Anyone involved in the development of high-tech products should understand the consequences brought about by such items as component die-shrinks, lower voltage levels, faster rise/fall times, higher switching speeds, etc. These elements often mean more diligence and thoughtful attention to detail is required in laying out circuits today, more so than they ever have been in the past. These extra precautions are required to ensure proper circuit functionality and to meet timing, signal integrity and electromagnetic compatibility (EMC) requirements, to name a few.
Have you ever done everything you could to ensure that your design met all the necessary requirements before building an actual prototype, only to discover issues once you had the functioning hardware in hand? These issues could be related to functionality, signal integrity, EMC, or something else, and they could prevent your design from being released to production on time. If you don’t have a way to quickly troubleshoot these high-speed or high frequency issues within your design, you will be left with guessing what could be wrong. You may make changes based on engineering judgment and intuition, but you will likely have to repeat the process until a solution is found.
It’s easy to see that the trial-and-error approach of guessing what is wrong with your design is flawed and consumes a lot of time and resources. Furthermore, upper management may not be pleased if a solution isn’t found quickly enough. To escape this seemingly endless and frustrating cycle, you must be resourceful and use high frequency probing and measurement techniques to gather actual measured data. You’ll need certain knowledge and tools to develop this skill, which will be discussed shortly. But first, let me clarify what I mean when I say “be scrappy.”
A Quick Note on What it Means to “Be Scrappy”
As engineers and technicians working in compliance engineering, we must always deal with constraints. For instance, we don’t always have the ultimate budget that affords us the luxury to buy the best, most expensive new probe or other fancy test equipment. Or we may not have exactly the right hardware to get initial “engineering scan” EMC testing completed on our new widget that’s operating under a tight schedule. Or we might not have enough time to allow us to test everything we want to validate a design. We must “be scrappy” in such situations and determine the best path forward. High-frequency probing is an area where we can “be scrappy” and shine. We don’t always need to buy an expensive probe. When making high frequency measurements, we can build our own that accomplish exactly what we need them to do and when we need them to do it! This is what I mean when I say “be scrappy.”
Basics of High Frequency Measurement Techniques
Before diving too deep into the nitty gritty of high frequency probing techniques, it’s imperative to understand a few basics, as described in the remaining portion of this article. The basics include probe calibration and null measurements and a description of the various types of voltage probes available, including their strengths and weaknesses.
Probe Calibration and Null Measurements
Just like you wouldn’t attempt to drive your car without first ensuring the tires are properly inflated and it had enough fuel or electric charge (if electric) to get you where you want to go, you shouldn’t attempt to take a measurement on a high frequency circuit without first ensuring the voltage or current probes you’re using are measuring adequately. This is a key first step to performing high frequency measurements properly and not doing it is a mistake that I often see even the most seasoned professionals make. Don’t be like the others! Be wise and diligently perform probe calibrations and null measurements for every high frequency measurement you intend to make.
Calibration is typically where you use the 1 kHz square wave output provided on the oscilloscope to “square up” the signal shown on the display by making an adjustment located on the end of the probe. If you move the scope probe to another channel, it is best to re-calibrate it for that specific channel. Don’t assume the calibration won’t change when you move the probe to another channel. If you don’t perform this simple step, you may see overshoot and ringing in your measurement data that really isn’t present in the circuit you’re measuring! These anomalies could thwart your efforts to understand the problem you’re trying to solve. Don’t say I didn’t warn you about the importance of proper probe calibration 😊.
For a voltage probe, a null measurement is a method for determining how much the common mode noise present in the measurement setup is affecting the measurement result. It sounds counter-intuitive, but for a typical scope probe, the null experiment is performed by first tying the ground lead of the probe to the probe tip and then touching the tip to the circuit under test while the rest of the circuit and any auxiliary equipment is operated normally. The amount of voltage picked up is the margin of error in the actual measurements you will perform when using this probe to collect data.
For a current probe, a null measurement has the same purpose as a voltage probe, except it is conducted a little differently. To conduct a null measurement on a current probe, loop the wire to be measured so that it bends back on itself. Where the wire bends, it should sit flush from the opposite side of the probe body and not go all the way through it. As with the voltage probe, the amount of voltage picked up is the margin of error in the actual measurements you will perform when using this probe to collect data.
There are three general types of oscilloscope probes. These include the high impedance (Z) passive, high Z active, and low Z probes.
High Z Passive Probes
High Z passive probes, like the typical 10x probes found in most test laboratories, are the most popular oscilloscope probes available. These probes are characterized by their capacitive input Z and must be compensated to the input Z of the scope used. To fully meet its specification, this type of probe must have a bandwidth that matches that of the scope. High Z passive probes are sensitive to high Z electric field coupling that could ultimately affect the measurement circuit.
High Z Active Probes
High Z active probes are characterized by their low capacitive input (~ couple of 2 pF) and high input resistance (< 1 MΩ); therefore, they are useful where circuit loading must be minimized. Like the high Z passive probes, these types of probes are also sensitive to coupling from high Z electric fields, except to much higher frequencies than is typically seen with high Z passive probes. Other downsides to high Z active probes include lower bandwidth capability for balanced designs (over what is available in passive probes), increased cost, and inability to take the normal handling abuse as typically seen in a normal laboratory environment. Experience has shown that the typical engineer and technician are not too kind to their equipment, especially in the heat of battle (product development or testing), when things get hectic and they are in a crunch for time.
Low Z Passive Probes
Characterized by a resistive input resistance over a wide range of frequencies, low Z passive probes are useful in many measurement situations. If you want to “be scrappy” these types of probes are easy to construct and provide accurate measurements up to 500 MHz. At low frequencies, low-impedance passive probes are not very sensitive to electric field coupling at very low frequencies. It’s interesting to note that > 50 MHz, the input Z of this type of probe usually exceeds that of high Z passive probes!
There are many other elements and depths to high frequency probing not discussed here but are definitely worth further study. These include probe ground lead effects (resonance and induction), high Z passive probe compensation, differential measurements, magnetic loop and other noncontact measurements, current probes (theory and uses) and measurements of pulsed EMI effects on electronic circuits. Please consult reference 1 for more details on these additional topics. Although written 30 years ago, reference 1 still contains valuable information that is even more applicable today.
References and Further Reading
- Smith, D.C., High Frequency Measurements and Noise in Electronic Circuits, 3rd Edition, Springer, 1993.