Appropriate Connector and Wire for ADC Input
KEggemeyer
Posts: 3
I am using a MAX1271 ADC to read the output of several LM34 temperature sensors (0.9 V to 2.0 V range).· The sensors need to be removable from the project.· I am thinking about using RJ-45 connectors (Ethernet-style) and Cat 5e wire (because I already have plenty of both).· My questions are these... Are these connectors appropriate for analog signals?· That is, is there a difference between connectors for analog signals and digital signals?· What constitutes a "noisy" environment where shielding would be necessary?·
I wouldn't even know where to go to find out -- well, except for here!· Thanks!
Kevin
St. Louis, MO
I wouldn't even know where to go to find out -- well, except for here!· Thanks!
Kevin
St. Louis, MO
Comments
Cat-5 and Cat-5e rated UTP cable is so cheap these days that I wouldn't hesistate to use it but RJ-45 plugs and jacks are pretty costly. Unless you have a monster supply of these (bulk priced?) I would buy a small supply of Cat-3 rated RJ-45 plugs and jacks to use on low frequency or DC applications like the one you've described.
In your application the cable just has to carry a slowly-changing DC signal (the 0.9V to 2.0V output of the LM34) so you probably can get away with a shunt capacitor (say 0.01uF from signal to ground) at the point where the cable connects to your circuit. The capacitor would shunt any high frequency DC noise that gets induced into the cable. This would be effective for very long runs...100ft or more. I wouldn't expect much current to flow to/from the LM34 so you shouldn't see much voltage drop on even a very long run.
If you find (or suspect) that the noise on the line is significant, you might want to consider using a low-pass filter (inline inductor, shunt capacitor) instead of just a shunt capacitor. You'd need this if your cable is run very close to house wiring (110V AC) and/or locations where there are EMI sources like flourescent lamps and microwave ovens. An oscilloscope comes in handy here because you could connect your oscilloscope the the prototype of your circuit and with the vertical gain turned up you could see the quantity and amplitude of the noise the cable picks up.
I don't have the formulas to calculate suitable values at hand but if you designed an LC filter for a relatively low pass frequency, say 10-20Hz, that would nicely pass your LM34's output (which changes very slowly with temperature) and filter out any frequencies above that. I'd Google "low pass filter" or "low pass first order filter" for a schematic and the relevant formula to calculate the pass frequency.
Hope this helps.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I wouldn't connect that if I were you...
Vive Le Tour!
Le Grand Depart July 6-8, 2007
Post Edited (aalegado) : 1/12/2007 2:25:35 AM GMT