I/P Transducer

I don't know what an HT water G-valve is (high temperature globe valve?) but it could be that the line pressure is greater than the control valve can handle. We used to have this problem. The water pressure had enough force to push the valve plug up. We solved it by adding more springs in the diaphragm housing.

...so it could be that the valve itself is installed sdrawkcab, for example.

The installation cannot be seen from here.

what is the pressure deference between the valve plate,,,,

"delta P" arround the valve gate ;)

Hi,

is the problem solved now?

you calibrate P from I/P ,if so i think that the problem was in the valve itself , some of valve constant changed with time like spring constant for diaphragm , corrosion on the screw and any other changes because of environment,

but if calibrate i get to the I/p i think that the problem from the safety barriers or PLC

If by calibration, problem is solved, that must be the problem.

But other problems may be

Some leakage in discharge line of I/P. If it is leaking, though the command signal is 20 mA (or 100%), output may not reach to 100%, as the situation is like infinite actuator volume.

Or may be PLC itself is not giving 100% signal, though your command ladder is designed OK.

You have to cross check and validate every loop.

I would make sure that your valve spring is deigned when ordered to operate at your current process pressure I really need more info to help you how big is your valve ? what type of process are you trying to control?

Dear

how did you know that the valve is not closing? do you have a feedback from the valve??

when you calibrate it, it closes??

did you measure the signal comming out from from the PLC?? it could be not adjusted??

it need to be re scalled??

Dear sir,

Thanks for ur reply.

i measured PLC output i.e. input to transducer. it is showing 20 mA & at this condition valve is to be in circulation mode ( temp. should increase but it is decreasing), also i valve indicator position but was not closing 100 %.

It is happenning sometimes ( not always). problem persisting again & again.

Ah! Intermittent problem. That changes things.

I still don't know what type of valve you're using (electro-pneumatic or motorized). You may have a loose connection somewhere in your signal lines.

In a pneumatic valve, if your compressed air line has a heavy load that switches in from time to time, it may cause a drop in line pressure causing the valve to partially close.

For a motorized valve, you may have a problem with the limit-of-travel switches or on the feedback resistor.

Your use of terms is a bit confusing (for me at least). You are saying that 20mA is equivalent to 100% and the system should be in circulation mode.

However, I find your symptoms to be a bit strange. You're saying that the valve has a 20mA signal and the system should be in circulation mode. Your temperature should be increasing but instead it's decreasing ? It seems that your valve is operating in reverse except you're saying that it's intermittent. Have you checked your PLC program? There may be a condition where the valve operation reverses (though I can't think of why that would be done).

Perhaps you can give us a more detailed description of your problem. You've been giving us bits and pieces which is not really conducive to proper problem shooting. Please be more forthcoming with your descriptions.

regards,

Vulcan

If the PLC signal goes into the I/P and the valve does not close at the proper signal (for example 4 mA) but the valve does close when you put an external 4 mA signal on the I/P then the I/P is not the problem. If both tests were done with the valve in operation then the problem is not the valve either. Pressure shouldn't be a problem in a 3-way system since the flow always has a place to go. If the only difference in your test is PLC or no PLC then you need to check the calibration of the PLC output. This is more easily done in some models than others, but you will need to put a mA meter in series with the valve signal. Then send your output to the valve. If you are not reading the proper mA value (say 4.02 mA vs. 4.00 mA) you need to adjust your output signal scaling until you get the value you want.

Hi andrew_ward!

I'm with you on this. The test signal operated the valve correctly. There doesn't seem to be an efficient signal from the PLC, since everything else seems to be in working order. Maybe a crowded activity schedule is preventing adequate signal supply. send2vaiju didn't tell us under what conditions (time of day, activity in the PLC, or even activity in the water lines) the problem is consistent, intermittent, or the valve works properly either.

Maybe his diagnosis needs a little more investigation around the circumstances, in addition to the work he has done on the valve and the power source.

If send2vaiju tells us that this turns out to be the answer, you'll get a GA from me.

Mark

Dear sir,

Thanks for ur reply.

problem is persisting intermittently, i.e. presently & at the time of calibration found all signals ok.( this malfunctioning happenned four times.)

Hi, send2vaiju!

OK. "Intermittent" is a valuable clue to solving your mystery.

It's time to include environmental causes in your search for the problem. Anyone reading this who can add to the following suggestions for environmental causes for malfunction?

Check your records for the following information:

1. What times of day/night did the valve not close properly?
2. How long does the PLC signal to close the valve last? Is it long enough to close it all the way?
3. What else was sharing the PLC signal output at exactly those times?
4. What else was sharing the same water supply (at any point from source to end-use) at exactly those times?

You are looking for

1. a very short signal from the PLC that was programmed without allowing the valve to close completely: in other words, the PLC signal shuts off too soon, and the valve has not closed all the way yet, and/or
2. an overload on the PLC output at certain times of day that would prevent it from signalling the valve for a sufficient time or with sufficient strength to have it close all the way, and/or
3. you are looking for variations in the water pressure on the line used by the valve and comparing them to the valve's pressure rating.

If you find that

• the PLC signal shuts off too soon and the valve has not closed; or
• the PLC has too many tasks to do simultaneously at various times during the day and so it can't send an adequate signal to the valve,

you will have to re-program some of its tasks.

If you discover that the pressure against the valve is too high for its closing mechanism rating, you will have to change the valve.

Mark

IF the valve works properly with an injected signal of proper value, then the problem is not the I/P, it is the input signal into the I/P from the PLC .Put a milliamp meter in series with the signal from the PLC to the I/P and you will find your error.Sounds to me like the zero (4 MA) has drifted upward on the PLC. The I/P does not care where it gets it's signal, it should respond the same across the full span of the input signal.

Hi, send2vaiju!

I don't think you have a ground fault in the feed line from the PLC to the valve. If you did, then your successful test imitation signal buzz would not have closed and/or re-opened it on command.

However, if you are going to check out a possibility of a fault, then please add the following further to what the Guest writing in suggests...

If you should find that the in-line throughput on the meter put is reasonable to what you would expect, put your meter right next to the valve for the first fault check test and send your test signal through it to be sure it is working as it did when you originally thought to check the line this way.

Then check back through the conduit link by link with the meter (standard fault location testing) until you find a length with the fault in it. Replace that length.

(Gee, I feel foolish telling you something you already know. Please accept my apologies. The description might be useful for a student reading this. )

Mark

The problem could also be a ground loop through the I/P and PLC ouput. Measure the current on the + and - sides of the current loop at the plc and see if they are the same. If not, there is leakage to ground or another loop somewhere. Lift the wires at the PLC and check to see if either or both show resistance or stray voltage to ground. I had a similar problem when a junction box got wet and the screws on a terminal strip coroded and developed leakage paths to the backing plate and each other. When we just looked into the box, everything appeared OK. IT WAS NOT! We only found the problem by doing the test above.

It could also be a problem with the valve if it is not shutting off tight. It could be tight packing, mineral build up on the stem, damaged trim or an actuator problem.

Please, let us know what you find.

Have FUN!
TT3

Some kind-hearted advisers giving good and solid tips like:

• The pressure difference of the fluid across the control valve could be higher than the pneumatic supply and output pressure to the control valve's diaphragm.
• PLC output could be out of calibration; zero point could not be at 4mA and 100% could not be at 20mA.
• Controller setting in reset mode, could be giving some offset reading.

First we have to establish the facts:

1. How is the system hooked up in the control loop?
2. Temperature sensor - electric signals of thermocouple / RTD into 4-20mA amplifier, and to PID Controller.
3. PID Controller (Electronic) gives output of 4-20mA to the IP converter, which converts 4mA = 3PSI and 20mA = 15PSI, this pneumatic signal then goes to the control valve's positioner. Inside the positioner, there is a corrective curve, for parabolic curve for globe valve.
4. Positioner, then gives output to the actuator of the control valve and pushes the valve stem against a spring to operate the valve to open or close (depends on the Direct (Increase to increase) or Reverse (Increase to decrease) action.
5. There is a feedback from the valve movement and position back to the positioner.

Could the calibration of the positioner be out? or the positioner is not set correctly at zero when the signal is zero?

Tijit

Tijit,

Those are all good tips as well. Unfortunately the original poster has been strangely absent for more than 2 days now. Maybe he's just away from the computer for the weekend and we'll all find out how it worked out Monday.