impedance measurement (AD5933)

Hello!

I referred to the datasheet of AD5933 and the user guide (UG-364)of the evaluation board for AD5933.
Following questions arose. Please help me to clear these doubts.
1. For low impedance measurement (<500 ohms). An additional circuit (figure35 in AD5933 datasheet)suggested. Is it built on the evaluation board, so that we can connect? Else we need to design separately.
2. In the GUI panel there is an option for multi-point calibration. Is it two-point calibration or more than two? 
3. About repeat frequency: I could not find this option on the GUI panel. Can we repeat the frequency sweep from the software accompanied with eval-board?
4. Is the output excitation frequency 100 kHz, when MCLK is at 16.67 MHz. Do increase or decrease in MCLK effect output excitation frequency. What is the maximum clock input we can give to the device, Is it 16.67 MHz? can we increase it further. Can we increase the excitation frequency beyond 100 kHz by increasing MCLK? 
 
Waiting for your reply.

Top Replies

Parents
  • While you are waiting for a more official reply from the ADI, I can try to answer some of your questions.

    For low impedance measurement (<500 ohms)

    If the version of your EVAL board is the same as covered by the UG-364, the limitation of the lowest impedance that can be measured is the output current that can be produced by the U1-A and consumed by U1-B OPAMPs, which is presumably ±80mA. So, if your excitation voltage is 2V p-p = ±1V, then an estimate for Zmin = 1V / 80mA =12.5Ω . If your excitation voltage is lower, say Range 4, which is 0.2V p-p, Zmin = 0.1V / 80mA =1.25Ω .

    Can we repeat the frequency sweep from the software accompanied with eval-board?

    I do not think you can, unless you use some third-party software that will repeatedly push "Start Sweep" button for you such as AutoHotkey or PTFB.

    Do increase or decrease in MCLK effect output excitation frequency.

    Yes, the output frequency is directly proportional to MCLK for any frequency code programmed into the chip. 100KHz is maximum operating frequency that is recommended in the datasheet, but with 16MHz MCLK you can program in frequency up to 1Mhz and the chip will generate it (it will look a bit ugly as it will contain only 4 voltage steps per period). 100KHz limitation comes form the receiving side of the chip, where there is a built-in low-pass filter (LPF) with 100KHz cut-off frequency. There is a bit of discussion on this subject on Page 20 of the UG-364 under "Measuring Higher Excitation Frequencies." From my personal experience the roll-over of this LPF is not as steep as shown on Fig. 28 and there is enough signal reaching the ADC even at 1MHz to perform measurements, albeit at some loss of accuracy.

    I overclocked the AD5933 at 36MHz MCKL without seeing any loss in performance and adverse effects such as overheating, communication issues, etc., although it is not within the recommended specs and chip longevity and performance cannot be guaranteed.

Reply
  • While you are waiting for a more official reply from the ADI, I can try to answer some of your questions.

    For low impedance measurement (<500 ohms)

    If the version of your EVAL board is the same as covered by the UG-364, the limitation of the lowest impedance that can be measured is the output current that can be produced by the U1-A and consumed by U1-B OPAMPs, which is presumably ±80mA. So, if your excitation voltage is 2V p-p = ±1V, then an estimate for Zmin = 1V / 80mA =12.5Ω . If your excitation voltage is lower, say Range 4, which is 0.2V p-p, Zmin = 0.1V / 80mA =1.25Ω .

    Can we repeat the frequency sweep from the software accompanied with eval-board?

    I do not think you can, unless you use some third-party software that will repeatedly push "Start Sweep" button for you such as AutoHotkey or PTFB.

    Do increase or decrease in MCLK effect output excitation frequency.

    Yes, the output frequency is directly proportional to MCLK for any frequency code programmed into the chip. 100KHz is maximum operating frequency that is recommended in the datasheet, but with 16MHz MCLK you can program in frequency up to 1Mhz and the chip will generate it (it will look a bit ugly as it will contain only 4 voltage steps per period). 100KHz limitation comes form the receiving side of the chip, where there is a built-in low-pass filter (LPF) with 100KHz cut-off frequency. There is a bit of discussion on this subject on Page 20 of the UG-364 under "Measuring Higher Excitation Frequencies." From my personal experience the roll-over of this LPF is not as steep as shown on Fig. 28 and there is enough signal reaching the ADC even at 1MHz to perform measurements, albeit at some loss of accuracy.

    I overclocked the AD5933 at 36MHz MCKL without seeing any loss in performance and adverse effects such as overheating, communication issues, etc., although it is not within the recommended specs and chip longevity and performance cannot be guaranteed.

Children