Hello,
I am trying to implement capacitive sensor using Arduinio. It is a simple RC circuit. It works as follows,
Concept
--> Self capacitance of a single plate capacitor
Only send pin and receive pins are used (Read from literature and applied)
Problem is most of the literature and videos show sensor working on ADC principle. Send pin sends a high signal and on the receive side time is counted till receive goes high as well. I want to get analog signal (Voltage) from the sensor as i have to use it as trigger for other data acquistion systems. I give 5 volts from send pin and at receive pin i measure analog voltage. It is not grounded. Big question is with respect to which reference voltage is measured here. Sensor works fine and shows voltage drop when touched.
I am beginer in electronics.
Any lead is welcome.
At start the capa is empty (U=0). When you connect the source a current will flow through the resistor and fill the connected plate thus the voltage difference between the plates (U) increases. The current is proportional to the voltage difference across the resistor and will decrease due to "filling" of the capa plate. The voltage of the plate is proportional to the integral current over time and you obtain a [1-exp(-t/RC)] evolution.
What you should measure is the voltage difference between the plates which grows from zero to the supply voltage.
Important is the impedance of your pick up since its resistance leads a current leak which will modify the measurement from its theoretical evolution.
In fact the current which will be integrated is reduced by this leak. The equation is U= integral(((Usupply-U)/R-(U-Uref)/Rtestpoint)/C*dt). Before every test you should short circuit the capa to put it at voltage difference =0. Uref is the voltage at the other end of your test point.
Hello Nick,
Many thanks for your answer. Please have a look at attached simple circuit diagram. In the circuit arduino send pin and receive pin are not connected to ground anywhere. In this case voltage read at receive pin is with respect to ground or any other reference.
Secondly, many literatures suggest it might be a floating pin as well. How can differentiate between floating pin and capacitance.
Naveen Shirur The output pin is indirectly referenced to GND because when switched to logical "1" the output driver connects the pin to the positive supply rail, and the power supply is connected to this rail and to GND.
The input pin is directly referenced to GND because the analog inputs of the ATmega328P are single ended. This means that the input voltage is always compared to the voltage between GND and the positive supply rail (or an internal 1.1 V voltage or the AREF pin); input potential = GND results in a "0" from the ADC, input potential = potential of positive supply rail results in the maximum number of the ADC = 0x3FF (10 bit resolution).
So, even if there is no connection to GND outside the microcontroller, all potentials are referenced to GND inside.
Herr Joerg Fricke, Many thanks for the answer. It gave me much clear idea since I am new for electronics. Following same question, connecting a multimeter to sensing pin of the capacitor shows different voltage than serial monitor display voltage. Is it because of capacitance induced by multimeter pin? If yes how can i read exact voltage using multimeter?
Hello, Herr Naveen Shirur , using a multimeter, you should measure the voltage in a static mode, i. e. with the digital output fixed at digital "0" or "1" because usual multimeters provide just 2 to 3 measurements per second.
In the static case, the input capacitance of the multimeter has no effect. The terminal "COM" of the multimeter should be connected to GND. The input resistance Ri of the multimeter, together with your resistor R, forms a voltage divider, resulting in a voltage Vin at the analog input: Vin = Vout * Ri / (R + Ri). Ri is mostly in the range 1 .. 10 Mohm, so with R also in the MOhm range, the input voltage will be considerably lower during measurement. But since the ADC sees the same voltage, both the ADC result and the multimeter result should be equal within the precision limit.
If they are not, my first suspect were the equation used to transform the ADC result into a voltage. With AVCC = 5 V as reference voltage, it should be something like V = 5 V * ADCresult / 1023.
In my 1st answer I mentioned the leak current or as was mentioned later the divider effect of the test point connection. This is most probably the reason for the differences.
You should look at R values in order to quantify the effect:
- for the Arduino : the internal resistance of the source, the R value in front of the capa, the impedance of the Arduino analog input which is the input impedance of the ADC connected to it.
- for the oscilloscope the impedance of the input.
Then you can compute the effects at least as order of magnitude.
There are digital oscilloscopes with a sample frequency higher than the above mentioned value. If you program the Arduino for sampling at same frequency you can better compare the results even if there is a time lag between the 2 series.If you are interested in the voltage evolution then your sampling frequency should be according to the time constant (RC). It depends on what you do with the acquired values.
Nick Hamburger and Joerg Fricke many thanks for you suggestions and answers. I could figure out the things.
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering