"Software Defined" concept, as applied to "Instruments"
What is the purpose of the "Software Defined concept", and how does it differ when applied to "Instruments" instead of "Radios", for which designs are already being developed in several laboratories? (See the Web-site: "Software Defined Radio", SDR, an industry organization for companies interested in the subject.
http://www.sdrforum.org )The primary reason both applications are in need of the new "software-defined-architecture" is the fragmentation of their markets. The mobile communication market is fragmented today, because different groups of nations have not and are not likely to agree on unified standards. Only a Software Defined transceiver circuit can be programmed to work on a global basis. The instrument market is fragmented, because instruments are specialized in hardware to serve thousands of slightly divergent test applications. The reason for this is "tradition" and the stagnated business model of the T&M industry.
Market fragmentation has a large impact on the production volume of a particular circuit or system, which is the deciding criterion for doing, or not doing, a multi million dollar System on Chip, SoC, development.
On the technical side there are also more similarities than differences between the two applications. However, for a "Software Defined DSO" design, which I have nearly finished, the "equivalent time" nature of the Gigahertz scope allowed me go a step further than what's being done for "Software Defined Radios". For these existing developments, the conventional "functional-modules" of a transceiver are either just software-controlled or completely replaced by software running on a fast DSP, while for my DSO design several functional-modules themselves are a mixture of hardware and software. (See Equivalency of Software/Hardware)
This method of extending the software-defined concept down to the "circuit level" can accomplish two things. First, it can drastically reduce the number of hardware components in all mixed-signal designs. This then means a much smaller chip size for SoC implementation. Second, it can provide automated adjustments for component variations (Temperature dependence, aging, manufacturing tolerances etc.) And here is a second purpose for "Software defined Systems" which, in the years to come, will mean a Revolution in Analog circuit design.
Apparently these are considered new concepts as an article in EETimes, Oct. 21, 02 "IMEC eyes paths to reliable SoC analog circuits", would indicate. (See Mixed Signal, this one page .jpg file, becomes readable when printed.) Described are the possible automatic adjustments for a Software Defined Radio" (a wireless LAN transceiver). For a "Software Defined DSO" these adjustments would concern measurement accuracy, such as for amplitude and time, but also linearity of time display and constant gain and phase inside the given bandwidth range. --- The DSO design described on this web-site utilizes these (new) concepts and also provides for auto-calibration and self-diagnostic. Look for the remarks at the end of Intel's report on their Billion-dollar project for creating "Software Defined Radio". (http://www.msmisp.com/futuretest/Intel-Free-SDR.htm
--- (Also a 600Kbyte Jpeg file, which needs to be printed for easy reading.)
Back to Index