I have been noticing some very significant indications in the industry where Analog functions are being tied very closely into the Digital realm. It’s not as if Analog and Digital are strangers to each other. From my circuit design experience in the 70s and 80s (And long before that as well) there was always the design-flow procedure of Analog circuitry being tied into the Digital via some sort of isolated interface. It’s just that recently, I see the lines between Analog and Digital beginning to blur more than ever.
I will give a couple of examples of recent developments that I have noticed.
Big Analog DataTM
The first is “Big Analog DataTM Solutions ” by National Instruments (NI) and Hewlett Packard Enterprise (HPE) where they are addressing pre-tested end-to-end solutions in this area.
HPE’s Moonshot Server System combined with NI’s Data Finder Server Edition will provide designers with an end-to-end system that is pre-tested and validated, that can manage structured and unstructured data coming in from any data acquisition analysis node.
Big Data is typically defined as four Vs: The system gathers large Volumes of data; the data which is gathered and analyzed by the system is in a Variety of formats and structures; the data is gathered at high Velocity and sample rates: and great Value is obtained from the analysis of that data.
Big Data comes from Industry/IT Data Sources in the form of events, logs and inventories; Social Sources from various Social media channels that are gathered by Facebook, Google, etc. containing people’s product preferences and companies they like; and finally Engineering/Scientific Data Sources resulting from the physical world of Analog phenomena via and Analog to Digital Converter, especially in the Internet-of-Things (IoT) arena.
The sources of this so-called Big Analog DataTM are RF, light, sound, temperature, smell, pressure and more—essentially mostly the things we, as humans, perceive in the natural world around us (Except for maybe RF)—need to be converted from their analog form into a digital format that the processing entity understands. That’s not new, but the end-to-end solutions that NI is in the process of creating, are not so straight forward anymore due to the immense amounts of data that have to be digested starting from the sensor to the decisions that will be made by the Processor based upon this huge amount of data. Figure 1 shows the three-tier architecture composed of Sensors, hardware and software working seamlessly together.
The distinctive three-tier solution architecture described by NI come together to create a single, integrated solution from the sensor to the decision-making entity.
The PSoC Analog Coprocessor
Recently, Aubrey Kagan, a very technical, creative and prolific blogger on Planet Analog, wrote a blog series entitled Measuring an RMS value on a PSoC5, Part 1: Signal Acquisition, Measuring an RMS value on a PSoC5, Part 2: Squaring a Reading, and Measuring an RMS value on a PSoC5, Part 3: Square Root and Result in which he was unfamiliar with the use of an FPGA or DSP, so he used a PSoC5 device for an RMS measurement. Being an Analog geek myself, I am not really familiar with using a PSoC, but Mr. Kagan opened my eyes to a useful tool for Analog geeks.
Used to be that an Analog geek like me would use an Analog solution like the Texas Instruments LM118 True RMS Detector.
Then, recently, I have also become aware of another similar tool by Cypress’ Development Community which has developed a PSoc Analog Coprocessor Pioneer Kit. This product will simplify the design of embedded IoT applications in home appliances, consumer and industrial systems and enabling systems engineers to more easily create custom interfaces for a variety of sensors.
These solutions are almost becoming the long sought-after Software-Configurable Analog IC or Analog FPGA-type of solution.
Is Analog disappearing? Absolutely not. The world around us will always be Analog-based and only Analog sensors, with conditioning, and fed into an Analog-to-Digital Converter can turn those signals into a format that will be understandable to a Processor (At least as far as I can see into the future).
Please share your thoughts on this topic. Am I off base?