I need some help from friends in this forum.
I have my system which has 16 channels sensor(E1,E2....,E15,E16 ) made from the conductive plate that put on the outside of the acrylic circle pipe wall. Each sensor is shielded by another conductive plate and this made the sensor function as a cap. Or it is known as electrical capacitance tomography sensor.
Now,what I want to do is to determine the resistive load of the i.e conductive liquid inside the pipe. So, i made one of the sensor as a source,i.e E1 by giving voltage input 10 Vp-p, 2 MHz sine waveform and all the remain sensors,E2...E16 as the detection sensor and measure the current value to the ground.
Thus, in processing the detection current value which is in between 0.1mA to 1mA, i need the I to V converter amplifier and other steps so that I can used it in ADC part. For time being, i used the LT1360 as my op amp to convert current to voltage amplifier with inverting input.. However, I noticed that, if i used the same gain of inverting feedback, i.e Rf=0.9K ohm; the output of course will be the invert of the input. The nearest of detection sensor, E2 showed no phase shift(that is totally inverting of the voltage input). But, the phase shift will keep increasing across the sensor, E3..E9 and for the farthest detection sensor,E9 ; it showed the phase shift is almost 180 deg.Then, it will decreasing again till no phase shift at E16. The output voltage peak to peak showed like a 'smile graph' which is between 0.1V and 1V.
However, the correct signal should be in zero phase shift for all detection sensors with the 'smile' curve of amplitude voltage.
What is the suitable approach/ic that I can use to make it all the signal in zero phase shift with the operating frequency 2Mhz?
I read some of people use phase shift demodulation,PSD technique, but i think it is complicated.
I also read some of them use programmable gain amplifier,PGA but is there any other method that I can use to solve this problem?
So sorry if this post is too long.
I attached the drawing for your basic understanding.
When one channel acts as transmitter with voltage source, all the remaining channels will act as receiver and will measure the I-V converter amplifier one by one in sequence.
Because the concern is the Rx value that is the resistive load only, the Cx is assume to be small by implementing the high frequency, 2Mhz.
That is why I want the signal in zero phase shift.
Im using the coaxial cable. Yup, I also think the same coz when I adjust the cable the output effect in term of the phase shift even though the effect is small. But, i dont think that it will cause a big issue. or is it?
how about the circuit of I-V that I implemented?is it correct?Or is there any other solution so that when I use the I-V converter the signal will be in zero phase shift?
I didn't quite understand what you are doing, but one degree at 2 MHz is 1.39 nanoseconds.
A few inches of cable can have that much difference.
Can you post a drawing?
With respect to the I to V, you are using an op amp from Linear Technology, so I suggest you ask them.