D Feucht

's profile
image
-----

Tektronix, Inc., 1968-1985, Design Engineer, test and measurement instruments Tektronix Laboratories, 1978-1985, Research Engineer, advanced instrument concepts, robotics Synektron Corp., 1985-1988, Senior Electronics Engineer, motor drive design Innovatia Laboratories, 1980 - , electronics research and development


D Feucht

's contributions
  • 05.11.2015
  • High-performance motor control at low speeds
  • The methods presented in this article for determining the motor phase at stall are among the better ones, though the best I have found (and used) is that of the U.S. Seagate patent of Cassat, which is based on saturation of the magnetic path of the stator phases. Whichever phase is most aligned with the rotor magnet field and driven with the same polarity will saturate most, the inductance will be least, and some simple analog circuitry can detect this. This determines phase to within 60 degrees, which is enough to drive whichever of the six phase-steps that torques the rotor most in the desired direction.
  • 03.11.2015
  • Quest for the Ideal Transistor?
  • Kevin, I think your point is basically correct. The low resistance at the inverting input results in a low time constant. In my (free! see www.innovatia.com) book, Transistor Amplifiers, I show that the same voltage gain expression for a VFA can be rewritten to be interpreted as a CFA. Whether feedback is called "voltage" or "current" is misleading. Those names have more to do with how the circuit is analyzed rather than what it is. Barrie Gilbert calls CFAs transimpedance amplifiers, for that is essentially what they are from a port to port standpoint.
  • 03.27.2015
  • Ode to the Oscilloscope
  • Roger, Thanks for recounting some old history! About the time you were leaving the USAF, I was 17 years old and joining Tek. They put three of us kids right out of high-school into the calibration class along with the Navy techs they hired to calibrate scopes. In final test and cal we used a stack of Tek-made equipment to cal scopes. One of them was a time-mark generator that did, as a separate piece of equipment, what your customized scope had built into it. If those scopes you customized had instead been Tek scopes, you wouldn't have needed the built-in time markers! However, maybe they were. Back then, the sweep generators that generated the sawtooth sweep waveforms in Tek scopes were op-amps called Miller sweep generators and the op-amps usually consisted of a pentode (for high plate resistance) followed by a triode cathode-follower. The pentode had all the voltage gain, and compared to today's op-amps, wasn't much. So the sweep linearity was maybe 1 %. (It seems so quaint to be writing about electron-tube circuits in 2015!) And by the way, the other two kids were Paul Magerl who went to MIT and Bill Den Beste who went to Reed C. (where Tek founder Howard Vollum went to school) - and we could perform as well in calibration as the older Navy techs. Our problem was that we also thought about improving the cal procedures and the scope circuits!
  • 03.27.2015
  • Ode to the Oscilloscope
  • Oliver, Actually, the "dual-delayed" part is simply because of the two-scopes-in-one aspect of this monster 'scope. As I recall, Tek introduced the true dual-delayed feature in 7000-series time-bases. Nowadays, DSOs have a similar function, though implemented quite differently. The traces on scope CRTs back then were "sharp" (thin lines on the screen) simply because the tubes were so long, as required by the physics for a sharp beam. As 'scopes shrunk (to fit under airplane seats - that was our size constraint in the Portable scopes group at Tek), the traces lost that crisp (and bright) look to them - another example of how older technology has some advantages. Probes are important enough that Tek had a probe design group. Those big, old grey probes are clunky on a bench but they had retractable probe tips that are lacking in too many DMM accessories of today. A voltage probe is essentially the upper half of a compensated voltage divider, but how to design it geometrically is not as conceptually simple.
  • 03.27.2015
  • Ode to the Oscilloscope
  • Sgt Jack, Living in 1968 is not all that bad in that some aspects of technology then are better than now: 1. The electronics was more robust, electrically and mechanically. Nowadays, the shrinkage of monolithic integration and lower voltages has increased electrical fragility of circuits. Surface-mount connectors do not have the mechanical strength of through-hole or panel-mount connectors. 2. The circuit technology was simpler and could be repaired and maintained by technicians who could understand the entire instrument. 3. The documentation on 'scope products (from both Tek and H-P) was substantial and open-source, so that it was possible to know what you bought. Instead of paying for it to use it, as is often the case nowadays, you paid for it to own it instead. 4. The effort and competence brought to product design (of 'scopes in this case) was generally greater than what is expended nowadays in producing short lifetime, "spin cycle" products. There are happily exceptions, but back then, reliability and durability of products in design were higher priorities. This does not mean that the technology of today does not allow for greater reliability and longer product life; it simply does not seem to be as high in priority as back then, when "quality" was the buzzword.
  • 03.25.2015
  • Slope Compensation in PCMC DC-DC Converters
  • Dr. Franco, I wanted to add that your article is clear and well-illustrated in showing how slope compensation affects stability for changes in input and output port voltages. It should be quite helpful to those new to peak current control. I know you as an op-amp and analog circuit expert and it appears that you (like myself some time back) have migrated into power converter circuits. I would be interested in any feedback you might have on my "refined model" of peak current control, given as a 6-part article on Planet Analog beginning at http://www.planetanalog.com/author.asp?section_id=3049&doc_id=563764 Thanks.
  • 03.25.2015
  • Slope Compensation in PCMC DC-DC Converters
  • Back in the early '90s, Ray Ridley's sampled-loop ("continuous-time") model of peak-current control was a major advancement over the low-frequency average model because it accurately predicted the subharmonic oscillation behavior. So start with Ridley's website at www.ridleyengineering.com/books.html If you then want to move on to a more refined model of peak-current loop control based on average cycle current rather than the current valley point each cycle, continue from there to the 6-part articles on it starting with Part 1 at http://www.planetanalog.com/author.asp?section_id=3049&doc_id=563764
  • 02.05.2015
  • The inductive nature of voltage-control loops
  • This article is a good lead-in to the more general theory of high-frequency impedance gyration. See http://www.edn.com/design/analog/4401129/What-circuits-textbooks-haven-t-told-you-about-feedback-amplifier-and-transistor-port-impedances
  • 11.26.2014
  • Peak oil in retrospect
  • Andy & aussie, For energy storage, which of course is critical in the use of intermittent energy sources, the best alternative I know for storage on the scale of residential and small business use, is nickel-iron batteries. They are safe, have essentially none of the problems of maintenance and life cycles that lead-acid batteries have, and can last for a century - at least the one in the Ford Museum in the electric car has lasted that long. I also agree that peak oil is still an issue. The two largest oil fields in the world, the Ghawar field in Saudi Arabia and the Cantarel field of Mexico, have been on the downside of their production curves (the Hubbert curves) for a long time. Fracking is a desperate attempt (and possibly not without negative side-effects in destabilizing terra firma) to extract what remaining oil can be obtained at a breakeven cost of $80/barrel in the Tuscaloosa reserve and $60/barrel in the larger ones. If oil is driven below $60/barrel by the geopolitical maneuvering intended to hurt Russia economically, it will cause "fracked oil" to be sold at a loss, thus endangering US oil supply, which has become more dependent on domestic sources. Political decision-making is also subject to the law of Unintended Consequences.
  • 08.19.2014
  • Transistor and PWM-Switch Analogs
  • Brad, The two-port model is implicit just under the surface in the article. Like the transistor, whichever terminal of the three is in common is the shared INP- , OUT- terminals of the input and output ports, while the remaining two terminals are the INP+ and OUT+ port terminals.