I am looking for information and ideas about permanantly mounted sensor wiring. Currently i am using switchboxes which work well. I have heard about hooking all the sensors with only one (two wire) cable. Using a "Bus"? My background is in mechanics and i don't understand "bus" or how all the sensors could share one cable. Does anyone use this kind of wiring? Is it possible? Any comments are welcome.

"not a vendor"
2 web sites to check out........

http://www.imi-sensors.com/

http://www.ctconline.com/

There is lot's of information on both these sites and yes you can use "bus" boxes or intermediate junction boxes to cut down on the wiring. Rockwell/Entek has the Enwatch Box that can handle 16 channels which you then send to a processor with ethernet cable.

http://www.entek.com/

Hope this helps.

Yes, I'm a vendor.

We do something similar; accelerometers are cabled to a signal multiplexer, and then made available on a signal port. There's local channel control, or there's a couple of options to string many of these controllers together with various control and channel ID options.

The big savings clearly comes in the ability to use Cat5 unshielded twisted pair cable to do long cable runs between boxes - about 6 cents per foot rather than northwards of a dollar. An example of where this can be useful is on a machine train where there's possibly hundreds of sensors over a structure that's 300 feet long by 30 feet high by 30 feet wide, and you want to bring all the signals out to a single instrument connection point. At least with our system (I can't speak to the pricing practices of others), this becomes cost-effective at an astonishingly small amount of cabled length reduction. With 150 sensors clustered in about 10 groups, the savings would be on the order of 100 feet of cable per sensor, or 15,000 cabled feet, for a savings of $15,000 on cable, replaced with $8000 of multiplexers. Add in the lower installation cost, smaller conduits, smaller cabinets and the labor savings on monitoring (more measurements taken at one space = more productive data collection) and the savings remains substantial.

In our system, centralized switching can easily be upgraded to continuous monitoring. . . depending on your specific requirements for scanning frequency, truly diagnostic monitoring (as opposed to mere velocity-based alarming) could be as low as $2950 for the entire machine, while still retaining the ability to collect spectra with the data collector of your choice.

The important piece to remember is, at this stage of the game, the "bus" systems are all multiplexing sensor data onto the wire by switching them on one at a time. Not all signals are available on the wire simultaneously.

Regards,

Eric Thompson
Spintelligent Labs, Inc.
www.SpintelligentLabs.com