Quote:
|
Originally Posted by stephenson
Glad got to the real story ... 95 C sounds high relative to what I have seen in the thousands of threads at this site. Most common numbers are mid 80s at normal highway speeds without fans running.
Anyone ever thought how weird it is to have computers running fuel, ignition, etc at multiple cycles per second, with a slow moving analog device like a thermostat controlling the water temp? Guess it means water temp, within a relative broad range, is satisfactory to the far more capable digital control systems ... does anyone have opinion (:-)) on what result would be if the water temp were regulated tightly? (I suspect imperceptible changes in efficiency - unless, the temp could be elevated and safely maintained much higher)
Thanks!
|
Well, the analog (seems to be a dirty word these days) system has built-in hysteresis (slow to react to change), which is actually a good thing in many cases, including this one. Analog closed-loop systems had been around for years and work well. The thermostat has a relatively predicatable lifetime and (generally) fails in a safe mode - open.
Digital closed-loop systems are actually "analogues" of analog systems!
I suspect that the variations that we see in thermostats is not in the closed and open temperatures, but instead in the "gray" area between completely closed and completely open - and these are caused by manufacturing variations in the thermostat. Maybe the real advantage of digital systems is the elimination of manuracturing variations in systems (of course, a digital temperature sensor has an analog signal and an analog to digital converter, both of which have their own pecularities...)
I like to point out to people that old technology is not necessarily bad. For example, the old-fashioned ignition systems failed in a predictable time frame, mostly from the wear of the points, cap, and rotor. With electronic systems, the switching transistor could last 500 years or it might poop out the next time it is called into action to produce a spark pulse - you don't know, and there is no nondestructive way other than statisically to predict it's failure. People laughed at the FAA for using computers with tubes - same thing. Tubes, once past infant mortality, fail in a predicatable time frame, and can be replaced proactively. Transistors might work for 1000 years, or they might fail at the next switching transition.