Skip navigation
All Places > Keysight Blogs > Oscilloscopes Blog > Blog > 2017 > February

What is the best oscilloscope for your application? The following areas will help you make an accurate and informed decision. Today’s complex electronics industries require a broad spectrum of test equipment, with oscilloscopes being one of the most fundamental tools used by engineers and technicians. Oscilloscopes provide design and manufacturing engineers with critical insights to signal properties suggesting additional design work needed, targeting manufacturing issues, or performing compliance and protocol testing per international standards.

Oscilloscopes fall into two groups, real-time oscilloscopes and sampling oscilloscopes (also called equivalent-time oscilloscopes) and it is important to understand the difference between the two types. Real-time oscilloscopes digitize a signal in real-time. Imagine a repetitive AC signal - the real-time oscilloscope acts like a camera, taking a series of frames of the signal during each cycle. The amount of frames the real-time oscilloscope captures depends upon the bandwidth, memory depth, and other attributes that we will soon discuss. A sampling oscilloscope, on the other hand, takes only one shot of the signal per cycle. By repeating this one shot, but at slightly different time frames, the sampling oscilloscope can reconstruct the signal with a high degree of accuracy.

The following topics can help you better evaluate which kind of oscilloscope will best suit your needs.


Sampling oscilloscopes are designed to capture, display, and analyze repetitive signals. If your oscilloscope solution needs to capture a single random event within your waveform, a real-time oscilloscope should be selected. Whether you are looking at intermittent signals during product design or manufacturing, real-time oscilloscopes allow you to trigger on a specific event such as a rising voltage threshold, a set up and hold violation, or a pattern trigger. The real-time oscilloscope will capture and store continuous sample points around these triggers and update the display with the captured data.



The frequency of your signal under test and the harmonics within it will determine the bandwidth of the oscilloscope that will fit your needs. Sampling and real-time oscilloscopes cover a wide bandwidth range and there is a lot of overlap. A sampling oscilloscope can acquire any signal up to the analog bandwidth of the oscilloscope regardless of the sample rate. But a real-time oscilloscope must gather a significant number of samples after the initial trigger to accurately display a waveform. A typical rule of thumb for a real-time oscilloscope bandwidth is 2.5 times your signal frequency to reproduce your signal with the best fidelity. So you can get by with an effectively lower bandwidth scope using a sampling scope as long as you have the trigger mentioned in the previous section.


Memory Depth

Oscilloscope memory depth is an important specification for only real-time oscilloscopes. A real-time oscilloscope captures an entire waveform on each trigger event. To do this the real-time oscilloscope captures a large number of data points in one continuous record. For a real-time oscilloscope, the memory is directly tied to the sample rate. The more memory you have, the more samples (sample rate) you can capture for each waveform.  The higher the sample rate, the higher the effective bandwidth of the oscilloscope.  There is a simple calculation to determine the sample rate given a specified time base setting and a specific amount of memory (assuming 10 divisions across screen): Memory depth / ((time per division setting) * 10 divisions) = sample rate (up to the max sample rate of the ADCs). This memory depth concept does not apply to sampling oscilloscopes because only one instantaneous measurement of waveform amplitude is taken at the sampling instant.


Analog to Digital Converter Bits

Sampling oscilloscopes can have as high as a 14-bit analog-to-digital converter (ADC). Consequently, they have a very large dynamic range, which enables viewing signals ranging from a few millivolts to a full volt without the need for attenuation. This allows sampling oscilloscopes to maintain very low noise levels at all volts per division settings. A real-time oscilloscope is limited in its dynamic range to 8 - 10 bits depending upon the model, but typically will show an effective bit number of around 6 – 8 bits respectively. Because of a real-time oscilloscope’s lower signal-to-noise ratio, it is designed with attenuators to correctly display signals at specific volts per division settings.


Frequency Response

Frequency response is another key consideration in your selection criteria. Sampling oscilloscopes do not use digital signal processing (DSP) correction, so the frequency response rolls off slowly and looks more Gaussian in shape. Real-time oscilloscopes can implement DSP to correct their frequency response. For instance, Keysight’s S-Series oscilloscope has a very flat frequency response across its bandwidth, which means its gain will not vary by more than 1 dB across the entire band.


Clock Recovery

The clock recovery component of an oscilloscope measurement is used for building real-time eyes, mask testing, and jitter separation. A recovered clock is a reference clock within the oscilloscope and used for measurement comparisons. Keysight’s Sampling oscilloscopes provide an accurate software-based clock recovery system. In many applications, real-time oscilloscopes have a software clock recovery and selectable hardware clock recovery frequencies. Please note that the advantage of a software clock recovery is that it is not prone to the hardware errors, and will land its edges where they need to be regardless of the data rate.



Sampling oscilloscopes, like real-time oscilloscopes, offer eye diagrams, histograms, and jitter measurements. With high bandwidths, modularity, and lower pricing, they typically fit manufacturing environments better than real-time oscilloscopes.


Many of Keysight’s sampling oscilloscopes have modular systems consisting of a mainframe and various electrical, optical and TDR modules. This allows the end user to customize measurement hardware to fit their solution. Sampling oscilloscope electrical and TDR channels can be integrated into a module to reduce cost or remote heads can be used to improve accuracy. Optical channels are always integrated creating a well-controlled 4th-order Bessel-Thomson frequency roll-off.


When making jitter measurements clock recovery systems play a large role. Understanding the clock recovery algorithm and the jitter transfer function used will help you determine your final oscilloscope selection. The sampling oscilloscope has a slightly lower jitter and a higher dynamic range making it ideal for characterization in a controlled environment assuming that your signal is repeatable. However, real-time oscilloscopes are great if you need to trigger on difficult to find events. Real-time oscilloscope users can choose from a long list of compliance, protocol triggering and decode, and analysis applications including jitter.


Form Factor

Your solution may require an oscilloscope solution with a specific size or configuration (form factor) to fit your needs. Keysight has both sampling and real-time scope solutions in a variety of form factors, from standard desk top and rack mountable frames to faceless (no screen) module solutions in a variety of AXI or PXI configurations. See the links below for sampling and real-time options.



On the surface there is a lot of overlap between sampling and real-time oscilloscopes but the differences in capabilities and performance that we have discussed can help you make an informed decision to tailor a selection to your specific application.


If you require measurements of a repetitive waveform with lower jitter and a higher dynamic range, a sampling oscilloscope is a good choice. In addition, sampling oscilloscopes have an advantage of a lower initial cost and modular upgrades, making them well suited for electrical and optical manufacturing test applications. Real-time oscilloscopes come in a variety of bandwidths, include the ability to capture single-shot events as well as repetitive signals.


Both Keysight sampling and real-time oscilloscopes are available in frequencies from 1 GHz to 50 GHz and beyond with a variety of modular and frame options to fit your specific requirements.

 keysight oscilloscopes samplingscope

If you didn’t get everything you wanted for Valentine’s Day, check out the latest Infiniium oscilloscope firmware (version 5.75) – it may have what you wished for. Its updates include a front panel macro recorder, the ability to load and save .mat files, multiple undo/redo capabilities, and more!

The front panel macro recorder allows you to record all of your actions with the keyboard, mouse, and touchscreen so that you only have to go through your set ups once – you can save and playback the macro record or load it to be executed as a set of SCPI commands. It retains up to 500 commands.

Macro Recorder

If you use MATLAB, you’ll enjoy the ease of saving waveform data as a .mat file and the ability to open a waveform .mat file as a memory waveform. Remember, Infiniium allows you to open and view up to 8 waveforms at once.

waveform files

Perhaps my favorite addition to the software is the new Undo and Redo capability. If you’ve ever accidentally clicked on a setting that you didn’t like or wish you could go back one step, two steps, five steps, etc. you can now do that with Undo/Redo. You can either step back through your changes one at a time or use the drop down menus to undo or redo multiple steps at once. Too bad we don’t have an Undo/Redo for any Valentine’s Day dates-gone-wrong (unless you’re spending the evening with your oscilloscope – then Keysight has your back)!

Scope controls

If you are testing PAM-4, check out the latest updates to our PAM-4 Compliance Application N8836A. Free trial here. We have added new Continuous Time Linear Equalizer for eye height, width, and symmetry mask width, new J4 jitter support, and PRBS13Q test pattern.

In addition we have added more bit error rates for jitter analysis (J2, J4, J5, and J9) and more hardware serial trigger data rates:

  • 2.4882 Gb/s
  • 3.7125 Gb/s
  • 4.455 Gb/s
  • 4.640 Gb/s
  • 5.5688 Gb/s
  • 5.94 Gb/s
  • 7.425 Gb/s
  • 9.95328 Gb/s
  • 12.440 Gb/s

If any of these look like the Valentine’s wish you were hoping for, update to latest software to your Infiniium oscilloscope & PC, or try the software for free.

Download Infiniium software version 5.75


Scope Month 2017

Posted by KeysightOscilloscopes Employee Feb 15, 2017

It’s almost that time again, we’re only a few weeks away from Scope Month 2017! If you missed out last year, or are just getting ready for this year, here’s what you need to know.


Just like you, we love oscilloscopes. So Keysight created an entire month to celebrate oscilloscopes and the great engineers who use them, that’s you! March 1, 2017 kicks off Scope Month, which will run through March 31, 2017, and will offer new measurement tips, oscilloscope resources, a new Keysight oscilloscope, and of course oscilloscope giveaways!

Everyone’s favorite part of Scope Month is the oscilloscope giveaways, and this year won’t disappoint. For Keysight Scope Month 125 oscilloscope giveawaysScope Month 2017, Keysight is giving away more than 125 oscilloscopes! We will be drawing new winners each weekday during Scope Month and posting these drawings on our YouTube Channel along with a helpful measurement tip for you. You will be able to enter the drawing once per day during Scope Month, and all entries will be eligible for the entire drawing period. And the best part is that you can get an extra chance to win: if you enter now, you get an early entry into the sweepstakes (only one per person during the early-entry period).


Enter now


Any questions? Check out the FAQs on this page. Did we miss anything? Ask your questions in the comment section below and we’ll get back to you.



Keysight Test to Impress contestNot quite ready to leave a new oscilloscope up to chance? You can also participate in the Test to Impress video contest. Just create and submit a video explaining why you need a new Keysight oscilloscope and how you would use it, and you could win one! Eligible entries will be reviewed by a panel of recognized industry voices who will choose 1 Grand Prize winner to win a 6-GHz 6000 X-Series oscilloscope and 2 “Runner up” winners to receive brand new Keysight 350-MHz 3000T X-Series oscilloscopes. Entries will be accepted March 1-31st, 2017 at Be sure to stay tuned after Scope Month, because we’ll announce the winners April 14th, 2017.



Keysight new oscilloscope

But wait, there’s more… this year the start of Scope Month also means a big SURPRISE. We can’t give it away just yet, but we can tell you there’s a brand new oscilloscope coming to the Keysight family and Scope Month will give you the first chance to see it and even get your hands on one for free (and you’ll definitely want to get your hands on one)!


Add the live Scope Month kickoff to your calendar to make sure you’re the first to see it!



While you wait for Scope Month to begin, we have quite a few great resources to help you with your measurement challenges:

  • Oscilloscope Learning CenterQuickly access video tutorials, application notes, white papers, and industry experts. Whether it’s your first time in front of a scope or you’ve been using one for decades, the oscilloscope learning center can help you stay ahead of next technology.
  • Oscilloscope blog – follow us to see a new post each week around topics from new releases to helpful tips to industry news
  • EEs Talk Tech – Join Mike and Daniel for an insider’s perspective on some of the latest technology trends and what they mean for you
  • Keysight 2 Minute Guru – Check out this series of 2-minute videos for tips on making better measurements
  • Digital Design and Test webcast series – Need a little more detail on some of the more complex measurements? Check out our free webcast series and join technology experts for more info on a range of topics


Check them out today!

The New Design and Test Challenges

If you plan on leveraging the work you previously did in your PCIe 3.0 design, you are mistaken. A doubling of the speed from 8 gigatransfers/second to 16 gigatransfers/second has a tremendous impact on both the design and validation of your high speed interconnect technology.


What are the Key Drivers for PCIe 4.0?

Big Data Needs Throughput. Big data is a term for data sets that are so large or complex that traditional data processing applications are inadequate to deal with them. Challenges include analysis, capture, data curation, search, sharing, storage, transfer, visualization, querying, and updating.

Networking Connectivity Applications.   Streaming movies, streaming sporting events, on-demand TV and the multitude of personal videos uploaded and downloaded is growing exponentially. Simply put, PCIe 3.0 cannot keep up with the latest Ethernet specs without increasing the number of lanes that are required. An increase in the number of lanes means increased cost in terms of power, circuit board layout and required components.  

Storage Technology. PCIe 3.0 has been pushed to its limits for SSD (Solid State) storage devices.   Greater interconnect speeds are required to take advantage of latest storage technologies.



Channel Attenuation – Higher frequency means greater channel loss. PCIe 3.0 circuit traces can be made to run up to 16-20 inches if design care is taken. PCIe 4.0, on the other hand, is expected to have a maximum trace length of 10-12 inches. It’s simply impossible to use low-cost FR-4, pass through two connectors, and retain enough signal integrity at these speeds; no matter how robust the transmit and receive equalization schemes at the source and endpoint devices are.

So how do you mitigate the effects of greater channel attenuation?    

The answer is retimer(s). A retimer is actually an extension component or, thought of another way, a smart repeater operating at the physical layer to fine tune the signal. So to achieve 20 inches you can place the retimer at 10 inches from both the transmitter and receiver to ensure the required channel length. The link initialization protocol still negotiates the amount of transmitter de-emphasis (to optimize the receiver equalization), but now this negotiation is done to and from the retimer instead of the transmitter.   Therefore, the Retimer, from a link equalization standpoint, behaves exactly like any endpoint or root complex device. It has an upstream and downstream side so when you boot up it starts the initialization. This essentially doubles the essentially doubling the channel length, but at an added cost.  

Signal Integrity – If you are driving a signal into the channel transmit lane (Tx) and it encounters a change in the impedance profile, it will generate a reflection and, if the return loss of the SERDES is sufficiently high, (meaning poor) it bounces back down the channel. If I have high return loss, then the reflected signal may significantly impact the integrity of the transmitted signal.

So whatever design you use for PCIe 3.0 will not work for 4.0. It is not simply a matter of doubling the data rate, because you now have to meet a more stringent return loss characteristic.   So, you actually have to change the design at the silicon level to effectively give you more return loss at the higher frequencies.

Receiver Calibration – The key challenge is the ever shrinking eye height specification. PCIe 4.0 specifies a minimum eye height of 15 mV after equalization with a maximum bit error rate 1 X 10-12.   You are totally dependent upon your receiver’s ability to maximize the eye height. You now have to calibrate to an even finer grain of detail.


How do you effectively validate and test?

Keysight is the first to market with full support of both PCIe 4.0 TX Tests under the 0.7 version of the specification and includes the extensive reference clock phase jitter tests required under this specification.

The N5393F compliance test software for transmitter testing of PCI Express 4.0 devices allows for BASE spec testing of new PCI Express silicon under the 0.7 version of the PCIe 4.0 specification. The N5393F product supports transmitter testing of speeds up to 16 GBits/s while also supporting intermediate speeds of 2.5G, 5G, and 8GBit/s. In addition, legacy PCI Express 1.1, 2.X, and 3.X transmitter tests are also supported. This software will run on real time oscilloscopes having 12 GHz (or greater bandwidth) including the Z-Series, V-Series, X-Series, Q-Series, and 90000A platforms.

The N5393F also includes reference clock tests defined in the PCIe 4.0 specification, covering phase jitter requirements. Since the vast majority of PCI Express implementations utilize a common reference clock architecture, it is critical to ensure that a candidate reference clock device meets the many different permutations of phase jitter required for each of the 4 data rates. For PCIe 4.0, this represents over 144 different tests that have to be performed on the reference clock.

The N5393F also adds support for transmitter testing over the new U.2 (or SFF-8639) connector. As PCI Express expands its application base to support Solid State Storage Drives (or SSDs), the U.2 connector has been chosen as the main interface used in computer server platforms.

 Keysight PCIe Compliance Application

Easily select any PCIe transmit compliance application.

PCIe Compliance Application Menu

All testing is automated and produces a hyperlinked HTML report file that makes it easy to identify and study any failed or marginal clock jitter parameters.

Learn more about all of Keysight's PCI Express (PCIe) design and test solutions

Alright, this tool may not actually be helpful in your search for dinosaurs if someone really were to figure out how to bring them back, but this makes a great example to help explain the concept of oscilloscope segmented memory.

So, let’s begin our journey to the greatest theme park of all time – we’ll call it “Dinosaur Island” for lack of a better name... Naturally you’re going to want as much footage as possible of your favorite dino so you can always remember this grand adventure (assuming it has a better outcome than in the movie). Let’s say you’re a huge T-rex enthusiast, so you want to capture them (and only them) running around the park each day. To do this, you can set up a video camera in a tree and retrieve the footage at the end of the day.


But when looking at the footage from the first day, not only do you see your beloved T-rex, but you also see all of the velociraptors, triceratops, diloposaurus, and whatever other Jurassic creatures were running through the park that day. Your camera only has so much memory, and it filled up half way through the day. So you’ve wasted the majority of the memory on these other dinos you don’t even care about. You only have two days left in the park and you really want more T-rex footage. What do you do now?


What if you could set up a sensor condition (or trigger condition) that tells the camera to only record when there is T-rex running by? Now, when you collect your camera at the end of the trip, the memory will only consist of T-rex footage! Not only do you save memory and use it more efficiently, but you also save yourself loads of time by not having to sift through all the footage you don’t care about.


Sadly, your amazing trip has come to an end and it is time for you to return to the real world as an engineer. However, you did learn some new skills that can be applied to making oscilloscope measurements. The concept of capturing T-rex footage using a sensor condition directly correlates to using segmented memory on an oscilloscope. Let’s say you have a signal that has infrequent pulses – like an RF burst (image below). There are about 4ms of dead time (miscellaneous dinosaurs) between each RF pulse while the pulses themselves (your T-rex) are about 700ns. If you were to acquire this signal as-is, including all dead time, you would use 0.0175% of that memory capturing the actual pulse (T-rex) and 99.98% of it on that dead time (misc. dinosaurs)! THAT’S INSANE! Almost all of your memory is being used on something that you don’t even want to see.


To solve this problem, you could always just buy an oscilloscope with significantly more memory, but that gets very expensive very quickly. A much cheaper solution would be to utilize the segmented memory tool, which is already integrated into Keysight oscilloscopes. This application comes standard on the 4000 and 6000 X-Series and all Infiniium oscilloscopes, and can be activated via software license on the 2000 and 3000T X-Series. With this application, you can set a specific trigger condition and tell the oscilloscope to only capture the waveform when that condition is met. So, once you set the trigger and segment parameters, your scope will only capture the pulses in the signal and ignore the dead time (image below). This means that 100% of your memory is being used to capture the pulses and 0% capturing dead time. It allows you to capture a long time span while still digitizing at a high sample rate. This way, you aren’t losing any signal detail for those pulses and you’ll be able to make even more accurate measurements.


Keysight Segmented Memory diagram


As I previously mentioned, this method will also save you a lot of time. Once all of the segments are acquired, you can easily scroll through a list of these segments and select which one you want to view (shown below). This list includes a time-tag of each of the segments which will give you insight into the frequency of each of the pulses. You can also view real-time and date information along the bottom of the screen, so you can see precisely when the pulse occurred. When using segmented memory for serial applications, the oscilloscope will automatically provide protocol decode for each of the captured packets.



Segmented memory can be especially helpful for many different applications, such as measuring an RF burst, decoding serial buses, finding glitches in repetitive signals, seeing the timing of single-shot events, the list goes on. This method gives you deeper insight in your design and helps you debug faster.


Want more detail? Check out these resources to understand how the segmented memory application works and how to set it up.

Segmented Memory Application Note

Using Segmented Memory for Serial Bus Applications