Chinese French German English
Welcome to E2E Engineer Discussion Forum Sign in | Join | Help

Welcome to the E2E Community Discussion Forum!

  • You are among the first to view this thought-provoking and educational networking tool for design engineers looking to discuss integrating electrical components into their products.
  • The forum is open to all who seek, or wish to share useful and meaningful information.
  • We encourage constructive dialog and independent thinking free from promotional innuendo.
  • For more information on the rules and guidelines, read our discussion forum guidelines

Sensor error

Last post 04-29-2005, 10:36 AM by Bobg. 3 replies.
Sort Posts: Previous Next
  •  04-10-2005, 10:34 PM 73

    Sensor error

    I have a question about a RTD sensor.  I am thinking of using a 3 wire RTD with 300 meters of lead wire to get back to my PLC.  Hope to have as low of error as possible (1 deg C).  Will this RTD selection work?

    second question is why do we use temperature transmitter instead of direct wiring to PLC? Is it not relate accuracy?
    "A scientist can discover a new star, but he cannot make one. He would have to ask an engineer to do that." — Gordon L. Glegg
  •  04-12-2005, 3:09 PM 86 in reply to 73

    Re: Sensor error

    Question 1 - Yes this can be done.  The only error you will get from the leadwire itself is if the resistance of the leads are mis-matched.  This can be avoided by using all one color wire that comes from the same spool.  Also, using a 1000 ohm part versus a 100 ohm part helps to keep the overall leadwire resistance a lower percentage of total resistance.  Finally, your part will be much less expensive if you allow the additional wire to be spliced on.  Manufacturing with 300 meters of wire directly in the sensor would be very difficult.
  •  04-12-2005, 3:39 PM 92 in reply to 73

    Re: Sensor error

    You may find it difficult to get agreement to eliminate color coding and use all three wires from the same spool. Assuming you use a more conventional color coding of two colors (Red/White/White or Blue/Yellow/Yellow, etc.), these will need to come from different spools. Make sure the wires are the same gauge and stranding. The mismatch error can be expected to be a maximum of 5% of the resistance of the run, so the wire gauge used should be larger (lower resistance per unit distance) for minimal error when direct connecting long runs. You can compare this error to the amount of change in resistance per degree your sensor provides to determine whether this error is significant.

    There are other potential sources of error with long runs that may need to have twisted and shielded cables to eliminate them. Signal conditioning such as a 4-20 mA transmitter can eliminate the need for anything more than a twisted pair of wires, which may substantially lower wiring cost vs.a long run of large gauge shielded cables. You also receive the benefits of the signal conditioning to linearize the sensor output and can even calibrate out sensor errors to make all outputs match.

    You might also find that you do not need a 3-wire connection to your transmitter if you can mount the transmitter close to the sensor. This can allow the sensor to be built smaller or more ruggedly with only two internal wire connections.


    Vitreous Humor
    "Eye can see clearly now..."
  •  04-29-2005, 10:36 AM 187 in reply to 73

    Re: Sensor error

    With a wire that long, you run the risk of error due to signal loss and EMI; a 300 m cable makes a great antenna. Consider using a 4-20ma transmitter with the element.
View as RSS news feed in XML
Powered by Community Server, by Telligent Systems