Microwave professionals with only a passing knowledge of National Instruments may not be aware of this, but NI likes to do things big - big investments in their R&D ($200M/year); big visions of a whole new approach toward engineering innovation and productivity for their customers and a big user conference to share this vision with the world.
There aren’t many companies that can get over 3,300 of their customers and partners to fly to Texas in the middle of summer (at their own expense) in order to see that company’s latest innovations and product releases. Who can get over 70 editors from as far away as South Africa to make such a journey? The answer of course is National Instruments and the event is their annual NI Week in Austin. It’s a marketer’s dream. And this year, NI was sharing their big vision and recent moves into the RF/microwave space with all these conference attendees and editors.
NI made a big splash just before IMS with their dual acquisitions of Phase Matrix and AWR. At NI Week these two moves were put into the larger context of the company’s vision for the future of the test instrument market, presented in the first day’s keynote talks kicked off by president, CEO and co-founder Dr. Truchard. Truchard introduced the overarching direction of the company’s technology development, which also served as the theme of the conference - Graphical Design Systems. This vision represents a unified platform based on open, graphic software and modular hardware for RF test, semiconductor, real-time test, and embedded applications. The company’s products are focusing on a convergence of design and test.
Dr. Truchard Talks Vision
For the standing room only audience, Truchard talked about the transformative years in the mid-1960’s when transistors and early computers revolutionized measurements, recalling his personal history of having to make impedance measurements using an RLC bridge, manually null at each frequency. Commenting that innovation drives the test industry, Truchard mapped the evolution of this innovation through the vacuum tube based era of a market dominated by General Radio (Gen Rad) from the 1920’s, through the transistor era successfully implemented by Hewlett Packard in the mid-1960’s, building the case that we are into the third phase of innovation where software and modular instruments take center stage and will dominate the world of test instruments.
The history of test and measurement progressed from stand alone, dedicated instruments that owed their capability to the hardware inside the instruments through the automation of these instruments with computerized control. NI ushered in the next phase with custom or virtual instruments made possible through digitizing data from the real world and manipulating it with virtual instruments in software. This was the concept behind the company’s flagship software product – Labview introduced in 1986. An interesting analogy repeated several times over the course of the day compared NI's Graphic Design System approach to what has been playing out in the mobile phone market, namely the rise of software driven products (such as Apple and Google) over hardware centric solutions. The argument being that software can greatly enhance the capabilities of the individual device.
An Eye on Digitizing Trends
Successful technology companies have a vision of the future that is unique, accurate and broad enough to allow them to plan long term and outmaneuver their competitors. This vision drives the company’s core philosophy and the direction of their internal R&D, acquisitions and subsequent products. Intel lead by Andrew Moore is a shining example of one such company. Today’s Intel products have 2.5 billion transistors on a single die. Moore saw the rate at which computing power would expand and he invested in the manufacturing and engineering resources to achieve that projected capacity. NI saw that engineers who needed to understand the world through measurement could leverage this expanding computing capacity through digitization and software. In fact, with the increasing complexity of our machines and electronics, software-enabled measurements are essential.
Evolution of NI's Digitizing Technology: Sampling Rate vs. Bandwidth (accuracy)
Where the lifespan of many consumer electronics today are becoming shorter than the return on investment of the automated test equipment needed to validate these highly sophisticated ICs, economics are driving the case for a Graphic Design system approach. The company’s “manifest destiny” has been to do everything they can do in software and the rest is done in hardware. In the early 90’s, the company’s executives began looking at how they could maximize value to their customers through combining what they could achieve in software with trends that were occurring in digitizing technology, namely increases in the bits of resolution and the frequency of the sampling rates.
The company’s quest over the next two decades has been to close the gap between what could be done with plug-in board and software versus traditional dedicated instruments. The focus on the speed and dynamic range of the available a-to-d converter hardware, needed to compete with traditional instruments was a leading factor behind the acquisition of Phase Matrix. NI’s goal is to compete head-to-head with the established T&M equipment manufacturers. Truchard noted that Phase Matrix makes the front-end for the highest bandwidth oscilloscope in the marketplace.
The entire NI RF R&D team has been busy expanding their RF platform and portfolio of products according to R&D Director Jin Bains. In particular they have extended the frequency range of their signal analyzer. The NI PXIe-5665 high-performance RF vector signal analyzer (VSA) extends the 3.6 GHz version into the 14 GHz frequency range. The VSA consists of the new NI PXIe-5605 downconverter, the NI PXIe-5653 local oscillator synthesizer and the NI PXIe-5622, a 150 MS/s intermediate frequency (IF) digitizer.
This combination provides spectrum and wideband vector signal analysis over a frequency range of 20 Hz to 14 GHz with analysis bandwidths up to 50 MHz. It features a third-order intercept point at +24 dBm with an absolute amplitude accuracy of ±0.10 dB as well as an error vector magnitude of 0.33 percent for a 256 QAM modulated signal. It also delivers an exceptionally low phase noise of -129 dBc/Hz at a 10 kHz offset at 800 MHz and an average noise level of -165 dBm/Hz.
Claiming best in class performance, the NI team demoed the VSA by measuring the instruments intermodulation performance operating under two-tone drive, with a center frequency at 10 GHz and tone separation of 1 MHz. The PXIe-5665 VSA performance was compared to the same measurements using the PXA VSA from Agilent. The 10GHz signal used to drive the device was provided by technology (Quicksyn microwave synthesizer) from Phase Matrix and represents the new extended frequency range that NI will be able to address thanks to that acquisition. Both Agilent and NI measured instrument intermod products at -73 db. For the next test, the engineers generated a WCDMA signal and measured the ACP levels of the two VSAs, which was around -81 dBc for both instruments. Where the Agilent instrument took over 400 ms to perform a sweep, the NI set-up took only 30 ms, making the NI set-up 15x faster straight out of the box. The performance was credited to the speed of the host PC (8183 controller featuring an Intel I7 processor) which was tasked with doing all the signal processing operations such as the averaging, mod/demod, FFT, etc.
True to form, the NI engineers upstaged their own impressive results by switching the processing from the PC to an FPGA that was on-board an NI FlexRIO plugged into their PXI chassis. The sweep time dropped down to 2ms, giving the NI set-up a 200+ speed advantage.
Shifting Economics for T&M Equipment
According to the NI executives, the last decade has not been kind to the measurement equipment industry with overall revenues being down about 40%. NI believes this reflects the markets need for instruments that can provide the high performance (speed, accuracy, dynamic range) required by designers and manufacturers but at a fraction of the cost. NI is able to achieve these goals by leveraging Moore’s law of increasing processing speeds, and with multi-core CPUs and FPGAs, controlled by Labview, which is uniquely able to program both Multi-core CPUs and FPGAs in the same environment.
An integral part of the NI graphical system design approach, NI RIO technology combines NI LabVIEW system design software with commercial off-the-shelf hardware to simplify development and shorten time to market when designing advanced control, monitoring and test systems. NI RIO hardware, which includes CompactRIO, NI Single-Board RIO, R Series boards and PXI-based NI FlexRIO, features an architecture with powerful floating-point processors, reconfigurable FPGAs and modular I/O. All NI RIO hardware components are programmed with LabVIEW to give engineers the ability to rapidly create custom timing, signal processing and control for I/O without requiring expertise in low-level hardware description languages or board-level design.
Throughout this evolution, Labview has maintained software compatibility as part of the company’s long-term commitment to its customers. Meanwhile the benefits have also included an ever-shrinking footprint for test equipment and an expansion of areas where devices can be plugged into and controlled by Labview. One such area is embedded industrial control, an area that will be big in machine to machine (M2M) and smart grid operations. At this point, Truchard introduced the concept of the Graphic Design System in greater detail. This is a unified platform that will allow engineers from any discipline to plug into a single environment framework to control and monitor any system at any level of abstraction.
Abstraction vs. Obstruction
This was another common theme at NI, one of how device complexity has moved beyond the grasp of the individual engineer to understand in the full extent of the device being developed. For instance, consider the 2.5 billion transistor CPUs. It is a team project far beyond any one person’s knowledge. This complexity and how we represent the engineering problems stand as obstacles in the way of innovation. Truchard talked about tools that liberate the engineer from spending their time and effort learning the nuances of a given tool or underlying c-code, model or whatever abstract representation that is used. The Graphic Design system is an approach to incorporate deep-level knowledge into models that are easily accessed and used to build greater devices of even greater complexity with less emphasis on the underlying models.
Parphrasing Charles Shroeder, Director of Product Marketing - A car driver doesn’t need to understand the detailed workings of the modern ignition, hydraulics, electrical, transmission or braking systems in order to operate that car and get to their destination. In similar fashion, the engineer doesn’t need to know the details of all the elements in a given system in order to focus on the aspect of that system that they are developing. So the Graphic Design system supports flexible representation and removes the need for individuals to focus on developing models outside their area of interest.
This theme was popularized several years ago when the company was talking about hardware-in-the-loop and the ability to design around existing hardware by way of measurement.
On this opening day of NI Week, Vice President of Product Manager, Erik Starkloff played Keynote host and master of ceremonies for the introduction of the latest products and technologies from NI this year. Starkloff explained how unrelated industries such as transportation, energy and medicine all rely on communications, processing and I/O to access the computing power they need to become more intelligent. This is how NI leverages Moore’s law to address development needs across such diverse industries. Starkloff discussed how NI is extending the convergence of design and test in a number of these industries, especially in RF and semiconductor test.
This convergence is apparent with the acquisition of AWR for RF/microwave design. AWR’s flagship product Microwave Office, allows RF designers to represent devices at the individual transistor, diode, passive element and interconnect levels. Where NI relied on measurements of actual hardware to provide information to designers, the company is now focused on providing a feedback loop at the simulation stage before a prototype has even been built. Starkloff invited AWR VP of Marketing Sherry Hess and Josh Moore, solutions architect up to the stage in order to explain a little bit about the company, its products and the challenges that face microwave engineers everyday. Ms. Hess gave a brief overview of AWR and used a simple example of two transmission lines in close proximity to explain the nature of unwanted coupling and its impact on circuit behavior. Mr. Moore than demonstrated how a Labview virtual instrument was used to capture the real-time response of a 2.5 GHz, 250 watt power amp from Infineon and embed the measurement-based model into a system level simulation in AWR’s Visual System Simulator (VSS). The demo was based on features not yet available for the public, but should be part of the next product release.
When the feature is available, users will be able to easily import any existing Labview VI into MWO, assign port numbers to the automatically generated component schematic for insertion into a block diagram. This powerful feature will support measurement-based models such as the amplifier used in the demonstration, as well as complex models for any existing communications standard developed in Labview. It is clear to see how this collaboration between NI and AWR is going to significantly advance the capabilities of both products.
NI is clearly doing a great job of inspiring their own engineers and developing a loyal and enthusiastic customer-base in all technology and medical R&D markets. The latest emphasis on RF test and design will make for an interesting development in the tools available to our own industry, a development that the Journal will keep a close eye on.