Getting wrong values while measuring capacitance as Device under Test (DUT) with Resistor as calibration impedance

This is the expected behaviour. With the mentioned configuration, the 47 pF capacitor produces a 400 mVpp signal at 30 kHz at the ADC, that is 8% of the dynamic range. That is the main source of error.

With 10 pF the signal is even smaller, so more error is incurred.

With 1 uF the signal overflows the ADC.

You need an algorithm to switch gains until a stable measurement is achieved.

Okay. Thanks for the explanation.

So need selecting PGA of 5 and to choose Zcal and RFB gain of 5 will serve the purpose?

Yes, you can either increase PGA gain or RFB value. Nevertheless these options result in lower SNR than increasing the transmitted signal amplitude.

I am trying to calibrate R in series with cap which results in to wrong phase values generated in the excel.

Rfb=570K, PGA=1 , EXTERNAL CLOCK, Zcal=100/k+10pF which is providing correct impedance value of 540K but not the phase angle of -79degree.

While clicking on view theoretical calibration profile the phase curve shown in the software is correct where as the measure(black line) is showing zero.

The values that are generated in .csv file after clicking on the download impedance is also different. The phase should be tan-1(Imaginary/Real) but which phase it is generating? 

When you calibrate using a reactive part, phase measurements are calculated relative to the phase of the calibration network. If you need to measure the phase of an unknown impedance, you must calibrate using a resistor.

The CSV file displays the phase in degrees.

The phase angle of -79 degrees corresponds to the phase of the RC network 100k+10pF at 30 kHz.

Okay thank you for clearing my doubt.

Is the ADC operating as a SAR type.

Yes, the ADC is a 12-bit SAR.

okay thanks

When I am using Rfb and Zcal 1Kohm with Zunknown=68nF,1nF,60nF iam getting correct phase angle and capacitance values but when i am replacing Zunknown as 1microF or in microF range getting phase angle in positive that is 20degree. Tried with PGA of 5 too.

Can you suggest why? What is the capacitor range which AD5933 can measure ? Is there any simulation online software of ad5933 where i can check these. Thanks in advance.

The impedance of 1 uF at 30 kHz is barely 1 ohm. With 1 kohm RFB, you get 1V/mA at the ADC input.

A rule of thumb is that RFB must be smaller than the impedance you are trying to measure. With RFB = Z you get 1:1 Tx/Rx voltage ratio with PGA gain 1, or 1:5 Tx/Rx voltage ration with PGA gain 5.

The device can measure a wide range of impedances by switching Tx range in PGA gain. Some users also add a multiplexor to use different RFB values.

Obviously, the SW must have some intelligence to switch ranges and decide when the measurement is good.