Reading output from 2-wire 4-20mA pressure transducers using NI9203 module

[CJTighe]

Hi everyone, I'm sure lots of people can give a straightfoward answer to this.

While reviewing the wiring diagram on some pressure transmitters (2-wire 4-20mA output, direct wire) I'm about to buy I noticed a slight inconsistency with the wiring in the NI9203 user manual. Here are links to the user manuals for the sensors:

Swagelok pressure transducers

Following the NI9203 manual, the +ve terminal in power supply (24Vdc in my case) should be wired to the brown terminal in the transducer and -ve in supply to COM ground on the NI9203. This leaves the green terminal in the transducer to be wired to AI0...AI7 on the NI9203 module. This arrangement makes perfect sense to me. However, the user manual for the transducer suggests the green terminal should be wired to zero volts i.e. the COM ground. This set up makes less sense to me.

My question is, will the internal electronics of the transducer still output the same mA current for the same measured pressure given by the calibration curves for the sensor, even if the green teminal isn't connected to zero volts?

I have a hunch it's okay but I wanted to be sure before I wasted a lot of money/spent years recording the wrong pressure. This should be a question for the manufacturers of the sensors I know, but I've found this forum to be *much* more helpful!

Cheers,

Chris

[Stephen_Moore · Accepted solution]

Hi Chris,

The 9203 is a current sinking device i.e. it is connected on one side to common or 0 volts.  You need to connect your pressure transducer in 2 wire mode.

+24 V DC to the positive terminal on your pressure transducer terminal 1 or brown flying lead, the terminal 2  or green flying lead is connected to AIx on your 9203 and the common of the 9203 connected to 0 volts on your PSU. 

A 2-wire device used with a current sinking input is always wired up as +volts then 2-wire device + then then 2-wire device - (or 0) then AI+ then common then 0 volts.

In my experience the 9203 can be rather prone to noise pick up for 2-wire devices and need an suitable clean power supply.  For 2-wire devices I take 1024 readings and then average them to reduce the noise.

Cheers

Stephen

[CJTighe]

Thanks Stephen, you're a great help. I'm going to to pick your brains a bit more to satisfy my curiousity!

I take it then the electronics in the transducer are smart enough to allow through the right mA current for the right pressure on the calibration curve, regardless of the voltage on the green terminal?

I'm not trying to measure any particularly fast fluctuations so your averaging method is a good work around for the noise. For future reference though, do you find a 3-wire transducer improves the signal to noise ratio much?

Cheers,

Chris

[Stephen_Moore]

Hi Chris,

The 2-wire transmitter limits the "voltage" on the green terminal so that the current that it passes is directly proportional to the item that is being measured.

That is the advantage of 4-20 mA in that there are 1-2 mA at the bottom end to power the transmitter beore the 4-20 mA output signal is reached.  4-20 mA two wire units are one of the industrial choices because they are usually less prone to noise in the industrial environment.  I found that the 9203 samples so fast and as it multiplexes the inputs I was seeing a lot of noise that was not being filtered out.  To compensate for this I sampled 1024 samples from each of the 8 channels of the 9203 over and averaged them for each channel on the FPGA.  This gave me an update speed of 50 ms to the RT which was more than fast enough for my application.  As I am using transmitters on equipment for control, rather than test instrumentation I would still choose 4-20 ma.

Cheers

Stephen

[CJTighe]

Brilliant, thanks Stephen!

Chris