PT100 Calibration and typical offset

[coachmike]

My PT100 probe requires an offset of about +34 degrees F. when used with my 612 controller. Is this within the expected range or excessive?

Also, is there any reason not to use an ice water slurry for calibration at 32 F / 0 C? It would reduce the calibration error due to different atmospheric pressures/altitudes when using boiling water.

Thanks,
Mike

[richiem]

Most Pt100 probes are much better than 1% accuracy, so that offset makes absolutely no sense. Did you mean +34 deg or +3(point)4 deg? Note that extra long connecting wires can significantly raise the resistance, since they're in series with the sensing element. Most probes have the three or four-wire connections for a bridge type measurement to solve this problem. But the probe has to be connected correctly!

An ice/water slurry is very good for the 0 deg C point, and if you have a digital ohmmeter/multimeter, you can check the disconnected probe for 100 ohms (plus the connecting lead resistance -- typically a total of about 1 ohm, max) at that temp. If it's not REAL close to 100 ohms -- say between 99 and 101 (minus the lead resistance) -- it's defective and should be replaced.

The boiling water test serves as a good rough check for both the probe and the 612 -- the probe alone should measure around 139 ohms at 100 deg C.

In my experience, few people require accuracy of better than 5% for nearly anything.

See the attached file on RTDs

[coachmike]

Thanks for the info and interesting article.

I really meant +34. I am using the probe in a process right now so can’t make the suggested measurements until finished. The system seems to be accurate and stable in my application. It has been running about 30 hours so far and I have a backup thermometer for verification but I would feel more confident with a new probe and a tighter offset.

It wouldn’t hurt to have a spare or two anyway!

[Tech_Marco]

Hello Mike:

The offset shouldn't be over 3 degree that much. I think you better send it back for exchange.


Marco

[coachmike]

Hi Marco,

Thanks. I just ordered a new probe and will send the defective probe back for exchange when the new one arrives. Even though the offset is excessive, it is functioning and I don’t want to be without one. I don’t mind having a spare probe.

Thanks again,
Mike

[coachmike]

I received the new PT100 today... Much better!

Offset was only +.9 Deg F at 32 Deg F and 140 Deg F... The defective probe was +23 and +35.5 across the same range (hugely not even linear).

I sent the defective probe back today for exchange so I'll have a spare.

Marco, thanks for the great service and quick response.

Mike

[Tech_Marco]

Great! I'm glad that it fixed the problem.

By the way, if you're gonna to use it in water, you need to clean it, dry it after use or it will get rusted. A lot folks though that it's fully water proof and submersed in water then complained that getting radom reading. I ook like it is water-proof but it's not. A couple guys gave me bad rating on the probe about this issue. I do carry waterproof (fully protected) just in case you need one for water.

Have a great weekend.

Friday~! Friday~! Friday~! Friday~! Friday~!

Marco

[coachmike]

The probe I have is the water resistant model. (ETC-CB-PTWP2M) Is there one that is more waterproof?

[Tech_Marco]

Nothing better waterproof than the one you have now

[coachmike]

Awesome! Thanks again.

[darkglobe]

Interesting, I'm having a similar issue. This is what has happened:

1 - Installed PID
2 - Connected SSR
3 - Connected PT100
4 - Applied power

I noticed that the probe was reading ~41 F in a 71 F room, therefore underrating the real temperature.

I grabbed the probe to warm it up in my hand to see if it was responding at all, and it did, but according to it, my hand was 101 F, so now overrating (!@#!)

Thinking this was strange, I connected the heater and I warmed the water bath up to the desired 135 F, measured with a known reference, and then dialed the PID's PSb setting in. Eventually I got it spot on, but at one point I was low, and then high again, eventually setting down to +3 F for a correction.

I checked things this morning after letting it calibrate, to my satisfaction the display was showing 135 F, but the probe drifted and the real water temp was 139 F.

I then used my Fluke to take a reading while connected to the PID between the red and blue wire, which made the PID lost communication with the probe (went out of bounds perhaps, just got EEEE on the display). I saw this on a site on how to measure this, but realized that it wasn't working in this case (got 89 ohms for a reading so clearly the PID was influencing the reading).

When I removed the fluke, the temp went from 135 F to 123 F. Figuring I did something bad, I disconnected the PT100 and measured it, seemed to be off a little (cannot remember how much, just performed a rough calculation to what it should be). I reconnected it and now it read 109 F. I added a massive correction factor to ( +26 F ) and it seemed stable for a while, until I disconnected it again and now it's closer to +30 F for a correction.

Interesting that this number is close to the initial offset at room temperature (30 F difference), so not sure what is going on here.

I'm not sure if the PID is the problem or the probe. I did install an XLR type connector on it so that I could easily connect and remove the probe, but I don't think the problem is there (although I am still looking at it). I think I will try the boiling water / ice bath and meter it right on the wires just to see (and then back at the wires into the PID if the PT100 checks out)

I think I may order another PT100, not sure if I'll order another all plastic / flexy one like this or get a steel one.

If anyone has any suggestions, I'm all ears.

James

[richiem]

You discovered that you can't easily test a resistor's resistance when it is in the circuit -- when you test with ice water/boiling water make sure the probe is disconnected from the controller. That test will tell you whether the probe is good. As I've mentioned in other posts here, use the highest resistance range which will give you a reading of the three least significant digits -- the 10k (20k) range or higher so as to limit the current through the probe, which is very fragile, electrically.

If the probe is OK, then be sure that the controller has the correct Inty setting for your probe -- all the controllers I've bought have been set to a K-type thermocouple by default...

[darkglobe]

Ya, testing while in a circuit does not yield correct results unless you know what the other devices do to the whole circuit. The recommendation didn't make sense, but thought I would give it a go anyway, complete with bizarre results.

I did measure the probe at room temp and discovered it reading a hair over 100 ohms, which should not be possible unless it is in an ice bath, so clearly its resistance has dropped. I am not sure how I could have caused it (I do a *lot* of electronic repairs, upgrades and power supply building so I'm not new to all this) unless the soldering iron has a short , but even that should require a complete circuit. Likely cannot return it due to me removing the spades.

The Inty setting is correct for the probe type p10.0.

Looks like I will have to get another since this one looks like it is shot... bugger!

Thanks for getting back so soon.

Cheers!