AD5933 Eval Board isn't response like AN-1053

Dear, ADI Eng Zone community

I tried to use the AD5933 Eval Board by following the application note: AN-1053. I started with the system setup. The result works well (like Pic.1 at below) as shown in Fig.2 page3 in AN-1053, then I moved to measure unknown impedance 15 pF capacitor as AN-1053 recommended, but the measuring result didn't change by the frequency (50-55 kHz) as shown in Pic.2.

What's wrong or what am I wrong? 



  • All settings seem to be correct. What is the value of your Rfb resistor? The capacitor reading is suspiciously close to half that of the resistor, while for 15 pF it should be about the same. What steps did you go through while calibrating the board? Could it be that instead of replacing your calibration resistor (of, presumably, 200 KOhm) with the capacitor you connected your capacitor in parallel to it? Could it be that the capacitor you are using is considerably larger than 15 pF? 

  • Dear Snorlax,

    Thank you for your response. My step is... 1. Put two 200kOhm to Rfb and Rcal (at Z) on the Eval board, then start sweep and results were shown in Pic1. 2. Leave 200 kOhm at Rfb on the Eval board, then remove Rcal 200kOhm (at Z) out of the board, and replace it with capacitor 15pF, finally to start sweep and results were shown as Pic.2. 

    Your question: Is my capacitor larger than 15pF?  Umm...I used the capacitor, which labels number (150) as shown in below picture. Could you have any suggestions for me, please?

  • Everything seems to be correct. The only simple suggestion would be to try measuring resistors of higher than 200 kOhm nominal, say 300 or 400 kOhm and see if the reported value is close to what it should be.

    If you have access to an oscilloscope, it would be worth trying to look at the voltage waveform at the RFB pin of AD5933 with the 200 kOhm calibration resistor and then with the capacitor: in both cases there should be a well-formed sine wave of about 2V peak-to-peak and baseline of about 1.65 V (= Vdd / 2).

  • Dear Snorlax, 

    1).  I tried to measure 300kOhm and the reported value was close to 300kOhm.

    2). I used an oscilloscope to look at waveform at RFB pin 200kOhm (CH1, yellow line) with 200kOhm as Rcal (CH2, blue line). The AD5933 Beta Version REV1.0 shown the same result as Pic1. and the results from the oscilloscope are shown at below.

    2). I changed Rcal to the capacitor (15pF, CH2 blue line) and leave R 200kOhm at RFB (CH1, yellow line) the results from the oscilloscope are shown at below and the result from AD5933Beta Version REV1.0 still shown the same result as shown in Pic.2.

    Well, I have some questions, Is it correct when I measured the capacitor? It got a square wave, not a sine wave. If it correct, why the impedance of the capacitor, which measured, is still a constant?

  • Thanks for taking and sharing the pictures!

    From the second picture it is evident that the capacitor is either way larger than 15 pF, perhaps 150 pF or it is not a capacitor at all (although it looks very much like a capacitor and the nominal should be read as 15 pF). The voltage on RFB pin is hitting the power supply rails - it is a sine wave which is severely clipped at the top and the bottom and thus looking like a square wave. The impedance under test Z and the RFB define the gain of the AD5933 internal input OPAMP G = Rfb / Z, so when Z is smaller than Rfb, gain is more than 1 and, if the excitation voltage is 2V p-p, it causes the output of that OPAMp to clip the output voltage at the power supply rails.

    Now, when this "square waver" is fed to the DFT, which is expecting the sine wave for the DFT algorithm to work properly, the result is certainly incorrect.