ECN Forum Forums General Topics of the Trade General Discussion Area Why is 4-20ma used..

0-20mA isn't used because you wouldn't be able to discern between zero output and cut wires. 4-20mA is a good method to provide supervised analogs over long distances in a noisy environment. The wiring resistance doesn't even have to be constant as long as the full 20mA can flow through the loop within the source's supply capabilities.

Let's say that I have a 0-60PSI ---> 4-20mA pressure sender but no cut sheet. I do know that it works off of a 24VDC supply. If I hook this thing up to my bench supply with no load,(output leads open), there will probably be close to 24 volts on those leads. Let's say that I measure 23.0 volts with a 24.0 volt supply. 23V / .020A = 1150 ohms. This is the highest load (loop) resistance that would allow a full scale output current. It would be 1000 ohms max. with a 20 volt supply.

My problem is not loop resistance. My issue is that my 2 SCADA (Supervisory Control And Data Acquisition) analog input cards use 0-1mA with 1000 and 5000 ohm load resistances. 1mA will develop either 1V or 5V across each of my analog input points with a full-scale output. So in my case, I need to shunt 19mA, at either 1 or 5 volts, around my input. 1V / .019 A = 52.6+ ohms. 5V / .019 A = 263+ ohms. Now the transducer outputs 4-20mA, but my SCADA sees .2-1mA through its input resistance. Master computer scaling and offset numbers are programmed in on the other end.

Finally, if I were crazy enough to want to do it, I could probably series my 4-20mA sensor output through several different analog inputs, and get accurate readings on every one. But again, the cumulative full-scale current voltage drops plus drop across the wiring, can't exceed the available supply. This strange connection would also require isolation on all of the inputs. This has never been a problem for me since the more I talk theory, the more isolated i seem to end up.
Joe

[This message has been edited by JoeTestingEngr (edited 02-07-2007).]

Joe,

In the old days it was very common to run multiple devices off of a single 4-20mA circuit. I have done sensor to chart recorder to digital display to loop controller. The only typical problem is if mutiple devices have one side of their signal tied to "ground".

rad74ss,
0-20mA is very common as the analog output of new generations of VFDs and "small" PLCs.

JBD,
Yes, we could have all matter of fun finding that not-quite-perfectly-isolated input, couldn't we. I should clarify something with the 0-20mA. I would never design around a 0-20mA output UNLESS my normal, non-alarm, operating point was well above zero. For instance, I wouldn't mind using a 0-100 degree C ---> 0-20mA transducer to monitor a process that ran between 20 and 100 degrees C. Any current < 4mA would be in the alarm range anyway. My programming would be very easy because I would be dividing a real world full-scale value of 100 degrees, by my full-scale A-D count of 2047 to come up with a scaling factor. There would be no offsets to program in. With 4-20mA, you end up dividing your real world range by 80 percent of the maximum count number, and using an offset value to slide back down to a zero reference. It gets a little confusing at times though. I have one RTU mfg. where I have to offset by 2047 X scaling factor, just to get my zero, Adding 4-20 to that mix is not something I can handle before my second cup of coffee.
Joe

4-20mADC process variable transmitters have the advantage over other types of signal transmitters that they are "2-wire" instruments. They get their power from the same two wires used for the 4-20mADC signal. In otherwords, they do not require a separate power input. This makes them convenient.

As mentioned by others, the current analog signal is less susceptible to interference. However, the only signal available is the process variable. The newer digital type transmitters can send a lot more information, such as, the process varible say pressure, the temperature, the tag number of the transmitter, the transmitter's health, etc.

How does HART factor in to 4-20mA control? Does it run on top of the loop or is it a seperate interface?