FROM SINGLE DIFFERENTIAL BAROGRAPH TO DOUBLE DIFFERENTIAL : FOM RELATIVE TO ABSOLUTE
The SINGLE DIFFERENTIAL DrDAQ barograph I have described in past postings has several shortcomings, as follows:
1. It is based on a ratiometric system which in turn requires a state of the art, precision 5V power supply having an exceptional long term output voltage stability.
2. This 5 V supply must be independently derived and the DrDAQ 5 V USB power cannot be used because it cannot provide the required precision & long term stability. This in turn means that, even using a battery powered laptop PC, the barograph circuitry ceases to function as soon as the mains voltage disappears, unless a separate PSU backup is provided.
3. The basic barograph system must be temperature compensated to overcome the temperature sensitivity caused by oil density and other changes with temperature. Temperature compensation is relatively simple (linear expression) within an ambient temperature variation of 1 to 2°C, but requires an exponential expression when temperature range increases. NOTE: this may no longer be true with the recently proposed modification.
4. The system requires a relatively large preloading to eliminate the dead band of the 200 mV initial output offset introduced by the pressure sensor.
5. Any change either in the preload oil level, or in the perpendicular position of the sensor assembly affects system accuracy to a high degree.
6. The Sensor assembly must be insulated from sudden environmental changes with a suitable sheath.
In short the described barograph is a RELATIVE system that needs complex technology and construction to provide accurate and reliable data. This is a far cry from the foolproof simplicity of a Torricelli (mercury column) barometer.
It is possible to eliminate ALL the above described problems and obtain an ABSOLUTE system, as absolute as a Torricelli type barometer, by simply building a DOUBLE DIFFERENTIAL barograph. This is easily possible because of the low cost of the Pressure Sensors used.
In a double differential system TWO differential sensors are used. SENSOR #1 measures barometric pressure variations as before (e.g. Port 1 preloaded, Port 2 sealed) while SENSOR #2, connected to the same preload oil column, has Port 1 preloaded and Port 2 open. In other words Sensor #1 monitors Barometric Pressure, while Sensor #2 automatically corrects any (parasitic) change in Sensor #1 output which does not depend on a change in Barometric Pressure: this feature because Sensor #2 cannot see Barometric Pressure variations, which are the same at both ports. The two sensor outputs go to two channels of the DrDAQ and hence to a math channel which simply subtracts Output Sensor #2 from Output Sensor #1. By this arrangement:
a) Ratiometric Power Supply variations disappear and the DrDAQ 5V USB supply can be used, because the ratiometric error in Sensor #1 is cancelled by the identical error in Sensor #2.
b) Any oil density variation and/or level, disappears as well, because both sensors share the same oil column weight. This feature eliminates the need for temperature compensation.
c) The deadband of 200 mV is also eliminated, allowing a notable reduction in preload column size.
d) Any output variation due to ambient temperature changes is automatically eliminated, therefore no compensation is needed and the protection sheath is no longer necessary.
In effect the output of the Picolog math channel is a voltage directly proportional to the barometric pressure and nothing else, the only hypothesis being that the two sensors be well matched. But small differences in the two sensors can be easily compensated by a calibration oil column connected to Port 2 of Sensor #2.
The double differential system is already built and is undergoing evaluation testing since one week. All data & info will be provided in future posts.