Sensor Valve Fluctuations

You can dampen the pressure transitions, either mechanically or electronically.

mechanically with a snubber, which be either
- a length of small diameter capillary tubing through which the pressure pulses are attenuated as they move through the restricted flow path.
- needle valve in a commercial style snubber

Electronically - averaging input readings over a period of time is a selectable parameter in all smart transmitters

The 3051 has this about damping in its manual:

If there's a PID controller controlling the position of the valve, then the PID tuning could be too aggressive, causing the valve to oscillate.

If so, you need to temporarily put the loop in manual and then with the valve holding its position, check to see whether the pressure fluctuates.

Otherwise the pressure transmitter might be showing you the results of poor PID tuning.

thanks for ur reply but actullay we r using only two wire current loop for this measurement(NO use of HART) and using WellWizard s/w and Datalog server. we just install the sensor and connect to JB. it give value to the server. but we r also using the mechanical dead cell for getting the pressure measurement. and we have installed 4000/5000/10000 psi rating sensor (8051 rousemount) but all the sensor are fluctuating. now how can I fix that and u can tell how we good calibration for this type of sensors

"Slug it" - i.e. install a low-pass filter.

From my seat, it could be one of two problems:

1. As stated above, the PID controlling the valve may not be tuned, assuming there is one set up for your application. I have used PLC's where the pressure reading MUST be converted or "scaled" to be used in the PLC's PID loop configuration. Is this your case?

2. The response time from the sensor to the server and back to the valve control output may be too long of a time frame. The valve may be reacting to a sensor input that has actually changed from the input used in the output calculation. Check your response time. You may need local control at this valve and only monitor the results back to your server.

With the pressures you have described, I would resolve this issue immediately for safety reasons. Please send additional information for a more thorough evaluation.

How are you fixed for test equip. over there? This calls for a 4-20 loop calibrator and in any Rosemount that dampening parameter is accessible with a 375 or 475 HART communicator.

By referencing your probem I'm agree with the solutions provided by the experts.

But I'ld like to know more to suggest you the right thing.

Till that as per my consideration (by understanding your problem) I'ld like to suggest you that check your transmitter location i.e. whether it is connected too closer to the process source? Many times it is observed that if the sensor(transmitter) is located too closer to the process source, then this kind of fluctuation problem persists.

asifidrsi,

Many good suggestions already--proximity to a pulsating pressure source, PID loop tuning, HART protocol parameters, and low-pass filter on sensor output. If you have eliminated or minimized the probability of pulsating pressure, closeness to an elbow (prefer at least 5x the pipe diameter as a minimum distance), and tuning---then consider possible sources of electrical noise. These could include grounding the analog cable at both ends (to form a "ground loop" that cancels most of the value of a shielded cable), and running the cable close to and parallel to noisy wires such as power conductors and particularly VFD motor leads.

More information would help us to give a more refined answer to your problem.

--JMM

thanks dear, by ur suggestion i have relocate the sensor position now the value is less fluctuating......