Minimum and Maximum Limits Analog Output Current (4-20mA) Deviation

This sounds more likely that you have an alternate current leakage path that is robbing your pickup electronics a fraction of the current that is produced by the sensor. You may also just be having part of your analog circuitry sliding into some voltage compression at some point. If this current loss is consistent and predictable then you maybe able to just ignore this difference.

Now the discussion on signal conditioning gets into a host of other concerns. How much AC noise is present and can be tolerated? What is the bandwidth of the data acquisition? Do you require therefore a Nyquist filter network as part of the signal conditioning. Differential or single ended cabling.

You also have to consider the possibility that the sensor precision may not be absolutely precise. What are the tolerances specified by the manufacturer of the sensor?

Hi Guest:

A 4-20mA device will rarely give you exactly that, there is always a few uAmps offset, but don't worry, this drift is a fix ammount and won't be changing from one measuring to another.

I assume you are using single ended connection, since your signal is not supposed to cross zero at any point thrughout the full span. If something else were robbing current from your sensor, I would expect a current increase rather than a decrease.; unless, of course, there is an external circutry that diverts a few uAmps from the analog input chanel. I wouldn't worry about high input impedance, since your device is capable of delivering both ends of the span. If you have concerns regarding the connections, I would advice to :

1.-Get rid of unnecesary conections.

2.-Eventhough the 4-20mA is has advantages over, say 0-10V as to the reach and noise immunity, run the signal wires on a separate conduit and use shielded cable.

3.- Ground the shield at ONE end (the input device end).

4.-Do good programming; if you are using a PLC, there are instructions that perform ALL signal processing for you; make the lower limit equal to a, the max limit equal to b, you can scale the span in-between to the proportions of your choice, and attain the linearity required for your app. You can program offsets and have an output in the user units of your choice.

Yahlasit

1) Is the 4-20mA output loop powered with an external DC power supply?

Failure to attain 20ma is symptomatic of insufficient supply voltage, particularly in loop powered outputs.

Can the 4-20mA output drive the analog input resistance(s)? How many? What resistance?

2) Are there zero and span adjustments on the device?

Are the low-output values given us the values after a zero and span adjustment have maxed out?

Sounds to me like it needs to be zeroed.

Thank you guys.

There is no any connection at the 4-20mA output but only the hard wire through the PLC. My concern here is to employ a signal conditioner along 4-20mA line through the controller input. Can you suggest me what particular signal conditioner has good quality and how much? thank very much for your ideas and suggestions.

Hi,

As it was suggested one option is to program the PLC to the zero and span values or outputs that you have for zero and span. In your case they are not 4mA and 20mA.Check moore industries for a signal conditioner,splitter or isolator, they can be calibrated to give you outputs of exactly 4-20mA.

Thank you

www.datexel.com has a DAT4135 which will help you. It can remove the offset and also provide isolation. It will also boost the output so you can drive into a impedance of 1000Ohms.

You need to ask a couple of questions before moving towards a signal conditioner. 1. At 3.7ma, does the display read negative or zero in weight when the load cell is empty. (if negative, your issue is with programming on the PLC) 2. At 19.5ma, does the display read exactly full load, or less than full load. (If reading full load, programming is the issue again) 3. Signal condition, noise reduction (separate conduit-grounding) is the issue if readings become erratic or cable lengths exceed 2500 feet. 4. To determine if strain gauge is the issue and requires span adjustments, apply a separate analog meter and power supply and measure ma at empty and full load. You did not provide enough to isolate root cause, so suggest your stay with basic approach before looking at resolutions which may snowball you into further confusion. Yahlasit is closes to solving your problem.

I read through the replies and nobody seemed to get to the heart of the issue.

1) A 4-20 mA loop does not output exactly 4 and 20 mA for zero and full load respectively unless it is precisely calibrated to do so. This calibration requires several iterations between zero and full load with proper adjustments of the zero and span potentiometers. Once precisely calibrated, temperature drift may make actual values deviate a bit form 4 and 20 mA respectively.

The 4-20 mA concept is excellent for process control where you are measuring variable(s) with 4-20 mA loop(s) and controlling process outputs with 4-20 mA loops.

If you need precise weight control, you may have to retrofit the installation with load cells, summing junction boxes and weight controller(s). Two companies that make such equipment are Hardy Instruments and BLH. Both of them are extremely accurate and both come with fairly hefty price tags. The Hardy solution even has interface modules that plug directly into an Allen Bradly PLC chassis for directly inputting the weight signals into a PLC. These are available for the PLC5, SLC500 and Control Logix families of PLCS. Hardy also has an automatic calibration mode available with their later model load cells and instrumentation 'C2' that allows accurate calibration without using test weights.

Old saying: You rarely get more than you pay for, and precise weight measurement is no exception.

Greg

P.S. Where I work there are dozens of Hardy and BLH systems in use for weight measurement of ingredients and products in the processes of the plant.

The upper and lower range limits also the zero and span adjustments are clearly outlined on the following link. http://www.kistlermorse.com/pdf/FAQ_SVS_reva.pdf Regarding the interaction of loads, it is not possible without seeing structural drawings to determine if interaction is likely however if it does occur you could set up trials to determine the interaction then profile to signal to compensate for errors. One point to note, the product may have different weight in different seasons. I don't know where your raw material is being stored before it enters the silos but if you set the system up for 20mA when you are full of dry product you will be over scale on wet product so it may need some consideration. Deano

The maximum load should be set little below 20mA to allow for overload. If max reading is 400V it should be set around 18mA to allow a certain amount of overload if voltage exceeds 400V

"Micro Cell" is a strain sensor which measures the deformation of the element on which bit was fastened. In combination with the carrying element it builds up a "load cell". This is for clarification of the component we discuss about.

Now when assembled the "strain sensor" is not always at "zero", its zero is in an "unmounted position" free of all constrains. Some of those components have a mechanical zero-ing device to make the zero adjustment on the spot other have a trimmer potentiometer in the box to make the same on the conditioner electric side.

Since the strain is function of the support the weighing MUST be calibrated ON THE SPOT since every support has an other stiffness thus an other sensitivity. This cannot be pre-set in the factory it MUST be set on the spot after assembly.

For silos this calibration of the sensitivity can be done either with additional weights or with a filling of the silo with a quantity weighed on a precise scale. There are -normally in the conditioner box adjusting possibilities for the gain of the strain conditioner so that the output signal at the end display corresponds to the loaded weight.

Procedures are as for load-cells only one has to remember that in this case the"load-cell" is the combination of the support with the strain sensor.

Hope it will help to better understand the problem.