sensors and long wires


I'm trying to use thermistors with my arduino. When the cables are short less than 1 meter the readings are very stable. However I have disconnected the phone lines in my house and plugged a thermistor at the end of 2 wires, and the arduino at the other end (obviously). I'm unsure about the length but one pair should be around 10m and the other one around 15m. And the reading are fluctuating a lot. it is related to the cable capacitance ?

is there a way to get the sensor to stabilize with such long wires?


Not capacitance but cable resistance. The wire's resistance will depend on what gauge the wire is. Then you can look up the resistance it has per foot. The longer the wire is the more resistance there will be. Here is a simple chart that covers some common wire sizes.

You may be getting some cross talk from other wires if they still are connected to a phone system. You may try placing a capacitor between the monitoring pin and ground to smooth out the varying input.

Two issues to manage here... divide and conquer...

(And at end: An alternative to consider)

When you make the change in your set up between short wires (bench tests) and long wires (though house's phone wires) two things will happen...

a) You will get an additional, relatively fixed, resistance in the circuit. This should be fairly easy to compensate for, in software. However, that extra resistance will vary from day to day if the temperature of the wires varies. And the tture in a room and the tture in the wall spaces, especially if it is an outside wall in Maine, won't necessarily be the same, so don't assume you'll be okay 'cause your thermostat keeps the room tture fairly steady.

b) You will get extra, and probably quite variable voltages in the wires, which they pick up by induction from the environment... in other words, your wires are acting as radio antennae.

=== Alternative:

If you take a little time to master the 1-wire chips... daunting at first, but not too terrible, really... then they will send the temperature they are seeing across the long wire as a digital signal. In essence, we're talking about the difference between the way LPs stored music for reproduction (or VHS video tapes) and (music) CDs (or DVDs).

You can "throw (a little) money at the problem", and get a little interface device to put between your 1-wire chips and your Ardy, and then things become really easy. Along the way, you pick up some general skills that will be useful in other projects....

The 1-Wire chips will probably work across the phone line (if the phones are disconnected), too. ("Probably" because of issues of how long? How good? Passing by what? Etc. At least with a 1-Wire result, it is either right or wrong. (You can use a checksum if you want to be sure, or just read 3 times, and use the best 2 out of the three. A wrong answer will probably be impossibly wrong, anyway)

1-Wire chips don't like star topology's though... but then neither would your thermistors.

Oh.. and you can put multiple 1-Wire sensors on a single channel, i.e. single pair of wires leading off to the places you want to sense.