What is a Floating Measurement?
A floating measurement reads the voltage between two points, neither of which is at ground potential. Floating measurements of fast AC signals are a challenge for conventional instruments. They have unique requirements over and above the usual considerations of bandwidth and resolution. Foremost among these issues is operator safety. In addition, functional concerns such as the form factor and cost of the measurement tools (many of which are designed for cost-sensitive field service applications) cannot be compromised.
A third, more subtle element is measurement quality. True and complete isolation of the measurement instrument from the ground connection of the unit-under-test is essential. Only then can the instrument produce accurate and repeatable ungrounded measurements.
A variety of solutions for floating measurements exists. This technical brief examines the available alternatives for measuring AC signals in an ungrounded environment. It will show how the balance between three characteristics safety, packaging, and performance determine an instrument's effectiveness for making floating measurements.
Where are Floating Measurements Needed?
The most demanding floating measurement requirements are found in power control circuits (motor controllers, uninterruptible power supplies, and switching DC power supplies) and industrial equipment.
Power control technologies use both high-power silicon components and low-power logic circuits. The switching transistors at the heart of most power control circuits, usually MOSFETs or IGBTs, require measurements not referenced to ground. Moreover, the power circuit may have a different ground point (and therefore a different ground level) than the logic circuit, yet the two often must be measured simultaneously.
In industrial applications, it's common to find two machines connected to different power distribution circuits with different ground reference levels. Sometimes it's necessary to measure one machine with reference to the other, yet it's unwise to connect the two grounds together the scope chassis becomes the conductor through which the two ground circuits try to equalize. The resulting current flow is limited only by the resistance of the connecting cables.
In all of these application areas, voltages and currents may be large enough to present a threat to users and test equipment.
The Three Facets of Floating Measurements
Instruments aimed at floating measurement applications must balance three facets: performance, form, and safety. The interplay between these three attributes defines the character of the power measurement tool; its usefulness, its versatility, and its cost-effectiveness.
Figure 1. Every instrument design involves conscious tradeoffs among the three facets of functionality.
Performance. The tool must have frequency response that matches the target application. Here, frequency response denotes both bandwidth and sample rate. While the gross bandwidth of most instruments is adequate for basic power measurements, fast signal transients and erratic glitches may require single-event capture at very high sample rates. Sample rate is also critical when the instrument must capture signals on two channels simultaneously.
High bandwidth and sampling rate usually come at a price: fast circuits generate more heat. This may not be a problem in full-size benchtop instruments, but in some packages especially compact handheld tools that heat is confined within the instrument's housing. Battery-powered devices face the additional challenge of minimizing the power consumption to conserve battery life.
Historically, handheld instruments have sacrificed performance in the name of heat and power conservation. Unfortunately this limits the utility of such instruments to lower-frequency applications.
Form. In keeping with the adage "form follows function," the instrument's form, too, must be tailored to the application. Weight and size, ruggedness (especially impact- and moisture-resistance), control layout, and display characteristics all play a role in the instrument's usefulness. If the tool is to be used in the field, it may need to include both battery- and line-operated modes.
The advent of switching power supplies and motor controllers brought a need to characterize waveshapes, timing, distortion, and other dynamics. Power measurements are no longer the simple DC or low-frequency AC readings they once were. Today's power measurements call for an oscilloscope but unfortunately the most common type of scope, the line-powered benchtop model, lacks a safe grounding scheme for floating measurements. And many compact battery-powered scopes have inadequate bandwidth and sample rate for accurate high-frequency waveform capture.
Safety. In floating measurements, isolation and insulation are the keys to safety. Isolation of input channels and their common leads, isolation of the instrument's own chassis ground from earth ground, and insulation of the housing and controls all combine to protect the user (and the equipment) from electrical shock.
Inside any instrument, conductor placement, length, and spacing are influential factors in the tool's bandwidth. In power measurement instruments, though, a certain minimum physical separation of these features is necessary to prevent dangerous arcing. Increased spacing means increased conductor length, which increases both inductance and stray capacitance, in turn degrading the bandwidth. Many power measurement tools trade off needed bandwidth for added protection against arcing at high voltages.
Table 1 summarizes the instrument characteristics that must be evaluated when planning for floating measurements.
| Table 1: Attributes That Make up the Three Facets of Floating Measurement Capability | ||
|---|---|---|
| Performance | Form | Safety |
| Real-time bandwidth | Portability (size) | Insulation |
| Sample Rate | Weight | Isolation from ground |
| Accuracy | Probing solutions | Isolation between channels |
| Channel count | Cost | Voltage rating |
| Triggering capability | Ease of use | UL/IEC1010 certification |
Choosing a Measurement Solution
The first step toward solving floating measurement problems is to choose among the available types of tools:
Battery-Powered DMMs. Within their bandwidth limits, battery-powered handheld digital multimeters (DMMs) are well-suited for many floating measurements. Their insulated housings make them safe to handle, and because they are not line-powered, these DMMs provide good isolation from earth ground. Even within their limits, though, some DMMs are susceptible to "parasitic" capacitances if their common lead happens to be connected to the active side (as opposed to the neutral side) of the measured signal. (Parasitics are a normal consequence of the interaction between two conductive masses placed in proximity to each other, in effect forming a rudimentary capacitor. Parasitics can cause erratic, unstable measurements.)
Battery-powered DMMs emphasize the form facet (portability and cost), trading off performance benefits especially bandwidth and signal quality that are needed for complex floating measurements.
Line-Powered Benchtop Oscilloscopes. Line-powered oscilloscopes are attractive because they have all the measurement performance needed for the job. Equally important, every lab, manufacturing line, and service shop already owns one or more! However, most line-powered oscilloscopes have a fixed chassis ground terminal as one of their two measurement contacts. Chassis ground is connected to earth ground through the instrument's power cord. This feature alone makes the standard general-purpose benchtop scope unsuitable for floating measurement applications. Unless steps are taken to isolate or "float" the scope, any attempt to connect the probe will create a short circuit.
A common, but risky, practice is to disconnect the scope's AC mains power cord ground and attach the probe ground lead to one of the test points. Tektronix strongly recommends against this unsafe measurement practice. Unfortunately, this puts the instrument chassis which is no longer grounded to earth at the same voltage as the test point. The user touching the instrument becomes the shortest path to earth ground. Figure 2 illustrates this dangerous situation. V1 is the "offset" voltage above true ground, and VMeas is the voltage to be measured. Depending upon the Unit-Under-Test (UUT), V1 may be hundreds of volts, while VMeas might be a fraction of a volt.
Figure 2. A floating measurement in which dangerous voltages occur on the scope chassis. V1 may be hundreds of volts!
Lifting the chassis ground in this manner threatens the user, the UUT, and the instrument. In addition, it violates industrial health and safety regulations, and yields poor measurement results.
Moreover, line-powered instruments exhibit a large parasitic capacitance when floated above earth ground. As a result, floating measurements will be corrupted by ringing (see Figure 3).
Figure 3. Ringing caused by parasitic inductance and capacitance distorts the signal and invalidates measurements.
Line-powered standard oscilloscopes emphasize the performance facet (bandwidth, versatility), trading off the ability to make floating measurements.
Benchtop Oscilloscopes with Differential or Isolated Probes. Differential or isolated probes offer a safe and reliable way to adapt a line-powered scope to make floating measurements. Neither of the two probe contacts need be at earth ground and the probe system as a whole is isolated from the scope's chassis ground. The probes are far less sensitive to lead length than standard probes; two widely separated points can be probed with minimal degradation of performance.
Differential probes add a layer of cost and complexity to the measurement apparatus. They may require an independent power supply and their gain and offset characteristics must be factored into every measurement.
Differential probe-equipped scopes emphasize the performance and safety facets (bandwidth, isolation), trading off form-factor benefits such as portability and cost.
Battery-Powered DSO. The most versatile tool for acquiring time-variable signals either periodic or intermittent is the digital sampling oscilloscope (DSO). The DSO's capacity to provide the necessary measurement quality, plus automation and documentation features, is well-established.
Handheld battery-powered DSOs have the potential to achieve the best balance of the three facets. They combine all of the attributes needed for effective floating power measurements: they are portable, well-isolated from earth ground, capable of both AC and DC measurements, and are cost-effective.
Therefore the handheld, battery-powered DSO is the recommended choice for many power measurement applications.
Measurement Quality Considerations Help Select a DSO
Underlying the Performance/Form/Safety issue is the concept of measurement quality. A host of technical aspects contribute to measurement quality. Both electrical and architectural issues are involved:
Bandwidth. Bandwidth is critical because even low-frequency signals may have high-frequency components. For example, when high-voltage signals are switched by a MOSFET power transistor, large transients overshoot or undershoot can occur. These transients embody harmonics many times the nominal frequency of the fundamental signal, and can affect the operation of the power supply or machinery. Low-bandwidth measurement tools are unable to detect wideband harmonics.
Sample Rate. Sample rate is to DSOs what bandwidth is to analog scopes. And like bandwidth, higher sample rates are better at capturing fast signal details. For regular, repetitive signals, the equivalent time sampling approach is adequate. The scope samples a succession of waveform cycles and reconstructs them as one composite waveform image. Details that vary from cycle to cycle are lost or are displayed in distorted form.
A more versatile approach one that doesn't require a repetitive signal is Digital Real-Time (DRT) sampling. DRT samples at a rate several times that of the input frequency, acquiring enough points from each cycle to reconstruct it faithfully. The advantage of this architecture is its ability to capture fast signal edge details and non-recurring transients (see Figure 4). This is common in field measurement situations, where signals may originate in the coils of a motor, for example, rather than a controlled, calibrated waveform generator.
Figure 4. Equivalent-time sampling (top) requires multiple triggering and may not accurately capture a signal that is shifting in time or amplitude. Digital Real-Time sampling captures the entire signal in one trigger event and accurately captures irregular signals.
Probing Architecture. A scope is actually a measurement system consisting of preamplifiers, acquisition/measurement circuits, displays, and probes. The role of the probe is sometimes overlooked. Nevertheless, improper probes or probing techniques can affect the measurement outcome. Obviously, it's essential to use compatible probes that match the instrument's bandwidth and impedance.
Less understood is the effect of ground-lead inductance. As lead length increases, parasitic inductance increases (Lparasitic in Figure 5). Lparasitic is in the signal path and forms a resonant LC circuit with the inherent parasitic capacitance of the scope (Cparasitic). As Lparasitic increases, the resonant frequency decreases, causing "ringing" (see Figure 3) that visibly interferes with the measured signal. Simply stated, the common lead must be as short as physical constraints of the circuit under test will allow.
Figure 5. Parasitic inductance and capacitance can affect measurement quality.
In regard to capacitance, even insulated, battery-powered scopes exhibit capacitance with respect to earth ground. In Figure 5, Cparasitic describes the scope's parasitic capacitance from its ground reference (through the insulated housing) to earth ground. Like parasitic inductance, Cparasitic must be kept to a minimum in order to force the resonant frequency of the LC circuit as high as possible. If Cparasitic is large, ringing may occur within the test frequency range, hampering the measurement.
An instrument's parasitic capacitance to ground is dictated by its internal design. The physical environment can also prompt ringing. Holding the instrument or placing it on a large conductive surface during measurements can actually increase Cparasitic and lead to ringing. For extremely sensitive measurements, it might even be necessary to suspend the oscilloscope in mid-air!
Insulation and Isolation
Many battery-powered instruments have "insulated" architectures that package a conventional, single-point-grounded scope in a plastic case. Although the insulation offers some protection, the system may still have some exposed points that could deliver a shock when touched. Figure 6 illustrates an insulated scope architecture. If, for example, the ground lead of Channel 1 is connected to a +100 V test point, the ground leads of both channels would be raised to that level because they are connected to the same point within the scope chassis. And connecting Channel 2's ground lead to any other measurement point would short-circuit the Channel 1 test point, with commensurate risk of shock or arcing. At the very least, the single-grounded connection would drastically affect the measurement on Channel 2.
Figure 6. Even "safety-insulated" instruments may lack isolated input channels and grounding schemes.
The insulated architecture also suffers from bandwidth limitations caused by parasitic capacitance. The relatively large mass of the measurement module, even though it is encased in plastic, interacts capacitively with its surroundings. This degrades signal quality and measurement accuracy.
Isolation, in addition to insulation, is the preferred solution for both measurement quality and safety. True channel-to-channel isolation minimizes parasitic effects; the smaller mass of the measurement modules is less prone to interaction with the environment. A properly isolated battery-powered instrument doesn't concern itself with earth ground. Each of its probes has a "common" lead that is isolated from the instrument's chassis, rather than a fixed ground lead. Moreover, the "common" lead of each input channel is isolated from that of all other channels. This is the best insurance against dangerous short circuits. It also minimizes the signal-degrading lead inductance that hampers measurement quality in single-point grounded instruments.
A Solution for Floating Measurement Problems
Recent advances in sampling architecture and semiconductor technology have spawned a generation of handheld, battery-powered DSOs that are well matched to power measurement application needs. Foremost among these is the TekScope® family of instruments from Tektronix.
Designers of the TekScope family have balanced the performance, form, and safety facets while offering the highest measurement quality of any instrument of its type. Its patented IsolatedChannel™ architecture provides true and complete channel-to-channel isolation for both the active and the common leads. Figure 7 illustrates the IsolatedChannel concept. The TekScope family is packaged in a compact insulated housing that further protects the user and the equipment. Its 1.5 kg weight is convenient for both field and factory applications.
Figure 7. TekScope's IsolatedChannel™ architecture provides complete isolation from dangerous voltages.
Equally important, the TekScope family offers the only handheld scopes with Digital Real-Time acquisition. The 1 GS/s sample rate of DRT is the key to the extraordinary bandwidth 200 MHz in the THS730A. This bandwidth/sample rate combination makes it easy to capture the high-frequency information (glitches, edge anomalies, etc.) that eludes other handheld instruments.
While all TekScope models are safe and suitable for power measurements, the unique THS720P model adds important features optimized for the purpose. In addition to the family trait of unsurpassed channel-to-channel isolation, the THS720P offers automatic harmonic analysis, phase measurements, and a host of pre-configured power measurements that may be selected from menus. Its range of triggering functions includes a "Motor Trigger" that responds to the conventional 3- and 5-level PWM (pulse width modulated) power signal. Lastly, the THS720P is equipped with a pair of 1 kV high-voltage probes that accommodates the vast majority of power measurement needs.
The channel-to-channel isolation of the THS720P provides a real-world measurement advantage in addition to its obvious safety benefits. Figure 8 is a screen image depicting waveforms taken at two different points in a power control circuit. Notice that the upper waveform is about 200 Vp-p, while the lower trace is less than 1 V. Because each of the two THS720P channels is fully isolated from the other (including the common leads), and equipped with its own uncompromised DRT digitizer, there's no crosstalk between the two signals. Were the two scope channels not adequately isolated, there might be misleading artifacts coupled from the 200 V signal to the smaller waveform; these might be misinterpreted as a circuit problem when in reality it's an instrument problem. The THS720P's ability to discretely capture two waveforms of vastly differing amplitudes reduces guesswork and improves productivity.
Figure 8. The THS720P's channel-to-channel isolation eliminates crosstalk effects when large and small signals are captured simultaneously.
Harmonics Measurements Reveal Unseen Power Problems
An understanding of the harmonics within a power grid is essential to the safe and cost-effective use of electrical power. Line harmonics are a growing problem in a world moving increasingly toward nonlinear power supplies for most types of electronic equipment. Nonlinear loads such as switching power supplies, unlike their analog predecessors, tend to draw non-sinusoidal currents. Their impedance varies over the course of each cycle, creating sharp positive and negative current peaks rather than the steady curve of a sine wave. The rapid changes in impedance and current in turn affect the voltage waveform on the power grid. As a result, the line voltage is corrupted by harmonics; the normally sinusoidal shape of the voltage waveform may be flattened or distorted.
There's a limit to the amount of harmonic distortion that equipment can tolerate. Load-induced harmonics can cause motor and transformer overheating, mechanical resonances, and dangerously high currents in the neutral wires of three-phase equipment. In addition, line distortions may violate regulatory standards in some countries.
The THS720P has comprehensive built-in facilities for measuring and analyzing line harmonics. Its "Harmonics" mode invoked with a single button captures the fundamental frequency plus harmonics 2 through 31. Using only the scope's standard voltage probe, it's possible to execute a harmonic voltage measurement. An optional current probe acquires current harmonics with the same ease. Using the voltage and current probes together, along with the THS720P's MATH command, provides power statistics readings.
Figure 9 is a THS720P screen image that documents a harmonic voltage measurement. The amplitudes are computed by the instrument's internal DFT (Discrete Fourier Transform) algorithm. In this case the bar graph reveals a very strong third harmonic level. Excessive third harmonic levels (along with certain other odd harmonics) are a classic cause of neutral-wire currents in three-phase systems.
Figure 9. The THS720P Harmonics display reveals potentially damaging signal content.
Power Readings More than Just Watts
Voltage and current measurements are by nature straightforward and absolute. A test point has only one voltage and one current value at a given instant in time. In contrast, power measurements are voltage-, current-, time-, and phase-dependent. Terms like "reactive power" and "power factor," which were devised to characterize this complex interaction, are not so much measurements as computations.
The power factor is of particular interest in these computations. This is because many electrical power providers charge a premium to users whose power factor is not sufficiently close to 1.0, the ideal value. At a power factor of 1.0, voltage and current are in phase. Inductive loads - especially large electric motors and transformers cause voltage and current to shift phase relative to each other, reducing the power factor. Some utility companies apply a surcharge in such cases because the inefficiency causes energy loss in the form of heat in the power lines. There are procedures to remedy power factor problems, but first the power characteristics must be quantified.
The THS720P embraces a full suite of power measurements. Among these are True Power, Apparent Power, Reactive Power, Phase between Voltage and Current, and of course Power Factor. Figure 10 shows a THS720P screen image summarizing these and other power measurements. The measurements require a current probe and a voltage probe working in tandem, and employ the instrument's one-button MATH function. Note that each reading is expressed as an instantaneous value (above the box) and also as a running average, maximum, and minimum value (within the box). The running figures are useful for longer-term monitoring to detect power fluctuations and surges. This feature might be used to gather data on random events (say, in the middle of the night) that are causing equipment shutdowns.
Figure 10. The THS720P power measurement statistics display shows a comprehensive analysis on one screen.
Specialized Triggering Simplifies AC Motor Drive Tests
The THS720P's Motor Trigger is unique in its field. This feature is designed to simplify the task of capturing the drive signals used in pulse-width modulated (PWM) variable-speed AC motor systems. Commonly, the last stage in a motor drive circuit is an IGBT responsible for converting an internal high-voltage DC bus into a synthesized AC waveform whose frequency controls the motor speed. The Motor Drive Trigger makes it easy to stabilize these complex waveforms. Equally significant, the THS720P's thorough isolation ensures operator safety during this high-voltage floating measurement. Using appropriate horizontal magnification and timebase delays on the THS720P, it's possible to examine individual pulses in the motor control waveform. This easily reveals signal aberrations, missing pulses, frequency and phase anomalies, and timing errors that can cause erratic-even catastrophic-motor behavior.
Conclusion
Every TekScope model delivers safe lab-quality measurements on power circuits ranging from computer power supplies to motor controllers. With its unsurpassed bandwidth, portability, and isolation, the TekScope family is a solution that embodies all three facets of the power measurement equation: Performance, Form, and Safety.