Well, my post was kind of a joke, having been involved in analog and digital for many years. The percentage of time even engineers have no clue as to what is causing a problem is much greater with digital electronics than it is with analog. You never have to "power down, wait ten seconds, power back up, and see if that works" with analog, except to reset a circuitbreaker but you know it's going to work. It's not a big mystery as with digital. I just like to poke at digital because it is usually touted as being superior but it's not, really. Just different. Doug
I thought the "power down, wait ten seconds" had nothing to do with digital vs. analog but rather letting the system go down to zero voltage by allowing power retention components like capacitors drain off.
Unless the problem is with an analog power supply, powering down and back up on a digital system is no different than triggering the power on reset, if it can be manually triggered. However, some system reset controls do not replicate the power-on reset (a design failure, IMHO), and leave some digital circuits uninitialized, hoping that software will take care of it (every time!) Complex digital designs no longer reset every register during power-on-reset, but implement a digital sequence to initialize every register before use. That's a "fun" requirement to verify in a complex digital design... On the other hand, digital circuits (outside of their power supplies) don't suffer from decaying components due to age and cumulative effects of high temperature. An aluminum electrolytic capacitor is nothing but a failure waiting to happen, especially if the equipment is powered off between uses. And they are ubiquitous in consumer electronics. I spent my career developing high reliability electronics. Getting digital electronics to last is not nearly as hard as analog electronics. On the other hand, if continuously operated, a slowly failing analog system will give you plenty of hints (if you know how to listen) how it will fail before it fails functionally. The bottom line is that most all digital systems are inherently more functionally complex. A small box replaces what used to require an entire rack of equipment, and does more in the bargain. A rack of digital equipment replaces a room full of analog functionality. Design errors escaping product verification testing become increasingly likely the more complex the system is, particularly with respect to the overwhelming state space of large digital systems. I remember seeing system requirements for dual-filament lightbulb indicators in older systems (for additional functional reliability), and the unconvinced spec-writers for the replacement system trying to flow and adapt that requirement to LED indicators... (adapt=delete was the right answer) We often still do automatic "lamp tests", usually during the power up or initialization sequence. Some things are just too comforting to get rid of.
Well, as far as troubleshooting, the flummoxing of even the experts began with digital as they had no clue how actually solve the problem so it was just turn it off and back on and see if that fixes it. With analog, everybody pretty much knew whether a fix was going to work or not. Doug
Clearly, my experience and yours differ. Perhaps it is the era (I started in the mid-1980s as an electrical engineer), but we were very busy replacing analog systems with digital ones as fast as we could, to keep up with the customers' demands. They absolutely loved the greatly increased reliability and functionality of our digital products. We could not eliminate (re)calibration tasks, but we reduced them tremendously, reducing downtime during re-cal cycles and in many cases, extending calibration intervals. Back to model railroading, I'm sure analog power packs and block switches are very reliable (for the limited functionality they provide, compared to DCC systems' functionality.) There's just not a whole lot that can go wrong. But there's also not a whole lot they can do when they are working, compared to DCC. How many of the "problems" with DCC systems are due to user inexperience, rather than the equipment itself? Following the user group forums for NCE, Digitrax, JMRI, SPROG etc. on groups.io, there are a whole lot of self-inflicted problems (compared to uninduced equipment failures). DCC systems (command stations/boosters and decoders) are both more capable and more complex. New problems come with the territory, but most I have observed are self-induced and/or due to lack of user knowledge.
