Pressure altitude from barometric sensors

Stuff I am working on
Post Reply
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Pressure altitude from barometric sensors

Post by Daniel Wee »

The formula for calculating pressure altitude in meters, from atmospheric pressure in kPa is given as:-

altitude = (1 - (P/101.325)^0.190263) * 44330.32172

So, a pressure of 1010 mbar, which is also written as 101.0 kPa works out as:-

(1 - (101/101.325)^0.190263) * 44330.32172 = 27.09m (rounded to 2-decimal places)

So, taking an SCP1000 sensor, for example, which has a pressure resolution of 1.5Pa in high-resolution mode, and 3Pa in standard mode, we can see that this works out an altitude resolution of 0.125m for the high-resolution mode and 0.25m for standard mode. Note that high-resolution mode only allows for 1.8Hz sampling rate with this processor and about 9Hz in standard mode. High-resolution mode for the SCP1000 uses 17-bits, while standard mode uses 15-bits.

What this tells us, basically, is that in order to get any kind of decent readings out of barometric pressure sensors, we should be looking at really high resolution ADCs, of the order of 14-bits and higher.

Now, if we take a sensor like the ADSX015A24R which give a pressure range of 0-15psi (0-103.42kPa) and feed this into a 12-bit ADC, we end up with a pressure resolution of:-

103.42kPa/4096 = 25.2Pa

which is nearly an order of a magnitude greater than what we were getting from the SCP1000. This is equivalent to a 2.1m altitude resolution, which is, quite frankly useless. There are several ways around this - either increase the dynamic range by raising the floor off 0kPa to a window that will still work, or use a higher resolution ADC.

Let us look at a cheaper sensor - the MPX4115A. This puts out 4.8V at 115kPa. Assuming a 5V full-scale ADC, this works out to:-

115kPa/4.8*5/ 4096 = 29.2Pa (2.44m) for a 12-bit ADC, and
115kPa/4.8*5/ 16384 = 7.3Pa (0.61m) for a 14-bit ADC, and
115kPa/4.8*5/ 65536 = 1.8Pa (0.15m) for a 16-bit ADC

Clearly then, our options for barometric pressure sensing lies with either sensors with integrated high-resolution ADCs or with external high-resolution ADCs such as the Microchip and TI parts - MCP3425A0T and ADS1100A1IDBVT. These are relatively affordable parts and offer 16-bit resolution on an I2C interface in an SOT-23-6 package (ADS1112 for 2-channels).

Daniel
Attachments
Motorola-MPX4115A.pdf
(735.11 KiB) Downloaded 787 times
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Pressure altitude from barometric sensors

Post by Daniel Wee »

A common type of pressure sensors come in the wheatstone bridge strain configuration. This might be cheaper and could be made to work at 3.3V (most of the above run at 5V). This could be coupled with some of the 16-bit ADCs we have been looking at for a reasonably cheap solution. The following link provides a good treatment of the Wheatstone bridge calculations involved:-

http://www.automotivedesignline.com/howto/202402040

What we need is to establish the dynamic range when used in this sort of configuration. There are two further problems when using Wheatstone bridges - the first being that the voltage output differential is normally very very small, thus requiring further amplification and signal conditioning using op-amps and such, thus increasing the complexity of the sensor design. The second is the need for a constant current voltage supply across the bridge - again complicating the design.

Daniel
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Pressure altitude from barometric sensors

Post by Daniel Wee »

There seems to be another sensor that we missed, the MPXV7002DP from Freescale. This actually reports a range of -2kPa to 2kPa but with the differential ports, could be the perfect thing for TAS sensing. Outputs analog so should be easy to use.

Daniel
Attachments
Freescale-MPXV7002 Airspeed.pdf
(215.29 KiB) Downloaded 902 times
Post Reply