RTD Calibration

[scootabug]

Hey all, I have an Pt100 RTD and I can't figure out how this needs to be configured. After trying a variety of set up options for the 1762-IR4 RTD I/O module for my MicroLogix, I have decided that I either have the wrong information or I am getting something wrong and just don't know it yet (but jeez I love doing things the hard way!). So here's the short-ish story: My first concern is that these RTDs have 3 wires. 1 red and 2 white. I assumed that the Red wire is the EXC, cos he is common among all RTDs but is the odd man out on this 3 wire RTD. Continuing on, in accordance with the AB "Field Wiring Connections" information, a 3 wire RTD is wired with the RTD's EXC (Red wire?) to 'EXC n' and the other two go to 'Sense n' and 'RTN n' (where n = whichever input you're using). So I did that, and I get a "short or open circuit" scenario, so I figured I got them wrong and continued to try all of the combinations in between without success. I ended up wiring it up like a two wire RTD with Red as EXC and one of the two white wires wired up to the EXC, because when measuring the resistance over the Red and either of the two white wires, I noticed that I was getting resistance not-dissimilar to what I should be getting given the approximate temperature. I am getting a reading doing it this way (perhaps a false reading, but it's better than a 0 indicating a short or open circuit!) Any thoughts in regards this phenomenon? Second issue is that the resistance of the RTD doesn't match up with their "conversion table". Regardless of anything else (PLC configuration etc), testing this with the multimeter alone, surely this information and how my RTD compares to my Fluke DMM's Thermocouple input should match, right? Which leads me to the thought that the RTD I have does NOT match up with their conversion table. For example: Yesterday it was 17 degrees (according to my Fluke multimeter with its thermocouple, positioned in close proximity to the RTD), and according to the OneTemp RTD Conversion Table ... http://www.intech.co.nz/products/temperature/typert.html ... the RTD should have been measuring at 106.63Ohm however I was getting a reading of about 109Ohm, suggesting it was about 24 degrees (quite a variation!). Do you think that the table is incorrect, like I do, or can you shed some light on an important factor I have overlooked? Thanks in advance, you guys are the shizzle! Scoota Edited June 3, 2008 by scootabug

[Mickey]

Normally the colors are just the other way around. Two red and one white. But you have what you have. As you have seen with your ohm meter the two whites are common and the red is on the other side of the RTD. You would then connect one white to the EXC term. the other white to the SENSE term. and the red to the RTN term. When you connect your ohm meter to an RTD the current running through the RTD will heat it up. The resistance reading is true though because of the increased temperature. Using an ohm meter will tell you if it is open or shorted and a ballpark reading.

[scootabug]

Thanks Mickey. I have wired it up in the configuration as you described, and it's now getting a reading. Yay for progress, thank you for the help. I have contacted the manufacturer this morning who is going to give me the correct conversion table...will post back when I make further progress.

[Mickey]

The error between your thermocouple and ohm meter is called "self heating" see link below for more reading on the subject. http://www.google.com/search?q=RTD+self+he...lient=firefox-a You will have to have an instrument designed to read RTD's .

[scootabug]

I'm not concerned about how accurate the RTD is, because for my purpose, I am sure that it will be fine. But surely a thermocouple and an RTD will be close, maybe even 1 degree different, but not 7. So I'm still working on the basis that RTD conversion table I have is incorrect. I realised after my first post that the table I was looking at was for another manufacturer and not the actual manufacturer of the RTDs I am using today. So I'll wait and see what I can find out from them, hopefully it will solve all my issues in one hit. Edited June 4, 2008 by scootabug

[scootabug]

Or are all RTD's made equal? I have found another conversion table and the data is the same as the first I found. How do I tell if I need to use "100 Pt 385" or "100 Pt 3916"? Edited June 4, 2008 by scootabug

[scootabug]

From another forum: "The two numbers represent the Thermal Coefficient of Resistance, Ohms/Ohm/°C, which is the amount the resistance of the RTD changes as the temperature of the RTD changes. The European coeeficient, 0.00385 Ohms/Ohm/°C is the most common, and if you have no contrary information is probably the correct value. The other coefficient, 0.00392 Ohms/Ohm/°C is the "American" curve, but it is not often encountered in either the US or Europe. Both types of RTDs have a resistance of 100 Ohms at 0 °C. At 100 °C the common European curve RTD will have a resistance of 138.51 Ohms, and the other curve will have a resistance at 100 °C of 139.16 Ohms. These two points allow crude calibration checks, since ice water is approximately 0°C and boiling water is approximately 100°C. There is variation with elevation and salinity and so on but for a quick check these values are close enough."

[Mickey]

Yup!! IMHO, you are not going to get your thermocouple and RTD to match using the method your using. Unless you have calibrated your thermocouple to some known standard along with your meter that thermocouple is going be +/- 4 deg out of the box. Add this to the self-heating of the RTD using the method you are using, then 7 deg. is not that bad.

[scootabug]

So what I just need to do is take 7 degrees off the reading I get, and call that an offset, and it'll be sweet?! ...which will at least make it consistent with my Fluke DMM?

[finfin]

Another error when reading with the ohm meter... The reason there are three leads is so the lead resistance can be subtracted from the reading. Measure the ohms between the two common leads and subtract that amount from the reading across the element. Edited June 4, 2008 by finfin

[scootabug]

Great point...TYVM