Which is better: Using NI 9203 OR USB Volt input differential card

[abvcf]

Dear All,

I have 4 pressure transmitters, 4 temperature transmitters with output of 0-10 Volt or 4-20 mA. Also I have 2 flow meters with output of 4-20 mA.

I have two options:

Either to get all the signals in mA and connect them to two C modules NI 9203 OR I can get 8 sensors with 0-10 V and then I can use for example NI 6218 where I can connect these signals using the differential configuration. While the last two signals (of the two flow meters) are to be converted using SSC components to yield Volt output so that I can connect them to the same card.

Which solution is better regarding the signal noise ?

Some one told me that NI 9203 does not support differential configuration as the mA signal does not obtain noise like the Volt signal so it does not need differential configuration.

Also NI 9203 has 0-20 mA input range and the sensors output is 4-20 mA ......... Is there a problem ?

Please advise .......

Yours,

Ehab 

[John_P1 · Accepted solution]

Hi Ehab,

Do you mean that you have the option between voltage and current transducers, and would like to know which is best?

A 4-20 mA current loop is inherently less sensitive to voltage noise.  4-20 mA sources have a much higher output impedance than a shunt resistor, so the vast majority of the noise is dropped across the source.  The following application note from BAPI and our own Developer Zone article both give a good overview of the subject:

Understanding 4-20 mA Current Loops

Current Measurements: How-To Guide

As far as the 0-20 mA vs. 4-20 mA, you won't have any problems here.  One reason 4-20 mA is used commonly for sensors is that a 0 can be used to indicate a bad (open) connection.

When taking a voltage measurement, the best way to reduce noise is to measure differentially, with your connections being made with shielded-twisted pair wires (wikipedia link).  The idea here is that any atmospheric noise would affect both the + and - ends of your signal.  Since you are measuring differentially, the noise would cancel out.

You will likely be fine in either case.  If you use the 6218 you can provide your own external shunt resistors to measure current, while using the other channels to measure voltage. See the article How to Build a Current Loop System with Isolated M Series.  It appears to not be updated for the 6218, but this board provides the isolation recommended for measuring current.

The 9203s have this built-in to the board and are meant specifically for measuring current.

Other recommended reading would be Field Wiring and Noise Considerations for Analog Signals.  When setting up this system you'll need to make sure everything is grounded properly (no ground-loops, floating connections, etc.).

-John

[abvcf]

Dear Sir,

Thank you for your reply.

What I understood now is:

It is better to use 2 NI 9203 modules for the sensors signals (either 0-20 mA or 4-20 mA) and these signals will be connected using single ended configuration.

This solution will be better regarding the noise issue.

Am I right ??

Yours,

Ehab

[John_P1]

Hi Ehab,

You could use either the 9203 or the 6218 to measure current.  The 6218 would require external shunt resistors.

In general, the 4-20 mA current loop will be more resistant to a noise (voltage is dropped across the resistor of the current source). 

From the 9203 Operating Instructions and Specifications, the signals would be connected as follows on the 9203:

From the link in my previous post, the 6218 would be connected as so (with an external Shunt Resistor--the shunt is the resistor across + and - channels):

Both boards would be viable options--I would recommend taking a look at the links I provided earlier for more insight about noise and how to connect your signals.  Thanks for posting!

-John

[abvcf]

Dear John,

Thank you for your reply.

There is something I want to make sure of it.

NI 9203 does not support differential configuration because single ended is sufficient for current inputs not like voltage inputs which need to be connected using differential configuration to eliminate noise.

Am I right ?

Thanks for your help .......

Yours,

Ehab

[John_P1]

Hi Ehab,

In a sense, yes this is true.  A possible benefit of using a differential input for current would be the ability to provide different references for your different channels (provided they are within the common-mode voltage range of the device).  

When using a differential board, you would often tie the (-) channels to the same ground to prevent the common-mode voltage from drifting outside of the range of the device.  The 9203 connects its shunt resistors internally to COM, so all of the measurements will share a common reference.  Here's another diagram from the 9203 Specifications that shows the shunt resistor configuration of the 9203:

Voltage measurements often benefit from differential configuration in the area of noise reduction, but current measurements do not work in the same way.  With current measurements, we are measuring the voltage drop across a shunt resistor that is internally (on the 9203) connected to a differential instrumentation amplifier.

-John

[abvcf]

Dear John,

Thank you for your reply.

I have another question:

Now the sensor output will be 4-20 mA and the NI 9203 input range is 0-20 mA. If the sensor gives 4 mA (zero reading), the 9203 will read 4 mA.

In this case I will use the lab view to modify this 4 mA to zero mA, so that I am getting correct readings.

Please comment ......

I am sorry for any inconvenience ........

Yours,

Ehab

[John_P1]

Hi Ehab,

You may indeed measure 0-20 mA on the 9203.  You will need to apply scaling in software to map the 4-20 mA signals to a physical phenomenon (e.g. temperature).  The sensors that you will be using should give you the scaling coefficients--they are usually a function of excitation voltage.  You wouldn't map 4 mA to 0 mA, but rather to 0 degrees Celcius (for example).

-John

[abvcf]

Dear John,

Thank you for your reply.

Will this scaling process increase the error of our readings ? specially if the calibration curve of the sensor is not linear.

Is there any other way for the NI 9203 to be compatible with the 4-20 mA directly.

Thanks for your continuous help .......

Yours,

Ehab

[John_P1]

Hi Ehab,

I'm not sure I understand.  The 9203 measures current signals within the specified accuracy tolerance--scaling this value in software would not increase the error of the current measurement.  A device that can measure 0-20 mA is inherently "compatible" with 4-20 mA.  You would just not expect to measure any values below 4 mA unless the signal was connected improperly.

For help scaling the measurements, you could look at the following resources:

How do I make a custom scale in the Measurement & Automation Explorer (MAX)?

Communities: DAQmx Acquire with Custom Scale

-John