Two good test probes are all you need

As with most successful technologies over the past few decades, test equipment has improved rapidly to adapt to ever changing requirements. When you look under the hood of a new car, for instance, you have to marvel at the engineering effort in automotives to get all of that "stuff" to fit and to work in concert. It's not strictly about getting from point A to point B anymore; it's about comfort, safety, ecology, and entertainment. It's different for drivers, too. Like an airline pilot readying for take-off, the modern-day driver has to scan through a mental checklist when jumping into the cockpit. Coordinates plugged into the navigation system? Check. Station selected on satellite radio? Check. iPod? DVD player? Bluetooth devices? Check, check, and check. Driving isn't what it used to be.

Beyond automotives, creating products having a suite of new gadgets and smarts requires increasingly sophisticated power system designs. Ultimately, the design engineer is called on to make accurate power measurements. Sitting in front of an oscilloscope with a current probe attached, the engineer also runs through a mental checklist. Degauss? Check. Auto balance? Check. Jaw open indicator? Thermal indicator? Deskew? Check, check, and check. Making accurate measurements isn't what it used to be either.

Here, though, the goal is to make it as simple as possible. To guard against falling into a testing trap, take an honest look at what you're asking your probes to do. Your best bet might be to adopt the old adage that less is more. So consider using two good probes—a general purpose one with a large dynamic range and sufficient bandwidth, and one with high enough sensitivity and bandwidth to handle peak pulse conditions. Your results and your test bench will be all the better for it.

Four probe specs to consider
Power components have evolved so quickly that connectivity, bandwidth, and dynamic range requirements pose significant measurement challenges. It seems that for each measurement challenge there is a different measurement application. This dilemma is particularly applicable to probing solutions. Current probes are generally equipped to solve specific problems. The probe may have a large dynamic range but not enough bandwidth. Or it may lack sensitivity at the low end. A probe with the capability of providing granularity down to the milliamp may lack dynamic range and may not be equipped to handle large peak pulse currents. There's nothing quite like getting your test environment set up and then having to immediately degauss the probe. With ever-changing design requirements and subsequent changes in test equipment, determining the capabilities and features that you need in a current probe can be confusing.

The objective in probing is to reliably deliver output from the signal source to the input of the oscilloscope with maximum signal fidelity and with as little impact to the original signal as is possible. Often, probes are the most overlooked part of the test solution. In some cases, engineers will grab a probe off the shelf that isn't equipped to accurately capture signals of interest. The danger of using a probe that's not equipped to do what you want is that you'll fall into the trap of believing what you see. If you're using a probe that doesn't meet the specifications required by the device under test (DUT), what you see on the screen may not be what's going on in your circuit.

In your test setup, you'd ideally want a single probe solution that you can hook up to the DUT and get a measurement without having to fight the test equipment. While there are many test solutions available that allow you to "brute-force" a measurement, ask yourself whether you really need an arsenal of test equipment to satisfy your testing needs. Does your equipment seem to be more of a hindrance than a help? How much range do you need? How fast does the probe need to be? What features are going to make your job easier? The four main characteristics to consider are:

•Dynamic range and sensitivity
•Bandwidth
•Usability features
•Reduced clutter on your test bench

Dynamic range and sensitivity
Power designs that are completely predictable do not require test equipment with broad dynamic range. However, most power designs do not reside in perfectly controlled environments and have to deal with high di/dt slew rates, spikes, and in-rush currents. Test equipment lacking dynamic range will misinterpret unpredictable events such as spikes, which will lead to unpredictable results. In some cases, the spike may be clipped and the designer may not even be aware of it.

Ideally, a current probe has broad dynamic range and is sensitive down into the milliamp range. However, there are still decisions to make. The designer must determine if he requires a probe that will perform high-level current measurements or low-level current measurements. Either you decide on a probe with enough sensitivity to provide accuracy down to a few milliamps, or you settle on a probe that will tolerate unpredictable peak pulse currents. The "one probe for each measurement" philosophy is favorable for test equipment manufacturers; but for designers, this may result in wasted time, money, and bench space.

If designers were asked how much sensitivity and range they would like in a probe, they'd probably ask, "How much can you give me?" The answer involves making a reasonable trade-off between sensitivity and broad dynamic range. A general-purpose current probe should provide a broad dynamic range from 5 mA minimum sensitivity to 200 amps peak continuous (500 amps peak pulse). A current probe with higher sensitivity should provide resolution down to 1 mA with 42 amps peak continuous (50 amps peak pulse).

Bandwidth
The high switching speeds of modern semiconductor devices creates a tremendous challenge for capturing events accurately. The combination of fast switching speeds with an increase in power density and performance requirements makes power analysis an extremely difficult task. How fast is fast enough?

The amount of bandwidth required is different for analog and digital measurements. Bandwidth for analog measurements is based upon the highest signal frequency of the device under test. Consider a switched-mode power supply (SMPS) with 100 kHz switching frequency. Non-sinusoidal switching signals contain harmonics that exceed the fundamental frequency of the signal. To accurately reconstruct the switching waveform, the test system needs to have flat bandwidth response out to at least the fifth harmonic. In this case, the measurement system requires 500 kHz bandwidth where the probe attaches to the DUT.

For digital measurements, the required bandwidth is based on rise-time considerations, not the repetition rate. A rule of thumb for determining the rise-time spec for your test equipment is that the combined rise-time of the probe and oscilloscope be five to 10 times faster than the measured pulse. Keep in mind that the amplitude response is attenuated -3 dB (~30 percent) at rated bandwidth. The test system needs a faster rise time to accurately represent the harmonics.

The bandwidth of a current probe is characterized by a frequency derating curve that charts the current versus frequency. The derating curve describes the potential of the probe to accurately measure current for a given frequency. General purpose current probes should have a target bandwidth of at least 20 MHz and a rise time specification of 18 nS. A current probe with higher sensitivity should have bandwidth capability of at least 120 MHz and a rise time specification of 3 nS.

Usability features
Apart from the electrical characteristics of a current probe, what probe features will impact how you take measurements? Consider the features that enable you to more quickly set up your test environment, the features that simplify the process of taking measurements, and overall flexibility.

1. Quick set-up
Consider the features that enable you to more quickly set up your test environment in terms of easy connectivity and seamless connection to the oscilloscope.

a. Connectivity to the DUT
The probe should easily attach to the DUT. The probe's slider jaw should have the versatility to clamp onto bare or insulated wires based upon the corresponding insulation ratings. Here are some recommendations for jaw-size based upon typical power supply voltage ratings:

•General purpose current probes should have a slider jaw with minimum 21 mm-by-25 mm jaw opening and should accommodate insulated or bared wire conductor sizes supporting 600 volts RMS CAT I & II or 300 volts RMS CAT III.
•A current probe with higher sensitivity should have a slider jaw with minimum 3.8 mm-by-5 mm jaw opening and should accommodate insulated wire conductor sizes supporting 600 volts RMS CAT I & II or 300 volts RMS CAT III.

b. Connectivity to the oscilloscope
A current probe should have a simple probe-to-scope attachment, with the probe being directly powered by the oscilloscope. When attached, direct scaling and unit readout should be seamless with the host instrument. Surprisingly, less sophisticated scope/probe combinations are unable to provide the measured current readout in amps. Also, the probe should seamlessly connect to either a 1-megohm or 50-ohm scope input termination. It's frustrating and expensive when you have to purchase an expensive, specialized adapter to plug your 1-megohm current probe into an oscilloscope that has 50-ohm input termination. In general, the probe should automatically change the scope's displayed units to amps, and the A/div readout should automatically scale to the probe's selected range. In addition, a current probe should connect to either a 1-meghom or 50-ohm scope input termination without the need for a special adapter.