waveform or analysing a frequency spectrum no longer required the purchase of a bulky piece of bench-top equipment (at considerable cost) and this definitely toppled the age-old model.

Through employing a high-speed analogue to digital converter (ADC) connected to a PC a VI can currently deliver a 100 Msample/s oscilloscope and a 50 MHz  spectrum analyser: more than adequate for the majority of test and measurement requirements and delivered at a fraction of the cost of traditional equipment. The resolution afforded by the latest VIs is also noteworthy; with high resolution (16-bit and above) ADCs able to distinguish a few grammes in a ton, or test high-end (‘golden ear’) audio.

As for the cost issue, traditional instruments like oscilloscopes cost around three times as much as their VI equivalents. The differences in cost can be best explained by looking at a block diagram of a generic instrument - see diagram.

Whilst the cost difference between traditional and VI was perceived as a key issue, the icing on the cake proved to be VI’s ability to be upgraded on the fly. New functions and features come with software upgrades - delivered (in some cases) free of charge. In addition, VIs are typically delivered with software drivers for DOS, Windows, Excel, Visual Basic and now Linux. Hence advanced test procedures and complex capture sequences can be written which are far beyond the button-clicking, menu-trawling capabilities of traditional equipment.

So how does the future look for the manufacturers of traditional test and measurement equipment? Not so good. They have made their beds, and are committed to evolution rather than revolution. Unable to ride on the back of the continuing development of PCs (and low prices through mass production) their solutions end up being low volume (when compared to PCs) products that cost a fortune to develop. Ultimately, there will be little call for ‘run of the mill’ oscilloscopes and analysers and the manufacturers will find themselves specialists in a very narrow market - which, incidentally, is how many started.

True, some of the established players in the T&M markets are beginning to edge closer to PC functionality with the latest products on offer. For example, Hewlett-Packard has chosen to build a PC running Windows 95 into one range of oscilloscopes, and Tektronix has also dabbled with the operating system.

Both players are making much of the familiarity of the Windows environment and the on-line help available with PC-based products. LeCroy, however has taken an even bigger step towards PC-based instruments by releasing a range of PC plug in cards together with oscilloscope software. But the use of plug in cards is unlikely to be the way ahead: early PC-based data-logging products, hard to install and configure, proved that. In addition, one is banking on the availability of card slots in present and future machines. It will be interesting to see how the traditional players react over the next few years with the almost guaranteed growth of PC-based T&M. It will be difficult for them to have feet in both camps without undermining their legacy products.

VI, initially perceived as ‘a neat trick for PCs — but not like the real thing’ has fast become the preferred choice of engineers; be they designers, technicians, field service engineers or scientists. Only a few applications, such as test and measurement for high-end communications, call for functionality outside of the scope of current VI and soon high-end machines will be the only bench-top equipment available. The common or garden bench-top instrument - a bulky piece of kit, unable to be upgraded, share and communicate data, and depreciating in value from the moments it is unpacked - has surely had its day and will go the way of once popular valve: now called upon only in the rarest of moments but otherwise extinct.

The prime users of VI at present are R&D engineers making measurements in order to optimise their designs. For example, many are interested in power quality measurement. For them, it is of ultimate importance to demonstrate that their equipment functions correctly on poor quality mains and that their equipment does not affect the quality of the mains.

The future though will most likely see another group becoming the prime users, as VI moves from the realms of T&M and into appliances and systems. With VI software now supporting Internet Protocol (IP), data can be transferred over a dial up line. This and the fact that there are no limits to the quantities that can be measured (such as temperature, humidity and pH values) means that a raft of remote data logging requirements can be met. The increasing support of the licence free Linux OS means many low cost data logging applications can now be realised and this will surely further fuel the popularity of PC-based data logging applications for both industrial and commercial applications.

Much of today's PC-based data logging technology will become building blocks for homes of the future. These homes will be networked and temperature, humidity and light levels around the home will be made visible to a central (or remote) PC. Even the amount of breakfast cereal in the cupboards could be monitored. In addition, the recent extension of Internet addresses from 32 to 48 bits, means even household appliances could have web addresses and be accessible over the Internet. The day when your fridge automatically orders milk from the supermarket or the toaster detects it is toasting the last slice of bread may not be that far off.

On a more practical note, VI could underpin major improvements in household energy efficiency. For example, today's central heating systems are currently limited by what the user can achieve with the three or four buttons on the controller. Household heating systems of the future could easily be controlled by a computer. A smart system could find the PC monitoring not just temperature but detecting which rooms are in use, working out patterns of hot water usage and downloading weather forecasts from the web to manage the system in an efficient way.

Another huge potential, applicable to industrial applications primarily but also applicable to the home, is power monitoring. Here, there are two main requirements: one is to save money and this can be done by monitoring which units or appliances are using the electricity; the other is by ensuring voltage and current are in phase by adjusting the power factor for the equipment.

Conclusion

Perhaps the one thing that has got the industry to where it is today, is its ability to measure and record progress. Unless we can measure that we are doing better, we are unlikely to be making progress. VIs and PC-based data logging have broken down the cost versus complexity barriers that have, to date, barred progress in many environmental monitoring and test and measurement applications. Conditioning signals outside of the PC has become the norm and VIs and PC-based data loggers will make full use of new interconnect standards, like USB.

VI and PC-based data logging, which has been warmly welcomed throughout the engineering community by young and PC literate engineers from virtually all disciplines, are destined to become the de facto tools of the trade for engineering in the 21st Century.

Block Diagram

With a traditional instrument, a large part of the cost is in the components such as displays, keyboards, processors and memories. With a VI, the cost of these elements is ‘mopped up’ by the high volume PC market.

With popular instruments such as oscilloscopes, the cost difference between traditional and VIs is about 3:1. With more specialised instruments such as audio spectrum analysers (which are produced in smaller volumes) the cost difference is around 10:1.

A Sound Idea

A prime example of where VI comes into its own is audio spectrum analysis. Spectrum analysers tend to fall into two categories: ‘swept’ spectrum analysers and Fast Fourier Transform (FFT) based spectrum analysers. Swept spectrum analysers work by using one or more notch filters (or mixers) to measure the signal amplitude at a given frequency, by changing (or sweeping) the frequency of this filter a plot of amplitude against frequency can be constructed. They still have their place in high frequency spectrum analysis, but for audio work they have the disadvantage that the signal amplitude must be constant for the whole period of the sweep.

FFT based spectrum analysers work by digitising the signal of interest using an ADC and the stored values are then processed using the FFT algorithm. The advantage of this method is that the spectrum of one-off or short duration events can be captured. Performing the spectrum analysis requires many calculations and some FFT-based analysers can take several seconds to update a trace.
Pico Technology's PicoScope product (software) uses an optimised, high speed routine for spectrum analysis in ‘real time’. Even on a relatively modest PC, the spectrum analyser can still update many times a second. As for the hardware, most of Pico's ADC range can be used for audio spectrum analysis and for high-end professional testing the ADC216 is hard to beat.

The two key specifications for any FFT analyser are sampling rate and dynamic range. A spectrum analyser will be able to display up to one half of the maximum sampling rate. To cover the entire 20 kHz  audio band this calls for a sampling rate in excess of 40 ksamples/s (the ADC-216 samples at over 300 ksamples/second). The dynamic range of the spectrum analyser is the next most important consideration. Most oscilloscopes (whether PC-based or bench-top) have an 8 bit resolution (256 steps). This limits spectrum analysis to 48 dB of dynamic range (20log256). The ADC216 with its 16 bit resolution (65536 steps) uses over-sampling techniques to achieve 100 dB of dynamic range. To put these figures in context, a typical tape deck would have 40 to 50 dB of dynamic range, a quality power amplifier 70 to 80 dB and a top-end CD player 80 to 90 dB.