FP-RTD-124 Measurement Error

[OKSparks]

I have a test cell with a range of 40 to 120°F with 3-wire PT100 RTDs (TCR=0.00385) for measurement.  During the initial calibration run it was noted that the error increased from nearly 0° at a calibration bath temperature of 120°F to about 5° with a 40°F bath.  This occurs on all 8 channels of the FP-RTD-124 and on multiple RTDs.  All 8 channels are configured for °F with the appropriate TCR.

The RTD's are wired to the FP module with #18 AWG shielded cable, routed in conduit away from power sources.  The cable length ranges from about 25' to 75', but the deviation from expected values does not correlate with the cable length.  In the same application, I have stand-alone PID controllers also using 3-wire PT100 RTDs.  These have slightly longer cable lengths of the same cable. Although these only display whole degrees, the display does match the bath temperature along the calibration range.

For troubleshooting, MAX is used to display the data values and the test is limited to using a glass thermometer and an ice bath.  I borrowed a "known good" RTD from another lab and obtained the following values:

                     MAX     Glass   PID

RTD              °F        °F        °F

Omega        37.11   32.4      32

Minco           36.47   32.2      32

Borrowed     35.75   32.2      32

Using the same bath, the borrowed RTD was connected directly to the FP-RTD-124.  MAX indicated a temperature of 32.6°F with a glass reference of 32.2°F.  This is well within the tolerance of the temporary ice bath.  This seems to indicate that the FP module is not grossly out of calibration.

The next thought is resistance of the cable.  If this is the case, why does the PID controller not suffer from the same problem, with no bias (offsets) applied or other compensation?

Any thoughts or suggestions?

Thanks

[Matt_S.]

One thought on this is that the PID controller might have internal compensation for line impedence. This is just a thought. It's obvious that impedence in the connecting cable is what's causing this issue. What's not clear is why the PID controller seems to have no problem with this. They may have hard coded some hardware on the input terminals to correct for this.

I'm looking into other possible reasons you're seeing this, but this sounds most logical.

Regards,
Matt S.

[OKSparks]

It is possible that the PID controller has undocumented compensation for lead length and the PID controller also accepts thermocouple and voltage inputs.  The PID controller uses constant excitation rather than pulsed.

I have 3rd party software to generate a calibration curve based upon deviation from a reference standard along the operating range.  It should be noted that the deviation from reference is NOT linear along a 5 point test from 40° to 120°F.  A 2nd order or greater fit is required to match the curve to the FP output.  I wouldn't expect lead resistance to have this effect. Could there be an issue with the cable capacitance and the pulsed excitation current? 

Unfortunately, my FP modules are temporarily inaccessible, so I cannot conduct tests until next week.  The first test will be the same test arrangement with discrete 18 AWG conductors of comparable length.  I will also connect a "PID" cable in place of the "FP" cable - this should have no effect, but could rule out the possibility of a spool of defective cable.

Thanks for your help.

[Matt_S.]

Hi again,

Definitely keep us posted on what you find. I'm very interested to find out what is causing this behavior.

As soon as you get more results let us know and we can go from there.

Hope you have a great weekend!

-Matt S.

[OKSparks]

We sent the FP-RTD-124 for an "extensive" calibration.  When it was installed, the PT100 values were about 5°F lower than the reference at 32°F, per raw values from MAX.  This was fairly consistent across our calibration range of 32 to 140°F. 

We have calibrated the MAX output (OPC) in our 3rd party system and the results are temperatures within 0.1°F of the standard across our range.  This accuracy level is acceptable.

It looks like the main issue was a problem with the FP-RTD-124 calibration.

Thanks to all for your help!

[Matt_S.]

Hi OKSparks,

That's fantastic. Thanks for keeping us posted. I'm glad to hear that this problem has been resolved for you.

If anything else comes up, let us know. We're always more than happy to help!  

-Matt S.

[Nac]

Hi,

I've faced same problem bcoz cable lenght is above 40 mtrs.

Better to use a temperature transmitter that will convert RTD reading to 4-20 mA and read this using AI module

In this case u can strech cable length up to 100+ meters !!!!!!!!!!!

i ve personally experienced

Bye,

Nachiket

[Nac]

Hi,

I've faced same problem bcoz cable lenght is above 40 mtrs.

Better to use a temperature transmitter that will convert RTD reading to 4-20 mA and read this using AI module

In this case u can strech cable length up to 100+ meters !!!!!!!!!!!

i ve personally experienced

Bye,

Nachiket