Post Go back to editing

Inaccurate impedance measurement using Eval-ADUCM350EBZ

Category: Hardware
Product Number: ADUCM350

Hello,

I am using Eval-ADUCM350EBZ board to measure impedance. My goal is to measure a difference in magnitude and phase of the unknown target in real time between 100~10kHz frequency range. What I am gonna measure should be capacitive, and the magnitude range is 1k~10k. Absolute accuracy is also important for me, but relative accuracy is more important.

FYI, I didn't change any component on the daughter board (Rcal: 1kOhm, Rtia: 33kOhm, Ctia: none, Raccess: 4.99kOhm, Rlimit: 3kOhm, Ciso: 47nF)

First, I tried 2-wire setup in IAR using the given 2-wire example code.

When I measured 1kOhm and 10kOhm, I could measure a correct result and the accuracy was good. However, If I switch to 1kohm + 220nF or a pure capacitance 220nF or 470nF, the measured value was wrong. In more detail, the results of 220nF and 470nF have a relationship between their values, but the absolute values were wrong. And for the 1kohm + 220nF, I could only see 1kOhm, which means the result was exactly the same as 1kohm.

My initial plan was to use just 2-wire setup because relative accuracy is more important for me. But the results of complex impedance measurement was bad, and now I am worrying to use this setup. I wonder if there is anything wrong what I did during the setup or measurement. I wanna know whether my setup is wrong, or I need to change some values or given code.

Second, I also tried 4-wire setup in IAR using the given 4-wire example code.

I tried to measure 1kohm, 10kohm, and 20kohm. But in this case, the absolute values were wrong. The magnitude values were like ~31.4 times higher at 1kHz. And phases were 180 degree. Especially, if I change the frequency, the mag and phase changed together, which I couldn't see during the measurement using 2-wire setup. I understood that 4-wire only concerns the impedance at the sensor node, so the frequency will not affect the measured value especially for my case with only resistance (1,10,20kohm). In this case, I also wonder if there is anything that I did wrong and have to change here. I wanna know whether my setup is wrong, or I need to change some values or given code.

FYI, I attached my setup photo. (black cable: for 4-wire measurement, red and orange cables: for 2 wire measurement)

  • Regarding my 4-wire question, I found out that the default passive components in the daughter board does not match to reported values on the datasheet. Especially, RG is 120kohm (100kohm in the document), and Rtia is 1kohm (33kohm in the document). So the absolute value becomes closer to the actually value after I change the values in my code in IAR. However, The value is not stable. It changes too much every time I measure it again. I tested with 1kohm ~ 20kohm with 100~10kHz range, and I see the same problems.