Keeping Technology Current in the Changing World of Power Electronics

Dr. Chris Hewson, CEO at Power Electronic Measurements (PEM) Ltd

Dr. Chris Hewson

‘Why isn’t there a simple to use wide-band current probe for power electronics?’ was the question that two successful power electronics professors, turned entrepreneurs, asked back in 1991. The two colleagues began to scratch an itch that had been a background irritation for over 30 years.
Rogowski coils seemed a suitable candidate. The technology first described in a paper in 1911 was potentially wide-bandwidth, had limitless current capability independent of the size of the probe and could be built with a thin clip-around coil making it easy to get a measurement in difficult to reach parts of a power converter.  Some innovative design, a first patent, a student project, a technical paper and a successful prototype later the two colleagues formed Power Electronic Measurements (PEM) Ltd,  wondering if this technology would be of interest to any other engineers in the field of power electronics? 

What they didn’t foresee is that 25 years later PEM would still be at the forefront of the design and manufacture of wide-band current sensors. As this technology developed the power electronics industry continued to grow beyond all recognition. Power electronics is now embedded in power utility and transmission, renewable power generation, industrial processes such as induction heating and welding, traction applications and electric vehicles.  Somewhere in the process of design and development, converter control, product approvals or diagnostics, current measurement is required and each application presents its own distinct set of challenges to the current sensor manufacturer.

Rogowski technology is just one among many methods of current measurement competing for the power electronic engineer’s attention. Current transformers (CTs) offer very high bandwidth and isolation, combination Hall effect and CT current probes offer very bandwidth and DC current capability, shunts range from low cost highly accurate devices to co-axial shunts offering GHz capability and Fluxgate sensors, though bandwidth limited, offer exceptional accuracy. Each has their place in power electronics. However, manufacturers of all these technologies are challenged with keeping their technologies relevant in a world in where power electronics increasingly infiltrates all areas of engineering and commands vast budgets ushering in rapid change. 

One such example of rapid change in power electronics is the challenge posed by the introduction of GaN and SiC switches. For Rogowski probes these new semiconductor technologies help drive new product improvement. Twenty five years ago a 10mm thick, un-screened, semi rigid coil, with a 1MHz bandwidth was state-of- the art. Now, engineers using such a probe to measure a current transient in a SiC device would observe a burst of noise on an oscilloscope rather than an accurate representation of a current waveform. Rogowski probes still offer ease of use, and incredibly low insertion impedance, but to stay relevant significant improvements have had to be made; 

  • Rejection of common mode interference from close coupled capacitive interference due to voltage transients >> 20V/ns, whilst retaining high frequency capability and accuracy with conductor position in the sensor loop
  • Maintaining accurate operation from -40 to +125degC with repeated thermal cycling
  • Measurements of rise times of 10 to 50ns with good fidelity and predictable delay for power measurements
  • Significant size reduction as semiconductor package sizes shrink

But of course there is still much to do, GaN demands even higher bandwidths and smaller coils and these solutions can’t take 25 years as the pace of GaN take up accelerates.

However, it is not just a case of smaller and faster. Other measurement problems can occur, often unforeseen by engineers pushing for ever more performance. For example, over the past 15 years in motor drives the installation of VSDs with faster switches operating at higher voltages has had unforeseen consequences for bearings. Harmonics created by these switching phenomena can induce high frequency leakage currents in the machine shaft causing the bearings to heat and lifetime to rapidly degrade. Tracking these leakage currents is not easy.  There is interference from strong magnetic power frequency fields and a large machine shaft requires a sensor with a very big aperture that retains the ability to measure spikes of current of a few Amps with sub µs duration. Difficult, but ideal for a carefully designed Rogowski probe. As measurement manufacturers we are always listening to the problems of engineers and advising on how our technology can solve a given measurement problem. 

Once that first ‘production’ Rogowski probe was built and began to sell, we sat discussing our future and briefly considered what technology we should tackle next. Ideas about speed sensors were considered – perhaps commercialising some of the large current test sets that we had developed? But it didn’t take long before power electronic engineers began to explain their particular set of challenges,  ‘could you customise your probes for induction heating, we have a protection application that needs volume but lower cost, we have a new semiconductor device that needs a higher bandwidth’, were some of the challenges. Twenty five years later and we’re still pursuing smaller, faster more accurate Rogowski probes but continuing to listen to engineers explain their new applications and directions to take this flexible technology……and we still haven’t got around to developing those speed sensors.  

Source: Bodo's Power Systems, October 2016