On July 24, 2013, Teledyne LeCroy demonstrated a 100 GHz oscilloscope. I recently caught up with Peter J. Pupalaikis, vice president, technology development Teledyne LeCroy; Roger Delbue, vice president, engineering Teledyne LeCroy; and Dr. Amarpal (Paul) Khanna, vice president, components and subsystems development, Phase Matrix, a National Instruments Company, to find out more about how the scope works and what the future plans are.
In the very beginning, the project just involved Delbue and Pupalaikis, who have been working in high-end scopes for the past 15 years. LeCroy launched products into the high-end scope market about 13 years ago. (Before that it focused mostly on mid-range scopes.) In the early days, Pupalaikis says, “We were disappointed. As soon as we launched our super high-end product, Tektronix launched a higher bandwidth one just before us.”
As a result, the team started thinking about a new strategy. Pupalaikis points out that, at the time, LeCroy was about one tenth the size of Tektronix and one fiftieth the size of Agilent, so the team needed a clever approach. “That was when we figured out that DBI technology would be a way to compete by doubling/tripling the existing bandwidth of the scopes.” (DBI technology is the company’s digital bandwidth interleaving technology. For more on this see the Teledyne LeCroy whitepaper.)
Since then, the team that is now Teledyne LeCroy used DBI with its chip technology in its scope development. “First we reached parity with the competition, then were able to exceed the competition,” says Pupalaikis, “So this 100 GHz scope has been a target for the last fifteen years.”
Along the way, though, the scope development teams began exceeding what their microwave partners could provide. In 2007, the group teamed up with Phase Matrix. (The company’s previous microwave partner maxed out at 18 GHz.) “Paul’s group at Phase Matrix is the only one I can imagine that we could partner with that has the microwave capability to work with these super high bandwidths,” says Pupalaikis.
So how did this 100 GHz scope prototype come about? “What blows my mind is where we are," says Delbue. "We started 15 years ago at 1 GHz. Now we are at 100 GHz. That’s just one crazy number.” To get to that crazy number, the team first put together a mock-up idea and then partnered with Phase Matrix for the microwave designs.
“About six or seven years ago,” says Khanna, “they came to us and showed us the design for their 18-plus GHz scopes using DBI. We had the technology to support them, but their application was unique. Working with LeCroy encouraged us to reach new heights, which were necessary in order to continue working with them. It has been great fun, actually, working with such a highly accomplished team.”
I had to know. At any point in this process of designing a 100 GHz scope, did anyone think they were crazy? Of course, this led to laughter. But Pupalaikis admitted when he first cooked up the idea for DBI (at around 11 GHz scopes), many people didn’t think it would work. With a background in DSP and math, Pupalaikis thought that the real question was how to separate the waveforms into pieces, downconvert them, and reassemble them in a coherent manner with the fidelity that the customers need. (He says he wasn’t worried about the rest of the design, because Delbue and Khanna were on it.)
Figure 1: DBI for the 100 GHz scope
Delbue, on the other hand, says he took the DSP for granted because Pupalaikis was on it. His concerns were the metrology challenges, making sure they were measuring the signal properly, and correcting the instrument. Up to 65-70 GHz, Delbue observes, the necessary equipment exists in a reasonable form factor (generator, power meter, etc.). “Above 65-70 GHz,” he says, “everything kind of falls apart.” The equipment exists, he hastens to add, but it is not easy to put together. “Now it is not just one generator, but a generator with different pieces attached to it. The speed of the signals has reached a limit of what people know as test and measurement technologies. So, you need to develop your own tricks to overcome that. And, in a case like this design, we have to make all of those tricks work in one shot. “
Khanna concurs with Delbue, noting that on the RF side, the connectors, equipment, and waveguide all becomes different after 67 GHz. He also notes that packaging is a challenge because dimensions become very small. Perhaps the biggest challenge for his team was the diplexer. “At one point,” Khanna admits, “we were not sure we could make it to 100 GHz using the technologies available. But, now we have a diplexer that covers 110 GHz.” As the speeds and frequencies have increased in the LeCroy (and now Teledyne LeCroy) scopes over the years, Khanna says that one of the greatest challenges has been to handle a huge 2.5-36.5 GHz IF with the necessary linearity, spurious, and dynamic-range performance. “We’ve had to face these challenges at all of the frequencies we’ve worked on with LeCroy. And, as we moved up, all of these issues got more challenging.” To his knowledge, no one is making broadband contiguous diplexers at these frequencies.
When Pupalaikis and Delbue first started using DBI technology, they could buy all of the components to make the DBI deck. Now the diplexer, the mixer, and the high-power local oscillator are all custom-made for them by Phase Matrix. Put simply, the guts of the scope are essentially the company’s 65 GHz scope with an additional band that handles 65-100 GHz. The DC-100 GHz is divided into DC-36 GHz, 36-65 GHz, and 65-100 GHz RF decks.
So what’s the status with the prototype? Right now it is at Phase Matrix, where people are working to optimize all of the connections. The team is still working on the software portion. When the components, subsystems, and hardware platform are finalized, the team members will put it all together for the final deployment. They estimate they are within one year of official product launch.
Figure 2: The development team for the 100 GHz Teledyne LeCroy oscilloscope prototype. L to R: Roger Delbue, Peter J. Pupalaikis, Dr. Amarpal (Paul) Khanna, Michael Neumann, Irfan Ashiq, and Alexander Feldman.
Now that they’ve achieved 100 GHz, what’s the new goal? Well they are not retiring! Delbue points out that, first of all, they have to finish this one. But they do not see 100 GHz as the upper limit. The caveat, however, is this: “In the past, we benefited from Moore’s law. People need the scopes to make the next generation of components, and we need those components to build a faster scope,” observes Pupalaikis. “But that’s not really happening today. They demand faster scopes, but it doesn’t help make the scope components better.”
Well, I for one have faith that the members of this team have many more tricks up their sleeves!