Resistance Measurement


Thread Starter

Randy Demoe

I am looking for a way to measure resistance and have the data go in to a Allen-Bradley PLC5. The present way of using voltage and current and then calculating the resistance is not working. Is there a device that will connect to the PLC that will measure resistance directly. The range is 0 to 10 ohms.

Kim L. Ground

REPLY: You do not give many details of the application, but I will share our experience in obtaining readings of low resistances with SLC500 plc.
1) any method which involves reading the resistance directly is doomed to failure with resistances under 10 ohms or so because of the proportionately large contribution fromlead resistance, contact quality, etc. This is why commercial milliohm meters which are any good use a four wire system to eliminate lead and contact effects. These cost thousands of dollars.
2) we read resistances on the order of a few ohms by applying a constant, known current across the unknown resistance and reading the voltage drop with a standard PLC voltage input card. Note that one end of the voltage card is commoned in AB cards (even the ones with 'differential' input capability) so one end of the unknown must be connected to signal ground. If not, you will get erroneous readings unless you either use one of the 'differential input' cards (like 1746-NI8) and use only one input channel (since the other channels are not isolated from each other) or you will need a signal isolator for each channel (we use Moore Instruments units).

As you say that you are having trouble when using a calculation based on current and voltage, my comment to that is that perhaps you are trying to use AC signals ? There are a number of problems with using an AC signal: 1) PLC doesn't scan all of his inputs synchronously so you may read the voltage or current at a point which is not the peak of the waveform. 2) if the unknown has any reactive nature, then current and voltage waveforms are by definition not in phase and so even if the plc sampled both simultaneously you would have the effect of 'power factor' causing errors. 3) plc analog inputs are generally not reliable in reading AC wavefroms at all, because of unpredictable instantaneous response of the input device and unpredictable timing of the sampling. This will become a much more difficult problem if you must use AC excitation.

I would be glad to address this issue at more length if you will provide more details about what you are trying to do by direct e-mail to me at [email protected].

Bob McWilliams

One way to do it would be to use an RTD input module, configure it for copper RTD's (0-10 Ohm) and use the 3-wire connection to subtract the resistance from the measurement loop. I have found the RTD modules to be surprisingly accurate with resolution of 0.01 Ohms. How much accuracy do you need?

Mark Hellyer (J & E Hall Ltd.)

Another method is to use an intrinsically safe barrier device (as used in hazardous areas). Pepperl& Fuchs have resistance to 0/4 to 20mA current output device (din rail mount, 24Vdc supply)that will connect to a current input of any PLC. Might be expensive compared to other options already suggested, the barrier is about £100!