Analog temperature indicator and HMI values not equal

We are having trouble minimizing the differences between the values from an analog temp. indicator and HMI.

The analog temp meter is installed in a generator while an RTD is installed in the same generator. The problem is that we have configured the PLC (HC900 c50) to read the values given by the RTDs but the output was way different from the values in the analog temp indicator. We have tried to minimize the difference by adding some bias. The method worked but when we came back the next day, the values have point differences again. We can't tell where the problem is but it really causes us some problems...

I assume both your local gauge and RTD are in the same place, if so either the gauge is wrong or you have a wiring fault. Was the local gauge reading the same both days? Is the RTD wired with 3 wires? Does your RTD type match your input card 00385 or 00392, other?

Assuming your RTD is not in a trip circuit it should be easy enough to check the wiring. If you have access to a decade box or good accurate Ohmeter and rheostat substitute a couple of known values to check the PLC e.g. 100 for 0C, 138.5 for 100C (assuming that your RTD type is 00385). If you can remove both devices, put them both together with a good mercury thermometer in water at different temperatures. Get back to us with the results and we will go from there.

Regards
Roy

If these are readings from two different RTDs, then there might be two different temperatures. If not then you need to calibrate the analogue signals. Adding an offset isn't enough, you need offset and gain. This is a y = mx + b type of equation. The offset ("bias") just adjusts the "b" and not the "m".

The normal way of dealing with this is the program in the PLC converts the voltage or current to theoretical engineering numbers (e.g. 10V = 300C, or whatever it should be) by applying a y = mx + b type calculation. The calibration adjustments on the signal conditioners are then adjusted to correct for actual electrical signal errors.

What kind of problem can a reading from analog indicator produce?

What is being measured?

How much error are you seeing between the two readings? What magnitude? Is this a 0.1°F difference or 10°F? At what operating temperature?

Are you recording the error or comparitive measurements?

Does the error change at different operating values?

Exactly what kind of analog indicator? bimetalic thermometer? mercury thermometer? bulb/gas/vapror filled system? capillary?

How, mechanically, is the analog device installed?

Is either sensor installed in a well? Does the sensor tip bottom in the well?

Is the installation indoors or outdoors? In the sunlight?

How is the zero adjusted on the analog device? adjuster screw?

When was the analog unit last calibrated? at how many points? At what temperaturess?

Is the RTD a 3 wire DIN 385? All three wires connected?

Was the RTD new, or salvaged from another project? Was it calibrated?

Has the HC900 AI been field calibrated? Was the calibration trustworthy?

The HC900 has 'universal' analog inputs: any input can be configured for T/C, RTD or mA. Has any other input on the AI card with the RTD in question not read correctly?

Bud

We found out that the analog gauge uses a 3 wire PT100. Also, we found out that the source of the plc's signal was tapped to the line of the rtd to the analog meter. It now bothers me because only the two wires were tapped from the rtd to the plc which should be 3 wires. Could this be the reason why im getting wrong values? another thing.. i have noticed that some of the values in the plcs suddenly drops to zero when it reach a certain value. Some says it might be shorted? or it was caused by interferences. However, this happens in a cycle...

Siopao,
From what you describe I think you have just one RTD with both the Local and PLC connected. Although not impossible this is very unusual.
The best solution would be to replace the single RTD with one with dual elements, dedicating one to the local indicator and one to the PLC.
You will also need to run a 3 conductor cable, this is the reason the readings drift, as the ambient temperature changes your lead wire resistance changes which causes a Zero shift. 3 wires overcomes this problem.

Try to find out a bit more about your instruments so we can help you further.
Make / model
RTD type 100 Ohm Pt 00385 or 00392?
Control system range settings.
Attach your e-mail address and I will send a typical sketch.

Regards
Roy
[email protected]

Okay heres the detailed info...
The analog Meter was a speedomax type which uses a 3-wire Pt100 rtd sensor. It was connected within the generator. When the system was upgaded, a PLC and an HMI(wonderware) was installed to provide remote monitoring. However, the rtd that the local analog meter uses was also the one used as the AI of the PLC. That means, the RTD was shared by both the analog meter and the PLC. Also, only two wires were used to run to the PLC from the field.Then the problem was different readings between the HMI and the analog meter. Could this be the reason why theres a difference between the values? ANd the other problem was the sudden drop of values to zero when it reaches a certain value. It happens in a cycle and it was reflected to the trends app of the HMI... I am planning to rewire the connections but is there a way to solve this one without the need of rewiring? Thanks...