RTD PT100 BEARING TEMPERATURE READING NOT STABLE

HELLO,,
we are facing problem on bearing temp. , RTD PT100 reading change too much for rising and falling 70 to 80
we tight the terminations also replace the sensor by new one and same the problem.

also, we observed the reading some times become smooth and straight, the terminal box have all PT100 for the pump and all the reading are fluctuating
we need the recommendations.
 
Have you tried:

1) Exchanging the non-working RTD for another working RTD on the monitoring system? If so, what were the results? (If the problem "follows" the intermittent RTD to a new channel/input on the monitoring device, then there's something wrong with the interconnecting wiring, and/or the RTD. If the problem remains on the same input channel, then it's the monitoring device that has a problem.)

2) Using a decade box/RTD simulator on the monitoring device input channel the intermittent RTD is connected to? If so, what were the result?(If you connect a decade box/RTD simulator directly to the monitoring device and the signal doesn't drift over time, then it's the interconnecting wiring and/or the RTD.)

3) Disconnecting the interconnecting wiring from the monitoring device and the intermittent RTD and meggaring the wiring? (wire to wire, and wire to ground) If so, what were the results?

4) Have you confirmed the RTD wiring is properly terminated at the monitoring device channel input terminals and the RTD itself? Some monitoring devices are VERY particular about which wire is connected to the power ("excitation") terminal, and which wire is connected to the COMP terminal, and which wire is connected to the sense ("signal") input?

5) Some monitoring devices can use three-wire or four-wire RTDs; some have only three terminals, but a four-wire RTD is used. Are you certain of the terminations (wiring configuration)?

6) It's presumed the circuit is wired using twisted, shielded wiring/cabling, and that the shield drain wires are only terminated at one end. Have you verified the shield drain wire terminations? (Ideally, the interconnecting wiring would be three-wire, twisted, shielded cables, with an over-all foil shield and a bare drain wire (sometimes called a "triad" cable); what wire/cable is usefor the interconnecting wiring at your site?

7) You've said you tightened the terminations; did you check the terminal lugs for tightness? (Sometimes, crimp-on terminal lugs are not properly installed and can become loose over time.)

8) Is there any high-voltage and/or high current device nearby to the pump that starts and runs when the reading difference starts to increase, or that shuts down when the reading difference normalizes? (This would indicate there is some kind of induced voltage/current on the interconnecting wiring leads which is causing a problem.)

Please write back to let us know what you find!
 
Have you tried:

1) Exchanging the non-working RTD for another working RTD on the monitoring system? If so, what were the results? (If the problem "follows" the intermittent RTD to a new channel/input on the monitoring device, then there's something wrong with the interconnecting wiring, and/or the RTD. If the problem remains on the same input channel, then it's the monitoring device that has a problem.)

2) Using a decade box/RTD simulator on the monitoring device input channel the intermittent RTD is connected to? If so, what were the result?(If you connect a decade box/RTD simulator directly to the monitoring device and the signal doesn't drift over time, then it's the interconnecting wiring and/or the RTD.)

3) Disconnecting the interconnecting wiring from the monitoring device and the intermittent RTD and meggaring the wiring? (wire to wire, and wire to ground) If so, what were the results?

4) Have you confirmed the RTD wiring is properly terminated at the monitoring device channel input terminals and the RTD itself? Some monitoring devices are VERY particular about which wire is connected to the power ("excitation") terminal, and which wire is connected to the COMP terminal, and which wire is connected to the sense ("signal") input?

5) Some monitoring devices can use three-wire or four-wire RTDs; some have only three terminals, but a four-wire RTD is used. Are you certain of the terminations (wiring configuration)?

6) It's presumed the circuit is wired using twisted, shielded wiring/cabling, and that the shield drain wires are only terminated at one end. Have you verified the shield drain wire terminations? (Ideally, the interconnecting wiring would be three-wire, twisted, shielded cables, with an over-all foil shield and a bare drain wire (sometimes called a "triad" cable); what wire/cable is usefor the interconnecting wiring at your site?

7) You've said you tightened the terminations; did you check the terminal lugs for tightness? (Sometimes, crimp-on terminal lugs are not properly installed and can become loose over time.)

8) Is there any high-voltage and/or high current device nearby to the pump that starts and runs when the reading difference starts to increase, or that shuts down when the reading difference normalizes? (This would indicate there is some kind of induced voltage/current on the interconnecting wiring leads which is causing a problem.)

Please write back to let us know what you find!

Hi, I am facing the pt100 fluctuation which is installed in a cold room temp -20C. We used Belden shielded cable and is connected with Siemens s7 300 plc n tp900 hmi.

Sensor is installed behind evaporator. When Evaporator is off, the readings are constant. However, when start the Evaporator fans, the sensor readings starts fluctuation. So it look like there's a power interference.

Tired to connect the Belden cable silver wire with ground on both sides but problem remain same.

We have same problem in all sensors and need an advice what further tests should be carried out to resolve this issue.
 
Siraj,

First of all, connecting the drain wire (the uninsulated wire in the Belden cable) to ground at both ends is completely incorrect and will only make matters worse. The idea of the drain wire (which is in contact with the foil shield of the cable) is to provide a SINGLE path to ground for any electrical noise which is "absorbed" by the foil shield. Connecting the drain wire to ground at both ends provides a circular path for current to flow (through the drain wire to ground at one end and back to the drain wire from the ground at the other end--ground (earth) is just like a conductor). The drain wire is intended to shunt any voltage developed on the shield to ground--at one end--thereby reducing interference with the signals carried by the conductors surrounded by the shield. By connecting the drain wire to ground at both ends, a path for current to flow is created and that will definitely negatively affect the signals in the wires surrounded by the foil shield. So, correct that problem right away.

Belden is a manufacturer of many types of wires and cables; their name has become synonymous with multi-conductor shielded cables (two-wire (plus the drain wire); three-wire (plus the drain wire); four-wire (plus the drain wire); and so on). Which type of shielded cable is used in your application?

There are at least three types of RTDs: two-wire, three-wire and four-wire. Which is used in your application?

How are the shielded cables routed to and from the Evaporator? At ANY point along the route are the shielded cables run in the same conduit as the Evaporator fan motors? Or in the same cable tray? The RTD signals are considered to be low-level (usually less than 50 V), as opposed to motor wiring which is considered to be high-level (usually more than 120 VAC). Mixing the two types of levels (low and high) in the same conduit or cable tray, even for a short distance (a few meters, even) is risking just the type of signal interference you are describing--noise on the low-level wiring caused by high current (even just a few amps) in the high-level wiring in close proximity to the low-level wiring.

Have you looked at the configuration software for the RTD inputs? Sometimes, there is filtering which can be added to the input signals which might be helpful in your case.

Without knowing a LOT more about the installation and facts at your site there's not too much more we can say. It might also be possible that the location of the RTDs themselves are too close to the Evaporator fan motors, and that the metal of the evaporator coils is aggravating that issue.

Hope this helps--please write back to let us know how you fare in resolving the problem(s).
 
Siraj,

First of all, connecting the drain wire (the uninsulated wire in the Belden cable) to ground at both ends is completely incorrect and will only make matters worse. The idea of the drain wire (which is in contact with the foil shield of the cable) is to provide a SINGLE path to ground for any electrical noise which is "absorbed" by the foil shield. Connecting the drain wire to ground at both ends provides a circular path for current to flow (through the drain wire to ground at one end and back to the drain wire from the ground at the other end--ground (earth) is just like a conductor). The drain wire is intended to shunt any voltage developed on the shield to ground--at one end--thereby reducing interference with the signals carried by the conductors surrounded by the shield. By connecting the drain wire to ground at both ends, a path for current to flow is created and that will definitely negatively affect the signals in the wires surrounded by the foil shield. So, correct that problem right away.

Belden is a manufacturer of many types of wires and cables; their name has become synonymous with multi-conductor shielded cables (two-wire (plus the drain wire); three-wire (plus the drain wire); four-wire (plus the drain wire); and so on). Which type of shielded cable is used in your application?

There are at least three types of RTDs: two-wire, three-wire and four-wire. Which is used in your application?

How are the shielded cables routed to and from the Evaporator? At ANY point along the route are the shielded cables run in the same conduit as the Evaporator fan motors? Or in the same cable tray? The RTD signals are considered to be low-level (usually less than 50 V), as opposed to motor wiring which is considered to be high-level (usually more than 120 VAC). Mixing the two types of levels (low and high) in the same conduit or cable tray, even for a short distance (a few meters, even) is risking just the type of signal interference you are describing--noise on the low-level wiring caused by high current (even just a few amps) in the high-level wiring in close proximity to the low-level wiring.

Have you looked at the configuration software for the RTD inputs? Sometimes, there is filtering which can be added to the input signals which might be helpful in your case.

Without knowing a LOT more about the installation and facts at your site there's not too much more we can say. It might also be possible that the location of the RTDs themselves are too close to the Evaporator fan motors, and that the metal of the evaporator coils is aggravating that issue.

Hope this helps--please write back to let us know how you fare in resolving the problem(s).
Firstly thank you for your prompt response and explanation. We highly appreciate that. Please see response below.




CA: Which type of shielded cable is used in your application?
SK: four-wire (plus the drain wire).

CA: There are at least three types of RTDs: two-wire, three-wire and four-wire. Which is used in your application?
SK: three-wire and I confirm configuration is good.

CA: Have you looked at the configuration software for the RTD inputs?
SK: Yes, I confirm it's correct


CA: How are the shielded cables routed to and from the Evaporator? At ANY point along the route are the shielded cables run in the same conduit as the Evaporator fan motors? Or in the same cable tray?
SK: One PT100 is behind the Evaporator (approx 1 feet) and one in the coil of Evaporator. The power cable and shielded cable run is same cable tray.

We will group the drain wire from one end and see the difference.

once again thank you for your support.
 
Siraj,

If you are using four-wire plus a drain shielded cables but the RTDs are only three-wire RTDs, then you need to do something with the fourth, unused wire of the cable to prevent any noise it "absorbs" from affecting the three wires with the RTD signals on them. There are a couple of options. One, you can simply ground one end of the unused conductor to drain away any noise from that conductor to ground. Second, you could connect it in parallel with any of the other conductors (so you would end up with two conductors for say the excitation or the compensation or the signal leg)--meaning that both ends of the unused conductor would be terminated with both ends of one of the other three conductors. But, you should do something with that other "antenna" (the unused fourth wire of the shielded cable--especially if the low- and high-level wires/cables are run in the same cable tray).

And, while I know it's probably too late to correct that problem--running low- and high-level wires/cables in the same cable tray--if possible, you should try to do that. Because, as you have already noted--the interference only occurs when current is flowing in the high-level wires....

You may see an improvement in the signals when the Evaporator motors are running when you disconnect one end of the shield drain wire from ground--but as long as the low-level and high-level signals are run in the same cable tray that's going to be a problem. That's a serious oversight on someone's part (equipment designer; equipment installer; someone), and while it's not usually a technical regulation or code violation--it IS a design/construction standard in virtually every part of the world. Designers should know this and take this into consideration when designing equipment, and also make notes to installers about the need to isolate different signal levels. And, construction superintendents and installers should also know this and follow appropriate construction practices to prevent mixing signal wires in the same cable tray or conduit.

You might try, if it's possible to separate the low- and high-level wires/cables as much as possible in the cable tray(s). Along with properly terminating the shield drain wire at one end only that may be enough to get the problem reduced or even eliminated. But, if the cables are run in close proximity over long distances, .... well that's likely to be a problem that's going to be difficult to solve.

I have seen--as a temporary measure--running a separate shielded cable outside of the same cable tray, say on the floor, or some other method as appropriate, for a test on one RTD signal (following proper wiring practices as noted above with the conductors and the drain wire!) to see if getting the low-level signal away from the high-level Evaporator motor wiring helps to eliminate the problem.

Please write back to let us know how you succeed in solving the problem(s)!!!
 
1. > The power cable and shielded cable run is same cable tray

That's your problem. The motor current is inducing noise in the RTD.

At 0°C, the RTD signal is only 100mV.

2. > If so is the sensor electrically connected with the sheath, so called grounded tip sensors?

Grounded tip sensors are common for thermocouples but would be serious construction fault/failure in an RTD.

The receiver device sources a constant current through the RTD element, something on the order of 1mA or less, and then reads the voltage drop created by the varying resistance of the RTD element. If the element were grounded to the sheath, there would be a high probability of a ground loop that would add or subtract current to the loop depending on the ground potential differences between the receiver and the RTD element.

A typical check of an RTD is a continuity check of each RTD lead wire to the sheath, which should be an open circuit/high resistance/no connection.

RTD's are by definition, "isolated" sensors and if the isolation is damaged then the RTD needs to be replaced.
 
Siraj,

If you are using four-wire plus a drain shielded cables but the RTDs are only three-wire RTDs, then you need to do something with the fourth, unused wire of the cable to prevent any noise it "absorbs" from affecting the three wires with the RTD signals on them. There are a couple of options. One, you can simply ground one end of the unused conductor to drain away any noise from that conductor to ground. Second, you could connect it in parallel with any of the other conductors (so you would end up with two conductors for say the excitation or the compensation or the signal leg)--meaning that both ends of the unused conductor would be terminated with both ends of one of the other three conductors. But, you should do something with that other "antenna" (the unused fourth wire of the shielded cable--especially if the low- and high-level wires/cables are run in the same cable tray).

And, while I know it's probably too late to correct that problem--running low- and high-level wires/cables in the same cable tray--if possible, you should try to do that. Because, as you have already noted--the interference only occurs when current is flowing in the high-level wires....

You may see an improvement in the signals when the Evaporator motors are running when you disconnect one end of the shield drain wire from ground--but as long as the low-level and high-level signals are run in the same cable tray that's going to be a problem. That's a serious oversight on someone's part (equipment designer; equipment installer; someone), and while it's not usually a technical regulation or code violation--it IS a design/construction standard in virtually every part of the world. Designers should know this and take this into consideration when designing equipment, and also make notes to installers about the need to isolate different signal levels. And, construction superintendents and installers should also know this and follow appropriate construction practices to prevent mixing signal wires in the same cable tray or conduit.

You might try, if it's possible to separate the low- and high-level wires/cables as much as possible in the cable tray(s). Along with properly terminating the shield drain wire at one end only that may be enough to get the problem reduced or even eliminated. But, if the cables are run in close proximity over long distances, .... well that's likely to be a problem that's going to be difficult to solve.

I have seen--as a temporary measure--running a separate shielded cable outside of the same cable tray, say on the floor, or some other method as appropriate, for a test on one RTD signal (following proper wiring practices as noted above with the conductors and the drain wire!) to see if getting the low-level signal away from the high-level Evaporator motor wiring helps to eliminate the problem.

Please write back to let us know how you succeed in solving the problem(s)!!!
Good that you highlighted to ground the 4th conductor. At the moment it's just present there and covered with insolation tap (From sensor end imagine only 3 wires of Belden cable are visible 2" length bottom is covered and it's connected to sensor wires through connector nice & tight. While from plc control side, 3 wires+ drain wire are visible and drain wire connected to ground)

However, if connect the 4th wire with one of the 3 other conductor (both sensor end and plc terminal end) then the temperature reading are incorrect. ( 10 to 15 degrees difference between actual and displayed temp on hmi).

I would try your tip to use separate cable with sensor and at distance from the power cables of evap fans.

Just wanted to inform, our application is cold rooms and the facility is quite big. So consider that the sensor distance from plc is min 100 or 200 meters plus.
 
siraj, are your RTD's sheathed? If so is the sensor electrically connected with the sheath, so called grounded tip sensors?
Yes Dave, the sensor is sheathed and it comes with standard 3 core 1.5 meter wire which we connect with belden cable (4 core, we used 3 wires and cover the 4 wire+ foil shielded+drain wire). We'll follow the tips mentioned above regarding grounding only one end of drain wire along with 4th spare wire.
 
1. > The power cable and shielded cable run is same cable tray

That's your problem. The motor current is inducing noise in the RTD.

At 0°C, the RTD signal is only 100mV.

2. > If so is the sensor electrically connected with the sheath, so called grounded tip sensors?

Grounded tip sensors are common for thermocouples but would be serious construction fault/failure in an RTD.

The receiver device sources a constant current through the RTD element, something on the order of 1mA or less, and then reads the voltage drop created by the varying resistance of the RTD element. If the element were grounded to the sheath, there would be a high probability of a ground loop that would add or subtract current to the loop depending on the ground potential differences between the receiver and the RTD element.

A typical check of an RTD is a continuity check of each RTD lead wire to the sheath, which should be an open circuit/high resistance/no connection.

RTD's are by definition, "isolated" sensors and if the isolation is damaged then the RTD needs to be replaced.

Alright David, we will try to keep the power cable away from sensor cable and also would check the continuity between RTD leads and sheath. Thanks for your suggestions.
 
Siraj,

Thank you for your responses and feedback; it's much appreciated by all who are contributing to this thread and all who are following it.

That's a phenomenom I haven't run into before--though I guess it could happen if you connected the 4th conductor with the signal leg or the compensation leg (resistances in parallel and all that--I should have though of that; sorry). Did you try connecting it with the excitation leg?

The last thing I would add is have you consulted the Siemens RTD information for maximum wire length?

Looking forward to hearing the results of your efforts!
 
> One, you can simply ground one end of the unused conductor to drain away any noise from that conductor to ground. Second, you could connect it in parallel with any of the other conductors (so you would end up with two conductors for say the excitation or the compensation or the signal leg)--meaning that both ends of the unused conductor would be terminated with both ends of one of the other three conductors.

You don't want to connect an used 4th wire in parallel with any of the lead wires because it will cause an offset error.

The assumption in the measurement of a 3 or 4 wire RTD is that all the lead wires are of the same length/gauge/diameter/resistance. Under that assumption, the lead wire resistance is taken into account in the 4th leg of a classic bridge circuit.

Here's a sketch of a typical 3-wire RTD circuit:

3-wire_RTD_bridge_circuit.jpg

If the 4th wire is paralleled to any of the other lead wires, the resistance for that paralled leg is NOT the same as the resistance for the other legs and when the assumption of equal-resistance-leads if false, the measurement is false.

Grounding the 4th wire on one end as though it were part of the shield/drain should suffice.

2. The Twist in shielded twisted pair is as important in noise rejection as the shielding/drain. You're sure the cable has twisted conductors?

shielded-twisted-pair.png
 
> One, you can simply ground one end of the unused conductor to drain away any noise from that conductor to ground. Second, you could connect it in parallel with any of the other conductors (so you would end up with two conductors for say the excitation or the compensation or the signal leg)--meaning that both ends of the unused conductor would be terminated with both ends of one of the other three conductors.

You don't want to connect an used 4th wire in parallel with any of the lead wires because it will cause an offset error.

The assumption in the measurement of a 3 or 4 wire RTD is that all the lead wires are of the same gauge/diameter/resistance. Under that assumption, the lead wire resistance is taken into account in the 4th leg of a classic bridge circuit.

Here's a sketch of a typical 3-wire RTD circuit:

View attachment 1159

If the 4th wire is paralleled to any of the other lead wires, the resistance for that paralled leg is NOT the same as the resistance for the other legs and when the assumption of equal-resistance-leads if false, the measurement is false.

Grounding the 4th wire on one end as though it were part of the shield/drain should suffice.

2. The Twist in shielded twisted pair is as important in noise rejection as the shielding/drain. You're sure the cable has twisted conductors?

View attachment 1160
David, thanks for the useful information on RTD sketch.

Refer to the picture below. Yes I confirm the Belden cable has 2 twisted pairs plus 1 drain wire. Both pairs are separately shielded (red&black in one foil, white n green in another. We are using red,black and white. While white and drain wires will be grounded from one end as per your instructions.61toilF3uKL._SS400_.jpg
 
Top