A solution looking for a problem
The first time I ever saw a photocopy machine was on an episode of Mission Impossible. In my defense I was only about 10 and not yet introduced to carbon paper, but I remember thinking “who would ever want to copy anything?” This is by way of pointing out that I am obviously not very good a detecting trends and seeing the future. In my travels through the electronic world, notwithstanding that I have designed with them, I have come across products and thought “that’s a solution looking for a problem”.
NKK makes a “Smartswitch” which is a single pole normally open switch with a screen embedded in the button head. I have worked with the monochrome versions which allow messages of 6 characters by 3 lines and I even developed a method of responding to setups by using single, double and triple clicks. (In truth I have dined out on my approach to interfacing with the device. There was the two-part article in Circuit Cellar “Driving the NKK Smartswitch” which became an application note with NKK for a while. There was also a blog “Implementing a Custom Serial Bus”.) However, the device is not cheap. I know it is used in aircraft simulators, customized (bulletproof) limos and according to what I see on NKK’s website, studio recording panels. NKK also has the same physical format with RGB capable of 65K colors and 64 x 48 pixels (see ISC15ANP4 for instance). Demos at shows have movies playing on the screen. Call me shortsighted (oooh, unintended pun), but I don’t know why I would want a TV show in the middle of a button - can anybody point me at a use?
Figure 1:Monochrome NKK Smartwitch showing the use of a 6 character x 3 line display. The button can be pushed. (Source: Author)
Echelon was into “IoT” long before it became a thing. They developed a “Neuron” chip (originally made by Motorola and Toshiba and replaced now by Cypress) that was programmed in C and had a built-in “LON” networking protocol to allow for interconnections. They targeted building automation quite heavily and seemed to have some success in Europe over the years. They are still in business. I was using their Neuron chip in 1993 and although it was simple enough to use, building a network was not so easy and I struggled and found customer support somewhat wanting. I never thought it would go anywhere. We sold the rather expensive development equipment and abandoned our efforts. Certainly Echelon had tremendous foresight on interconnection, but success in the IoT field still seems to elude them.
Hewlett Packard designed the HCPL-3700 a long time ago. I cannot find a date on my early data sheets, but I suspect it was 1980 or earlier. Through several name changes, Broadcom still manufactures the device. To my mind it was a marvelous design, incorporating a Zener diode bridge with hysteresis driving an optocoupler. This allows AC/DC input with threshold switching levels along with the aforementioned hysteresis. And yet I have only managed to use it once in a design! Obviously there is a demand somewhere since it is still in production, but where is it being used?
Figure 2: The schematic of the HCPL3700. There are variations with lower input current. (Source: Broadcom, Inc.)
Have you ever dismissed a product as pointless. And were you right?