Learning how to check pt100 sensor accuracy is one of those skills that saves you a massive headache when your readings start drifting or just don't make sense anymore. It's pretty frustrating when you're looking at a display and it's telling you your room is 150 degrees when it clearly isn't, or worse, when the sensor just dies completely in the middle of a job. Fortunately, these things are actually quite simple to troubleshoot if you have a basic multimeter and a little bit of patience.
You don't need a degree in electrical engineering to figure out if your sensor is toast. At its heart, a PT100 is just a platinum resistor that changes its resistance based on the temperature. The "100" in the name just means it's designed to have exactly 100 ohms of resistance when it's at 0° Celsius (32° Fahrenheit). If you can measure resistance, you can check your sensor.
Grab Your Tools and Get Ready
Before you start poking around, you'll need a few basics. Most importantly, you need a digital multimeter. It doesn't have to be a high-end, professional-grade model, but it needs to be able to measure Ohms (resistance) with at least one decimal point if you want any kind of accuracy.
The first thing you should do—and I can't stress this enough—is disconnect the sensor from whatever it's plugged into. If it's wired into a PLC, a temperature controller, or a transmitter, pull those wires out. Trying to test a sensor while it's still powered up or connected to other circuitry will give you weird, "phantom" readings that'll just confuse you. Plus, you don't want to risk shorting anything out.
The Basic Resistance Check
Once the sensor is sitting on your workbench or hanging loose, set your multimeter to the lowest Ohms range (usually 200 ohms). This is the quickest way to see if the sensor is even alive.
If you're at room temperature—let's say around 20°C to 25°C (68°F to 77°F)—you shouldn't see 100 ohms. Because the temperature is higher than freezing, the resistance should be higher too. Usually, for a healthy PT100, you'll see something between 107 and 110 ohms at room temperature.
If your meter reads "OL" (Over Load) or an infinite number, the internal platinum wire is broken. It's an open circuit, and unfortunately, there's no fixing that; it's time to buy a new one. On the flip side, if you see 0 ohms or something really low like 2 or 3 ohms, the sensor is shorted out. That usually happens if moisture gets inside the probe or if the insulation has melted.
Dealing with Different Wire Types
This is where people sometimes get tripped up. PT100 sensors come in 2-wire, 3-wire, and 4-wire versions. Don't let the extra wires scare you; they're just there to help the controller ignore the resistance of the lead wires themselves.
The 2-Wire Setup
This is the simplest. You've got two wires (usually one red, one white). Just put your probes on them and take the reading. The downside is that the meter is also measuring the resistance of the wires leading to the sensor, which can add a tiny bit of error.
The 3-Wire Setup
This is the most common one you'll see in industrial settings. You'll usually see two wires of the same color (like red) and one of a different color (white). If you measure between the two identical colors (red to red), you should see almost zero ohms—usually just 0.1 or 0.2. That's because those two wires are connected to the same side of the sensor. To actually check the sensor, you need to measure between one red and one white wire.
The 4-Wire Setup
These are the gold standard for accuracy. You'll have two pairs of colored wires. Just like the 3-wire version, wires of the same color should show zero resistance between them. To get the temperature reading, measure between any wire of color A and any wire of color B.
The Ice Bath Test (The Gold Standard)
If your resistance reading is "close enough" but you're still suspicious that the sensor is lying to you, you need a more controlled test. This is where the ice bath comes in. It's a classic move because it gives you a very reliable 0°C reference point.
Don't just throw a few ice cubes in a glass of water, though. That's a rookie mistake. To do it right, you need to crush the ice as finely as possible. Fill a thermos or an insulated cup with the crushed ice and add just enough cold water to fill the gaps between the bits of ice. You want a thick slush, not ice cubes swimming in water.
Stir it around and let it sit for a few minutes so the temperature stabilizes. Then, stick your PT100 probe into the middle of the slush. Make sure it isn't touching the bottom or sides of the container. Wait a minute or two for the reading to stop moving.
If the sensor is perfectly calibrated, your multimeter should show exactly 100.0 ohms. In the real world, if you see anything between 99.8 and 100.2, you're in great shape. If it's off by a whole ohm or more, your sensor has "drifted," which happens over time due to heat cycling or vibration.
Checking the Insulation
Sometimes a sensor works fine most of the time but acts crazy when it's installed in a machine. This is often an insulation resistance problem. Basically, the electricity is leaking from the platinum wire to the metal outer sheath of the probe.
To check this, set your multimeter to the highest resistance setting (mega-ohms if possible). Touch one probe to one of the sensor wires and the other probe to the metal body of the sensor probe itself. You should see "OL" or an extremely high resistance. If you see any kind of low reading here, the sensor is grounded out internally. This will cause all sorts of erratic behavior once it's hooked up to a controller, especially if there's electrical noise nearby.
The Boiling Water Test
If you want to check the other end of the scale, you can use boiling water. It's not quite as perfect as an ice bath because the boiling point of water changes depending on your altitude (if you're in the mountains, water boils at a lower temp), but it's still a great sanity check.
At sea level, boiling water is 100°C. A PT100 at 100°C should read roughly 138.5 ohms. If you're at a higher altitude, it'll be slightly less. It's just another way to make sure the sensor is actually responding to heat linearly like it's supposed to.
Why Do These Sensors Fail Anyway?
You might be wondering why you're having to figure out how to check pt100 sensor health in the first place. Platinum is pretty stable, but it isn't invincible. The most common killer is vibration. If the sensor is mounted on a shaky motor or a high-pressure pipe, those tiny internal wires can eventually fatigue and snap.
Moisture is another big one. If the seal on the end of the probe isn't perfect, humidity can get inside. Water and electricity don't mix, and even a tiny bit of moisture can create a parallel path for current, which makes the resistance look lower than it actually is. This results in the controller thinking the temperature is lower than reality, which can be dangerous depending on what you're heating.
Wrapping Things Up
Testing a PT100 doesn't have to be a chore. Most of the time, a thirty-second check with a multimeter will tell you everything you need to know. Just remember the magic number: 100 ohms at 0°C. If you keep that in mind and know how to navigate your 2, 3, or 4-wire setup, you'll be able to spot a bad sensor from a mile away.
It's always a good idea to keep a spare sensor on hand anyway. If your tests show that your current one is acting flaky, don't try to "calibrate" your way out of it in the controller settings. A drifting sensor is a dying sensor, and it's usually better to just swap it out for a fresh one and get back to work.