AD5933EBZ evaluation board : oscillation for one test impedance
the board works quite accurately for resistive loads. However for the test case of a 620 pf capacitor at the test port , there is an oscillation at about 300 KHz when using u1a and u1b as buffer and tia. The oscillation is visible at pin 4 ( output of the internal op amp in the ad5933 that is input to the ad)
the calibration is done with a 8.2K across the Z terminal and 8.2K for the Rfb using the eval board sw rev B. . The oscillation is observed for all the output excitation values when the capacitor is placed on the Z terminals. the pga control is left at 1. the test frequency is 10000 with frequency increment of 1 and number of increments 20. On an oscilloscope the pin 4 (input to the adc ) is at the full 3.3v pp for this load. Normal resistive loads follow the normal gain of the op amp and tia.
the same procedure is followed as with the resistive loads : program device registers, calc gain with the cal resistor in place , replace the cal resistor with the capacitor , single measurement.
quite remarkably, looking at the download impedance data to the file , for 10 KHz the impedance is reported very close to 24959 which would be a 637pf capacitor , the phase angle is reported as -89.2 ( a capacitor). when compared to a de5000 lcr meter, it would be about 3.4% off
I am disappointed of the undesired oscillation observed on this eval board , but i understand it has in this case little consequences.
Does anyone have any suggestions to prevent undesired oscillations on this reference board .