I recall a data sheet for a Watkins Johnson limiter that had a family of graphs for AM-to-PM conversion (something you want to minimize when limiting a signal feeding a phase detector in a PLL). There were multiple traces on each chart for different frequencies and then there were several charts for different temperatures. The correct title for this would have been "AM-to-PM conversion vs. frequency vs. temperature" but the engineers got one past marketing by making it "AM-to-PM conversion vs. frequency vs. phase of the moon". Since both temperature and phase of the moon are in degrees, it was dimensionally correct.
The "Anticipating timer". I believe it was an EDN design idea. It had a 555 timer with an op amp configured as a negative resistor. This negative RC time constant could then be used to drive some relay contacts. Set it to 10 hours, and if you got home and found you had left your air conditioner running all day, you could use the -10 hour time constant to turn off the air conditioner ten hours before you pushed the button.
ELEKTOR once had a rain predictor consisting of a resistor, an LED and a battery (no switch). When the LED was lit, it was raining. They predicted an accuracy of greater than 50%. They then went on to analyze the rainfall patterns in various parts of England and yes, the LED would be correct more than 50% of the time. LOL Brian
I worked for many years in the R&D labs of a huge now-defunct Canadian telecom manufacrurer.
The old timers once told me of an April Fools prank they pulled on a PhD whom they thought was too arrogant for his own good. They somehow rewired the guy's lab bench for half voltage, changed all the pilot lamps in his test equipment to compensate, then changed the shunts in his voltmeter so that it would read what all the voltages were supposed to be. Needless to say his prototype circuit and none of his test equipment operated correctly though there were signs of life, and the meter said everything was normal. Even his soldering iron got warm but not enough to be useful. This went on for a week until someone took pity on the guy and confessed.
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...