Using RE0 and SE0 with HS loop for 2-wire impedance on the AD5941

Hi, Akila .

Can your team please help look into this query?

Warm regards,
Jo

I turned up more information by plotting the raw data. Using RE0 and CE0 for calibration is causing noise on the data acquired with AFE1 and AIN1!

I get good data when:

• RCAL0 and RCAL1 are used for calibration
• The device under test is connected to AFE1 and AIN1

I also get good data when:

• DE0 and CE0 are used for calibration
• The device under test is connected to AFE1 and AIN1

I get good calibration data but bad test data when:

• RE0 and CE0 are used for calibration
• The device under test is connected to AFE1 and AIN1

Also, everything worked on my prototype. I am only getting this error on the production version. I am carefully going through the hardware and code for any change that could relate to this. So far I haven't found anything, I am setting up the measurements in the exact same way. It is almost like something in the chip is automatically closing a switch somewhere when I use RE0 and CE0, connecting something to the high-speed loop that I don't want connected.

I tried the same code on the same microcontroller with both my prototype board and production board. It works on the prototype board but not on the production board. As described, the calibration data from RE0 and CE0 does not have the noise, but when I use those two pins for the calibration sample, the test data from AFE1 and AIN1 is unusably noisy. When I get calibration from other pins and use AFE1 and AIN1 the data is good.

The prototype was made about a year before the production board. Both use an AD5941BCPZ. The CHIPID register returns 0x5502 on both of them.

I am reviewing my design for anything that could cause this, but so far haven't found anything. I am also checking all my data collection and analysis to see if I may have a mistake there.

It would be very helpful to have some insight as to if something inside the AD5941 may cause this, or if the problem is that the chip does not support what I am doing.