Use 1KΩ resistor Calibration Impedance , change unknow resistor --> "L " , the Inductance parameter completely wrong.

Can provide reference information? I have read AN-1252 ,no anly help.

Use 1KΩ resistor Calibration Impedance , change unknow resistor --> "L " , the Inductance parameter completely wrong.

Can provide reference information? I have read AN-1252 ,no anly help.

Hi,

I did not notice the attachment, my apologies.

Using your settings and the formula below, we will obtain 800ohms for Zmin.

and using the formula below, we will obtain 2200ohms for Zmax.

and converting this impedance range to inductance range using the formula below, we will obtain 4.24mH to 11.67mH.

I would suggest that you determine first the range of inductance that you want to measure then convert it to impedance range then compute for the Rcal and Rfb.

Best regards,

Mark

Hi all,

I am the end-user who found this problem and ask Jambo for help.

I registered an account just now.

First of all, I have to apologize that I type the inductance value wrong...

L for measurement is 500 uH (micro).

But still, the calculated value is twice of the true value.

After a few days of observation, I found that when we measured calibrated resistor, it has a big imaginary part.

I calculated the phase, about -27 degree. (From our own board)

Then I tested another size of pure resistor.

Good news! Still -27 degree.

Therefore, I treated this phase as the so-called system phase from 5933 datasheet.

The key point is next.

After we calibrated from a resistor, recording a gain factor and the system phase, directly change Z to what we want to evaluate (pure L or C) is not right, because we would lost the system phase we just got.

The correct way should be keeping the resistor serial to the L or C under test.

Then, we can bravely say that the L or C hides inside the sin(unknown phase-system phase) of the new calculated impedance value .

I don't know whether this is a common way to measure impedance by using other devices.

But at least it works for AD5933.

Just share for you.

I also suggest that if there will be next generation, make this process inside would be much more convenient.

Regards,

Webb

Hi Snorlax,

I'm not sure to catch your point. I would definitely not expect to measure the inductance at DC. My testing frequency is about 10k~100kHz to find the above mentioned way applicable. What I was trying to say is that the serial resistor is necessary when we want to evaluate inductance or capacitance because only in this way, we can truly determine the phase change and find the correct projection of Z.

Best regards,

Webb

Any DC coupling (resistor including) between VIN and VOUT shifts the operation point of the internal input OPAMP and PGA so they do not necessarily perform their AC function as intended . It becomes particularly messy as the DC level at the VOUT depends on the excitation amplitude settings (datasheet, page 13 under "Output DC Bias Level"). Once the DC interfering with the device operational point is eliminated through AFE or some decoupling - there is no need for any additional resistors to be able to measure inductances correctly. Measuring capacitance is usually no problem as it naturally blocks the DC between VIN and VOUT.

Hi,

I will move your question to Direct Digital Synthesis (DDS) community. Someone here should be able to help you.

Regards,

Mark