Advertisement Feature Cover Story Measurement & verification:


A step-by-step guide I


n oscilloscopes or oscilloscope probes, bandwidth is a measure of the width of a range of frequencies measured in Hertz. Specifically, bandwidth is specified as the frequency at which a sinusoidal input signal is attenuated to 70.7 percent of its original amplitude, also known as the -3dB point. Most oscilloscope companies design the scope/probe response to be as flat as possible throughout its speci- fied frequency range, and most customers simply rely on the specified bandwidth of the oscilloscope or oscilloscope probes, wondering if they are indeed getting the bandwidth performance at the probe tip. Now you can use these step-by-step instructions to simply measure and verify the bandwidth of your probe with an oscilloscope you may already have. To measure the bandwidth of an oscillo- scope probe, a VNA (vector network analyser) is often used, which is usually expensive and difficult to learn. Also, because typical passive probes are high impedance probes that should be termi- nated into 1Mohm of an oscilloscope, it makes the traditional VNA s21 method hard to implement because it is a 50ohm based system.


The other way to get bandwidth is to use a sine wave source, a splitter and a power meter and sweep the response directly. If you do this, you must set this up to run using a remote interface such as GPIB or USB. Doing it manually is very laborious, subject to mistakes, and requires extensive effort every time you want to evaluate a tweak, etc. An easier way of measuring probe bandwidth especially for the lower band- width probes (<1GHz passive probe) is the time domain approach utilising only an oscilloscope with the built-in step signal source, the differentiate function and the FFT. To use this method, your oscilloscope should support the function of another function output. If you can't, an alternative is to pull the time domain waveform data out of the oscilloscope, import it into a PC-based analysis tool such as Matlab, and apply math functions to the step data there.


When you apply a step function to 14


your system, you will get the step response. If you then apply the differentia- tion (or deriva- tive) to this step response, you obtain the impulse response; then take the FFT of the


impulse response to obtain the frequency response of the system. Agilent’s Infiniium real-time oscillo- scope is an excellent tool for this quick bandwidth testing. Here is the step by step procedure of the testing. For this bandwidth measurement example, an N2873A 500MHz passive probe with an Infiniium DSO9404A 4GHz oscillo- scope is used. 1. Use a performance verification fix- ture such as Agilent’s E2655C with a 50ohm BNC cable to connect the Aux output of the oscilloscope to the input of the oscilloscope. The Infiniium oscilloscope has an Aux output port with fast edge speed (~340 psec, 10-90 percent for Infiniium 9000 Series) for probe calibration. It is very important to note that the rise time of the signal source should be faster than the probe’s rise time and the frequency response of the source is reasonably flat over frequency (fig. 1). 2. Connect the probe to the PV fixture to measure one edge of the source. Use as short a probe ground as possible to reduce probe loading associated with ground leads.


Ch 1 (yellow) = Source (Aux output) as loaded by the probe


Ch 2 (green) = the measured output of the probe (Fig 2) 3. Place the rising edges at center of the screen, trigger on the measured output of the probe (ch2) and use averaging, or high resolution acquisi- tion, to reduce the noise on the waveform.


Figure 2 (below): Probing fast edge


Figure 1 (above): Probing 25ohm signal source with the Agilent E2655C performance verification fixture


Jae-yong Chang at Agilent Technologies explores a simple method to verify the bandwidth of your probe and how to maximise on the test procedure


5. Apply the built-in FFT Magnitude function on the impulse response (F1) of the measured step signal. Rescale the FFT to 100MHz/div (the center fre- quency at 500MHz with the 1GHz of frequency span across the screen) and 3dB/div vertically (Fig 4). 6. Now you have a plot of bandwidth. Since the vertical scale of the FFT plot is set to 3dB/div with the horizontal scale set to 100MHz/div. You can see the probe has ~530MHz, as you pick the point in the FFT trace falling by 3dB (Fig 5). There is one catch to this. The way we do differentiate in some of the oscil- loscopes is taking the best fit slope to three adjacent points and then assigning this


slope to the


center point. This can really hose the bandwidth meas- urement up if you


Figure 4 (above): Apply the built-in FFT Magnitude function on the impulse response


don't have enough sample density on the edge, so experiment with sample density and make sure it doesn't affect the bandwidth.


Figure 5 (below: Now you have a plot of bandwidth


Utilising the


built-in mathe- matical capabili- ties available in modern digitising oscilloscopes, it is possible to derive


the fre- quency response


or the bandwidth characteristics of a probe based on the measured step response of a fast step signal.


Among those several test methods, the time domain approach is the easi- est to duplicate without needing expensive test instruments. Agilent Technologies www.agilent.co.uk


Enter 203


Jae-yong Chang is Probes and Accessories Product Marketing Manager at Agilent


DECEMBER/JANUARY 2014 Electronics


4. Use the oscilloscope’s built-in math function


to differentiate the step


response. Now you get the impulse response of channel 2 where the probe is connected. Assign the differentiated output of the step response into the F1 of the oscilloscope (Fig 3).


Figure 3 (below): Use the oscilloscope’s math function to differentiate the step response


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44