Yes, that is normal.
The output is taken from the difference of the two outputs.
Look at the schematic for that project. There is an instrumentation amplifier between the two outputs of the Ptx and microprocessor.
Read the datasheet, the full scale output of that Ptx is only 40mV.
The 10 volt supply needs to be connected to pins 1 and 3. The output voltage is taken between pins 2 and 4.
The MPX does not need to have a load resistance connected between pins 2 and 4.
The output voltage is measured BETWEEN the pins 2 and 4 and these should not be connected to ground.
The full scale output is 40 millivolt and this occurs at a pressure of about 16 feet (5 meter) of water.
Your sensor has a range of 0 to 50 kPa. The full scale span is 40 mV. The sensitivity is 0.8 mV/kPa. Therefore 40 mV/ 0.8 mV = 50 kPa, so 0 to 50 kPa is the pressure range of the sensor.
1.0 centimeter of water column = 98.06649999980075 pascal so 30 centimeters of water column = 2941.9949999940227 or about 2.942 kPa. Since the sensor sensitivity is 0.8 mV/kPa and the pressure for 30 centimeters of water column will be 2.942 kPa we can multiply 2.942 kPa * 0.8 mV = 2.3536 mV.
Since the maximum pressure of the sensor is 50 kPa we can say 50,000 / 98.066 = 509.86 or about 510 centimeters of water column. That would be the span of your sensor expressed in centimeters of water column. I am basing all of this on Sensitivity (ΔV/ΔP) taken from **broken link removed**which covers theMPX2050 series 0 to 50 kPa (0 to 7.25 psi) 40 mV Full Scale pressure sensors.
Keep n mind with 30 centimeters of water in the tank you are pretty low in the sensor range of about 510 centimeters.
The LM324 operational amplifiers are configured as an Instrumentation Amplifier to amplify the low differential output of the sensor to something the uC can use and read.
Going bck to your post #4 and without ignoring the other contributions, it seems to me that you might choose a different pressure sensor unit. You have a 50 kPa model and are using it to measure only about 3 kPa. You might be better with say a 2010 model which measures up to 10 kPa so your normal working range is 1/3 of the full scale value.
If you have the 2050 and want to use it, and if you need an amplifier, I would suggest that a very low drift design would be needed. With the 2050, your output voltage is only a few milli volt and so an amplifier drift of a couple of millivolt will give you significant reading errors.
Yes, you can use that circuit with changes to the gain but using the LM324 configured as an Instrumentation Amplifier is really a poor choice. When doing a circuit like this all of the resistors should be 1% precision film resistors or better. They need to be matched perfectly or you will have all sorts of gain and offset issues as well as temperature coefficient problems. You would be asking for problems. The LM324 is a pretty old chip and never intended for IA circuits. Personally I wouldn't do it.
A much better choice that will eliminate the problems I mentioned would be a chip like the INA128 which is an instrumentation amplifier designed for exactly what you want to do. All you need to do is add a single gain resistor and the use of a good quality 1K Ohm pot will get you there. This will make for easy and accurate calibration using a 1K 20 turn pot. The INA128 is very, very common and very inexpensive averaging between $9 and $11 USD depending on flavor.
Based on your application the MPX2010DP is a better choice as long as it affords the range you want.
Mounting the pressure sensor like that is a really bad idea.
You already have a very small depth of water (30cm max) to give pressure head.
Using this method, the pressure will be (30cm - water column in the tube).
The sensitivity will be greatly reduced.
As the temperature changes, the volume of air in the tube will change, this will change the height of the water column in the tube and hence the indicated pressure.
To use this sort of measurement technique you need to feed air into the top of the tube to displace the liquid from the tube.
That way you get a true level measurement.
Try googling for "bubbler tube level measurement" for mor information.
Otherwise, you need to mount the pressure sensor so that it is below the bottom of the tank and in direct contact with the liquid.
Mounting the pressure sensor like that is a really bad idea.
As the temperature changes, the volume of air in the tube will change, this will change the height of the water column in the tube and hence the indicated pressure.
Otherwise, you need to mount the pressure sensor so that it is below the bottom of the tank and in direct contact with the liquid.
I used MPX2010DP and INA114 amplifier. I got very nice results.
In the first time I used +12V -12V as supply for the INA114 with RG=100 ohm, so I had a voltage from 0 to 2.5V to my microcontroller. But when I changed the supply to +14V -14V, I have now 0 to 27V !!!!!
I used MPX2010DP and INA114 amplifier. I got very nice results.
In the first time I used +12V -12V as supply for the INA114 with RG=100 ohm, so I had a voltage from 0 to 2.5V to my microcontroller. But when I changed the supply to +14V -14V, I have now 0 to 27V !!!!!
Something went wrong. With Rg = 100 Ohms the gain should be about 500. The supply voltage can be as high as +/- 18 volts. Have you looked at the data sheet?
Yes I looked it... I solved this problem. But now When the level is 0 cm the output of the INA114 is -180mv although the offset pin (5) is in the GND !!! What can I do now ?
Following the datasheet of the microcontroller STM32F10 and Fio std board, the input signal must be between 0 and 3.3V. It is normal that is -163mV ??
No the negative 163 mV is not normal. I am wondering if the INA was maybe damaged? With the offset pin at ground and zero volts input the output should be zero volts. That -163 mV is quite a bit. Could the INA have been damaged with whatever happened earlier?