8 Replies Latest reply on Apr 21, 2014 5:20 PM by jmarrina

    Oscilloscopes: Past, Present and Future


      Hello Folks,

      I'm John Marrinan, and I'll be your visiting expert this week on all things oscilloscopes. I work as a field application engineer for Tektronix in the UK. My core job is to support our customers for complex time domain applications such as high speed serial analysis, embedded system debugging, optical telecommunications and high energy physics. I'd like to start though by giving a brief history of scopes, the role the play in society and what I think the future holds for oscilloscopes.


      The traditional scope as we know it was based on Cathode Ray Tubes (CRTs) invented in the late 19th century by Karl Braun.  These replaced order oscillographs and galvanometers giving clearer and more efficient views of voltage over time. CRT based scopes used electrically charge, beam forming plates  to excite a phosphors coated screen with in incoming oscillating signal. By sweeping one of these sets of plates we could steer the beam across the screen, drawing the waveform. In the early 1920s they became much more common in research labs. After World War II Howard Vollum and Jack Murdock, the founders of Tektronix, invented a "triggered" giving the ability to steadily display repeating signals.

      The next major advancement in scopes was the introduction of the digital storage scope in the mid-80s by Walter LeCroy. Digital scopes allowed for higher bandwidths, complex triggering, single shots and post-processing. Digital scopes are fundamentally based on A/D converters and Nyquist theorem. This technology is still the backbone of scopes today and will be into the future.


      Today the drive in the scope market leverages improvements in signal process to integrate more of features and functionality. Logic analysis, protocol decoders, spectrum analysis are just some of these features we see today.  There still is incessant drive for more bandwidth at the bleeding edge, with applications should as coherent optics pushing bandwidths to 65GHz today, and even 70GHz or 100GHz in the near future. At these speeds T&M companies are lusing cleaver techniques such as digital bandwidth interleaving or asynchronize time interleaving to break through the limitations of A/D and Pre-amplifier technology. The bulks of scopes and scope users don’t need these bandwidth of course, but in the middle-ranges multi-domain features are such as combined protocol, scope and spectral analysis are now needed by embedded engineers to quickly integrate wireless chip sets into our ever connect world. Other trends I see are for more channels (4 phase buck converters), or modularity and shrinkage through PXI and USB.

      The oscilloscopes really touches every aspect of modern life whether we know it or not. Whether it is being used as an embedded system debug tool in the design of the latest smartphone or if it is assisting in the creation of defence technologies or being used in coherent optical research for the delivery 100G , 400G or 1Tb communication links used for long haul telecommunications. The ever bandwidth hungry world of Facebook, Google and Netflix to name just a few require more and more capacity in the backbone. Research in fibre, photonics and DSP essential to this industry heavily relies on data acquisition from scopes.

      The other trend in our lives is RF Everywhere. From the Internet of Things to Smartphones to Wireless access we all embrace and rely on connectivity. Bluetooth, Zigbee, RFIB and WLAN chipsets are becoming more and more pervasive in embedded system. The challenge for digital design engineers is to integrate the scary world "dbs" and "phase noise" into their "bits" and "bytes" product. Integrated scopes such as mixed signal oscilloscopes or mixed domain oscilloscopes, give engineers the tools to view the interaction between signals and system in both the time and frequency domain.


      What's the future for scopes?! Your guess is as good as mine. What I certainly think you will continue to see is more integration of boxes on your lab bench into a single product, and I believe that this product will fundamentally be a scope. There will be more miniaturisation and improved performances in modular based scopes. And there will be a greater influence by the consumer world on the test and measurements world. By this I mean smartphone type interfaces on scopes, instrument control though tablets, and more customisation with 'widgets and  apps'.

      At the very highest end of research things will stand still and there will be no change at all in my opinion…they'll just want higher bandwidths, faster sample rates and more accuracy like they've always done!!


      I would love to hear your feedback and comments on any of these topics over the coming weeks


      I'll also be talking about the following topics this week and next

           1. So Much More For The Student, Hobbyist and Lab Bench. How Entry Digital Scopes Now Give A Raft Of Features

          2. Integration, Integration, Integration. The Drive To Integrate More Bench Instruments Into The One Box

           3. All You've Ever Wanted To Know About A Modern Scope's Architecture








        • Re: Oscilloscopes: Past, Present and Future

          Hi John,


          I look forward to your posts.

          I got my hands on my first Oscope back in the early 1970's and found it an invaluable tool.

          In the mid 1980's I programmed one of the first digital oscilloscopes for Radar signal analysis.

          The whole line of modern scopes is just amazing to me.  The basic features are very powerful and easy to use.



          • Re: Oscilloscopes: Past, Present and Future

            I agree with DAB, they have certainly come a very long way since the weak/faint trace across the screen (usually green).


            While a small percentage of users need all the features (along with a desire for more bandwidth) most times the requirement is much less.

            Is there a market for feature based pricing using firmware upgrades to get the extra features.?


            The reason I ask is that the modern DSO is realtively highly priced for new engineers or home hobbyists.

            They tend to aim for a logic analyzer when the use of a DSO could improve/assist their design or project.


            Access to a tool that can grow as their needs grow would encourage more use and less 'luck' in some cases.




              • Re: Oscilloscopes: Past, Present and Future

                Hi Mark,

                Thanks for the reply.

                Feature based pricing is already part of the market but it's not available across the board. In the performance space users can purchase extra hardware features such as record length, sample rate or triggers. They typically can also purchase software tools such as USB or Ethernet compliance software. A lot of the products on the market focused more at the industrial design labs also allow you to buy trigger and decode options for serial bus standards like I2C or RS232. Or power analysis tools compliance to IEC standards. And, a lot of scopes today will also offer bandwidth upgrade options (where the sample rate will allow), or the ability logic analyser capabilities. In all these cases hardware features are usually designed in up front and then unlocked as the user needed. Just as you suggest.


                However I think in the entry level the game is a little different. Frankly, cost is extremely important! Putting too many unused features into these models pushes up the build cost. The market is full very good companies serving the market really well and making it a very competitive space. But digital processing and integration is driving change here. For example, Tektronix have this month released the TBS1000B model. For less than $600 you can get a 2-chanel, 50MHz, 1GS/s scope, with dozens of automated measurements, datalogging and limit testing. Typically these features would be in a higher performance scope costing a few thousand dollars. My gut feeling is that this trend will continue in the entry level, with more and more features being added for free as technology allows!



                  • Re: Oscilloscopes: Past, Present and Future


                    Thanks for the reply.

                    Your TBS1000B (Model TBS1052B-EDU) is available in element14 for $581 (+ GST) which is an impressive price.

                    It is limited by the 50MHz bandwidth but is certainly getting comparable with Logic Analysers.


                    You have a Road test challenge out for its big brother/sister .... maybe element14 should have doubled the number available and used this to get the lower end of the market into the game.


                    Thanks again


                • Re: Oscilloscopes: Past, Present and Future
                  sheldon bailey

                  I remember my first scope, a hand me down from a family member, single channel less than 1Mhz and almost no trigger. I use an MDO4000 almost daily now.


                  the only thing that has stayed the same is the wait, either for tubes to warm up or software/firmware to load.

                    • Re: Oscilloscopes: Past, Present and Future

                      Reminiscing about early scopes.... in my first proper job I used an analogue storage scope, I think it was Tektronix, with an interesting screen coating technology that created a persistent glow. The image lasted long enough to read off the measurements so long as you were quick! Anyone else remember those? How did that technology evolve?

                        • Re: Oscilloscopes: Past, Present and Future

                          Hi Joy,

                          Thanks for the reply. The phosphorus coating on the screen gives persistent glow. I used analog scopes in University and spent a lot of time playing with the intensity to hold images on the screen! I remember when I started at Tek and been told stories of how engineers would use this persistent screen, a cloak over their heads and a Polaroid camera to capture a "screen shot"!! When I hit the road as a field engineer and repeated this story I was surprised how often I'd come across an engineer who'd used this technique.


                          One of the great advantages of a traditional analog scopes is that the only deadtime between acquisitions is the time for the horizontal timebase to reset and the trigger to rearm before it can capture another waveform. This gives analog scopes a very fast "waveform update rate", and make it a very useful tool to finding intermittent glitches or events. The phosphorous screen then allowed us to hold the glitch on the screen for some time to analyse it. Digital scopes on the other hand are a different beast altogether - it's got a pre-amplifier, then an ADC, then some DSP done typically in a ASIC. The sample points are then stored in memory, transferred to a PC and then processed before being displayed as pixels. During the transferring/processing time the scope is not acquiring any new waveform, and depending on the length of the waveform this processing can take 100s of milliseconds or even seconds. So there can be huge amount of dead time between acquisitions on a digital scope. On the other hand digital scopes have higher bandwidths, complex triggers and the ability to capture and post process the signal.


                          To put some figures on it, an old analog scope could capture around 500,000 waveforms in a second. On the other hand, a modern digital scope might have a maximum capture rate of 1000 waveforms (it all depends on the settings for both). Since digital scopes have become dominant in the market Tektronix has tried to bridge this gap between the analog and digital scopes. They've invented a fast acquisition mode called DPX which uses dedicated ASICs to process the waveform and perform the rasterizing the samples into pixels. With DPX we can achieve capture rates up to 300,000 on some models. Check out this video of our DPO5000 scope showing DPX in action to capture some illusive activity on a bursting signal. While in the past analog scopes would allow you to find the illusive problem but not capture it, while the digital scope would allow you to trigger on the event if you knew it was there. DPX based scopes allow use to both find the problem and see it or analyse it, making them greater debug tools. At Tek we also leverage DPX speed up automated compliance testing for serial data standards such as Ethernet, HDMI and DisplayPort to name just a few by acquiring statistical information much quicker.


                          We refer to scopes with DPX as Digital Phosphors Oscilloscopes or DPO scopes. This is why about 7-8years ago Tektronix switched from calling their scopes TDS (Tektronix Digital Scope) to DPOs in order to distinguish it from other instruments on the market. One of these DPO scopes released recently is our MDO3000 model (6-in-1 scope). It is the first bench scope to include DPX running at rates around 280k waveforms per second.

                          On another note, I gotta take this chance to mention how we've migrated DPX technology into the frequency domain with the invention of the Real Time Spectrum Analysers from Tek. This is genuinely cool! DPX in a spectrum analyser allows us to see the time varying nature of modern digital RF signals. It allows you to see RF signals at the same frequency but separated either by time or power. Check out this video.

                          This has opened a wide variety of applications from debug, to EMI diagnostics to surveillance




                        • Re: Oscilloscopes: Past, Present and Future

                          Can't agree more with you Sheldon! The "instant on" scope is like the holy grail!