Quote:
Originally Posted by AC_Hacker
So far on my system, I am doing everything wrong... I have two single-sensor branches (short ones) coming off the main line, and no termination resistors and no power line caps anywhere. I'm getting good, reliable data. My total length is about 30 feet, excluding the short branches. I figure I have another 25 feet, and three sensors to go.
Aside from violating all the rules, I have been careful in the construction and deployment of my sensors... I check for good data, at each step. So far, I have not entered into the ZONE OF UNRELIABILITY. If things begin to get sporadic, I'll start with a termination resistor... If that doesn't do it, I'll go with the tiny caps... if that doesn't do it, I'll replace the branching topography with a linear one.
-AC
|
Cool, that's great info to have! I'm getting some wire this week to rewire the controller and put the sensors and controller into a semi-final position. I'll post some more pictures of the setup when it's done.
BTW, I've also, as Vlad suggested, posted the controller piece into a separate thread:
Custom hydronic heating controller
As for the branching setup, electrically it should make no difference to a linear setup - the only difference is that you're attaching a bit of wire to the pins of each sensor - if you think of the sensor's legs themselves as short runs of wire, there's no practical difference between putting all the sensors onto one long wire run, or giving each sensor its own - you're only moving the place where all the signals start running in the same wire, but since two sensors aren't sampled at the exact same time, that shouldn't matter.
In a linear setup, you'll have
more wire for each sensor, and hence more resistance and capacitance. Since it's DC, capacitance of the wires is irrelevant, and the resistance should only lead to a negligible voltage drop - at 50 feet with halfway decent wire, I would expect 1% at the very most (or 0.03 of 3.3V, which should be well within the tolerance of the sensors).