Welcome to the world of differential pressure measurement. All
differential pressure transducers exhibit a zero shift as the line pressure
changes. This is usually due to the physical deformation of the
transducer's case as the line pressure changes. There are at least six
ways to handle this. I will list them in order of least to most capable
(my opinion). You will have to decide the exact desireability and utility
for your application. The first four only require one transducer. The
other two require two transducers.
1. The simplest way is to get a better transducer that has less "line
pressure effect." The error due to line pressure should be given in the
transducer's data sheets. The units are usually something like "percent of
full scale error per 1000 psi of line pressure." In good transducers, this
is on the order of 1 to 3 %FS/1000 psi.
2. The most popular, but the most limited technique is to do a single
calibration at a single line pressure. This works well for a process
plant. The plant tries to run at nearly constant conditions. If the
transducer is calibrated to operate at the system's design line pressure,
(which is where you operate 99% of the time) it does not matter if there is
some error at off design conditions.
3. Another way to handle is to use a set of solenoid valves to re-zero the
transducer at line pressure. Two single solenoid valves or one three way
valve is necessary. The idea is to disconnect one side of the transducer
from the line and then to short the high and low pressure sides of the
transducer together. Use this new reading as your zero and go to town. I
have used this technique sucessfully. It was not as hard to implement or
as clumsy to use as you might expect. It did not require extra
transducers.
4. The last single transducer method is to put your transducer inside a
pressure vessel. The vessel is pressurized to the low side of the
transducer. The high side is connected to a port in the side of the
vessel. The suitability of this depends on your
pressures/temperatures/fluids. This works because the line pressure effect
is primarily the indication of the transducer's response to gage pressure.
The pressure vessel does not need to be anything special. I have used a
piece of two inch pipe with some pipe fittings on each end. You need a
high pressure port, a low pressure port, and a place to pass the wires.
There are a variety of bulkhead wire passthroughs available. These are
expensive for high pressures. For low pressures, drill hole in the vessel
and epoxy the wires in place. Since this is a pressure vessel, make sure
you know what you are doing so that it does not blow up in your face. A
pressure vessel method like this can measure DP with a gage pressure
transducer instead of a DP transducer. Gage pressure transducers are
generally more accurate, less expensive, and more widely available than DP
ones.
You may need to experiment to find the best transducer for size, accuracy,
etc. The advantage of this method is that it can give the accuracy of a
three dimensional calibration of a two transducer system with a single
transducer and no special software.
5. This is the technique that you hinted about. Use two transducers and
a zero offset calibration. One transducer measures the line pressure (gage
or absolute, up or downstream pressure--it does not matter). The other
transducer measures the DP. Do a zero offset calibration of the DP
transducer by applying the same pressure to both sides and recording the
output. When your system is in operation (not calibration), measure the
line pressure and subtract the calibrated zero offset from the differential
pressure. This procedure works well because the line pressure effect is
usually an offset rather than a slope change.
6. The most comprehensive two transducer method is a full three
dimensional calibration with line pressure, differential pressure, and
output. In this case, the calibration curve becomes a calibration surface.
The software effort with this system is much more complicated.
If any of these are not quite clear or if you need any help, let me know.
Paul Brown
"John Brohan" <[EMAIL PROTECTED]>@pica.army.mil> on 02/19/2004 05:56:07
AM
Sent by: <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
cc:
Subject: Calibrating a differential gauge
Hello
We have a differential pressure gauge which shows a different zero
depending
on the absolute pressure. The curve of offset to pressure is a parabola
over
the area of interest. How do people normally deal with this issue? Is it a
bad gauge or do we just incorporate an offset calculation into the
conversion routine?
Yours Sincerely
John
John Brohan National Instruments LabVIEW expert in Montreal
Traders Micro "We connect all sorts of things to computers"
317 Barberry Place DDO Montreal PQ Canada H9G 1V3 Tel (514)995-3749
[EMAIL PROTECTED] http://www.TradersMicro.com/Index.htm
