4-20 pressure sensor

[Jaime Rodriguez]

I am working with a 4-20 mA output pressure sensor for a 0 - 1000 psi range. And I am using the CFP-AI-110 for the Compact field point to acquire this signal which is 16 bit. Does this seem to be too big of a range (resolution) to work with for precise measurements???? Were talking that 1 psi = .016 mA.

Should I do something to this signal to minimize the error???

[upchuckjunk]

Generally speaking, I like to size my transducers so that my maximum possible measurement is around 70-80% of the transducer full scale range.  This gives a reasonable over and under-range cushion.  Others have told me that 50% of full scale is the target range.  But in either case, it sounds like your 1000 psi full scale is a bit large and you should use a different transducer if your normal readings will be in the 1-5 psi range.  That's my 2 cents.  

[hrh1818]

On May 19, 5:40 pm, Jaime Rodriguez <x...@no.email> wrote:
> I am working with a 4-20 mA output pressure sensor for a 0 - 1000 psi range. And&nbsp;I am using the CFP-AI-110 for the Compact field point to acquire this signal which is 16 bit. Does this seem to be too&nbsp;big of a range (resolution) to work with for precise measurements???? Were talking that 1 psi = .016 mA.
> &nbsp;
> Should I do something to this signal to&nbsp;minimize the error???
> &nbsp;&nbsp;&nbsp;

The short answer is no. Your CFP-A1-110 has a software configurable 4
t0 20 mA input range. Hence you do not need to do anything to the 4
to 20 mA pressure signal to make it compatible with your CFP-A!-110.
You don't mention what the accuracy of your pressure transducer is. I
suspect that might be the biggest limitation in the accuracy of your
measurement.

Howard

[Jaime Rodriguez]

I understand what you guys mention, but my concern is having measurements being "unstable" due to very small variations in mA which in turn (because of the conversion) happen to be a significant amount in psi.

This is, I can be reading 345.876 psi and be seeing the measurement be unstable by oscillating between 342.xxx and 348.xxxx (for example), which in terms of mA, is a very very small amount of variation but due to the conversion of 4 - 20 mA to 0 - 1000 psi is a significant amount. Remember 1 psi = .016 mA, so everytime the fieldpoint reading changes .016 mA, (which is a VERYYYY small amount), my psi reading will change by 1.

Just would like to know if the Compact field point would be precise enough to work with this kind of resolution. 

[GerdW]

Hi Jaime,

have you checked the accuracy of your pressure transducer? Better than 0.1% of full scale? That's like using the LSB of a 10bit ADC to get "precise measurements"...

[Jaime Rodriguez]

The specs on the pressure transducer say the following:

High Accuracy --- 0.25% Typical

What are your thoughts??

[GerdW]

Hi Jamie,

ok you got: "0 - 1000 psi range" and "Accuracy --- 0.25% Typical". This gives you an accuracy of 2.5psi (typical. may be even worse than that) just from that transducer. And you still ask for a 0.1% conversion accuracy???

But anyway: you wrote the CFP module works with 16 bit ADC. What about any specs for accuracy here? Typically something like "0.05% +- 4 LSB" is given here... Check this too!

[hrh1818]

On May 19, 11:10 pm, Jaime Rodriguez <x...@no.email> wrote:
> I understand what you guys mention, but my concern is having measurements being "unstable" due to very small variations in mA which in turn (because of the conversion) happen to be a significant amount in psi.
> This is, I can be reading 345.876 psi and be seeing the measurement be unstable by oscillating between 342.xxx and 348.xxxx (for example), which in terms of mA, is a very very small amount of variation but due to the conversion of 4 - 20 mA to 0 - 1000 psi is a significant amount. Remember 1 psi = .016 mA, so everytime the fieldpoint reading changes .016 mA, (which is a VERYYYY small amount), my psi reading will change by 1.
> Just would like to know if the Compact field point would be precise enough to work with this kind of resolution.&nbsp;

The specifications for the CFP-A1-110 says it has a resolution of 16
bits. A resolution of 16 bits is equivalent to 1 part in 65536. For a
4 to 2o mA range this is 16/65536 or 0.000244 Ma which is far smaller
than the 0.016 mA you mention in your first message. National
Instruments doesn't say what the accuracy of the CFP-A1-110 is but in
order to obtain a resolution of 16 bits the noise and DC drift of the
CFP-A1-110 referred to its input must be less than 1 part in 65536.
You should also notice that National Instruments sells Compact Field
Point modules for use with Thermocouples. The signal levels from
Thermocouples are much lower than they are for a 4 to 20 mA DC
signals. Yet they still claim the same 16 bit resolution. Hence if
National Instruments can meet a 16 bit resolution specification with
Thermocouples they certainly can meet a 16 bit resolution
specification with a 4 to 20 mA DC signal.

As I suspected your pressure transducer is the limiting factor on the
accuracy of you measurement. A accuracy of 0.25% is one part in 400 or
0.04 mA

Howard.

[Wayne.C]

There is one very important factor affects the resolution of an AI-110 module.  If you carefully read the fine print in the Operating Instructions, it states that if a channel is configured to use the 50 or 60 Hz filter, then the resolution is 16 bits.  BUT, if the filtering is set to 500, then the resolution is only 10 bits.

I currently use AI-110 to read 1000 psi transducer w/4-20mA output with no problems.  One thing to remember is that a 4-20 mA transducer does not always have an offset of exactly 4.000 mA and a span of 16.000 mA.  You should have received a cal sheet that called out the exact values to use.  I would recomend averaging several readings as well.

Message Edited by centerbolt on 05-20-2008 01:01 PM