Post Go back to editing

How to power the AD5933 evaluation board?

I recently bought  EVAL-AD5933EBZ. 

I have a question to ask before using the device: 

Will, connecting the eval-board to computer PC ,going to power the onboard components like AD5933 , MCU and others ? or I have to design separate power supply.

Please reply

Parents
  • Please refer to your board user guide UG-364. On the first page it says: "The user has the option to power the entire circuitry from the USB port of a computer." Most likely the board has arrived configured to be powered from the USB port. You can double-check by making sure the jumpers on the board are set as described in Table 1. of the guide: LK5 should be in place and LK6 in position A, also refer to Figure 33. Linear Regulator on the EVAL-AD5933EB Evaluation Board.

  • Thanks, 

    Coming to calibration of the evaluation board,

    I don't find any option to adjust the feedback resistor (RFB) in the front panel of the evaluation board software. While reading the user guide UG-364, I found this " note RFB = 200 kΩ. Is it already programmed value. Cant we change this value ? How to change RFB in the evaluation board? Please reply. 

  • What should be its amplitude?

    Most lab generators have digital output (meander) of the set frequency. The amplitude depends on VDD supplied to the AD5933, on the evaluation board it is 3.3V, so the amplitude should be for 3V digital circuits. 

    Can I connect the ground of the general-purpose lab generator

    Yes, you certainly should. It should happen automatically if you use cable to connect the generator output to  CLK1.

    say,4KHz

    Could be a bit low, some folks in this forum reported I2C communication issues, when MCLK was too low. Better to start at 100 KHz or so, make sure that I2C works and then go lower. I wonder if somebody from ADI could comment on this as operation at lower frequencies is not adequately documented.

    how to select the excitation voltage ranges

    It is selected based on your desired impedance range, PGA gain factor and the value of your RFB. For the chip to measure correctly, there should be a well-formed sinewave at the input of the built-in ADC. If gain throughout the system is too high the input internal OPAMP or the PGA may clip the sinewave at the power supply rails, so the ADC will digitize the distorted waveform and feed it to the "DFT", which assumes that it is processing the sinewave, while it does not, causing errors in the Re and Im values.
    Therefore, when, for example the Range 1 (2V) is selected, the total system gain cannot be noticeably more than one or the sinewave at the ADC will be distorted. This means that when PGA gain is 1x, measured Z cannot be lower than Rfb, or the sinewave will be distorted and Re and Im readings will be erroneous. If PGA gain is set to 5x Z cannot be lower than 5*Rfb. The formula connecting the Rfb with VOUT (called Vpk in the formula) and other parameters can be found at the bottom of p.5 in this application note.There is a bit more discussion on the topic here.

    I am seeing a jump from 1 K ohm(calibration value) to the unknown value at the initial or starting frequencies.

    This is hard to diagnose remotely... What are the settings when you observe this behavior? A screenshot of your demo window with the settings could help.

  • Thanks.

    You asked for the screenshot to understand the initial jump at low frequency.

    In the above figure pink highlighted is the jump I am talking about. Starting freq: 1000Hz.

    When I increased the start frequency to 5 KHz , the error is reduced and the same is true above for frequencies > than 5 KHz.

    What is the reason for this ?

  • I am seeing a jump from 1 K ohm(calibration value) to the unknown value

    The jump probably does not have much to do with the calibration resistor value, likely a coincidence.
    The settings seem to be in order on your screenshots, everything should work as advertised. Would it be possible for you to:
    1. Connect the calibration resistor and click the "Start Sweep" button and then "Download Impedance Data"
    2. Do that with some different resistor
    3. Do that without any resistor connected
    4. Post all the files generated in the process? 

  • Hi

    Sorry for the late reply.

    As you asked, I did the following:

    1. Connected, RFB= 1 K ohm and Rcal = 1 K ohm.

     and start sweep, saved data as "calibration_1kohm "

    2. connected 10k ohm and start sweep, saved data " connected_10k ohm "

    3. connected 5k ohm and start sweep, saved data " connected_5k ohm ".

    4. open circuit between terminals where unknown impedance connected normally. run sweep and saved data as- " connected_nothiong_opencircuit " ( RFB is connected always from 1 to 4)

    5. i could not upload the excel file. Through " Insert". So uploading the image of excel files

    1. connected, RFB= 1 K ohm and Rcal = 1 K ohm.

    2. connected 10k ohm and start sweep

    3. connected 5k ohm and start sweep.

    4. connected nothing, open circuit, with RFB only. 

    Please check

  • The chip seems to operate as expected. What you are bumping into is the systematic error (or an artifact) of the particular flavor of the "DFT" implemented in the chip, which causes the accuracy to generally degrade at frequencies lower than 10 KHz. tfsoft has found an article detailing the causes of this effect and methods of reducing/eliminating it in/from the data. The easiest way to get the chip to operate at lower excitation frequencies is to reduce MCLK or go through a somewhat cumbersome math from the article.

    Some web wiseacres decided to block certain file extensions form the forum posts and seriously scrambled formatting when trying to generate a preview, so you need to zip your files and attach the resulting archive. Or save those files as pdf-s or txt-s and attach, please see your first file in the next message in all three formats.

  • 3187.Picture1.zip

    Frequency,Impedance,Phase,Real,Imaginary,Magnitude
    1000,1000,0,-16876,-10947,20115.57
    1050,999.925,-3.76E-03,-15116,-13215,20078.09
    1100,999.8346,1.98E-02,-13101,-14960,19885.62
    1150,999.7058,2.21E-02,-10998,-16143,19533.37
    1200,999.6371,3.14E-02,-8908,-16849,19058.89
    1250,1000.067,-4.84E-02,-6942,-17124,18477.63
    1300,999.7635,0.010787,-5125,-17064,17817.01
    1350,999.9394,-1.64E-02,-3538,-16709,17079.47
    1400,999.9154,4.32E-02,-2160,-16145,16288.85
    1450,999.6509,-2.09E-02,-1048,-15415,15450.58
    1500,999.6537,-1.55E-02,-170,-14570,14570.99
    1550,999.5052,-3.02E-02,435,-13679,13685.91
    1600,999.6785,7.80E-03,808,-12784,12809.51
    1650,999.777,-2.02E-02,949,-11948,11985.63
    1700,999.3124,-1.85E-02,900,-11212,11248.06
    1750,999.55,-1.79E-02,707,-10618,10641.51
    1800,999.7055,-7.00E-04,422,-10169,10177.75
    1850,999.5962,-6.02E-03,94,-9881,9881.447
    1900,999.3837,8.28E-04,-228,-9730,9732.671
    1950,999.3833,1.88E-03,-517,-9702,9715.765
    2000,999.4984,8.06E-03,-741,-9760,9788.089
    2050,999.5068,8.36E-03,-891,-9874,9914.119
    2100,999.5151,8.43E-03,-976,-10016,10063.44
    2150,999.5129,2.71E-03,-986,-10169,10216.69
    2200,999.432,8.54E-03,-945,-10315,10358.2
    2250,1000.15,3.77E-02,-853,-10429,10463.83
    2300,999.2561,1.39E-02,-732,-10515,10540.45
    2350,999.439,7.22E-03,-589,-10562,10578.41
    2400,999.5201,-9.71 E-03,-435,-10572,10580.95
    

  • Hi 

    I am currently working on AD5933 chip . I want to do frequency sweep from 1khz to 10 khz in steps(freq. increment) of 1k khz . And then from 20 kHz to 100 Khz in steps of 10 Khz. This should be done in a single frequency sweep.  

  • I mean in a single sweep, How to give different start frequencies ( in my case, 1 khz and 20 khz) and different increments ( 1khz and 10 khz) . I am unable to code this. Please help.

  • It is not possible to perform what you described in a single sweep. The chip performs a "hardware-assisted" sweep (meaning that your code can poll certain hardware register to see if the sweep is complete while the chip increments the frequency and counts the number of frequency points in the hardware) with a pre-programmed start frequency, frequency increment and number of frequency points in the sweep and non of the above can be changed in the course of the sweep.
    A somewhat close to what you describe can be two sweeps with different initial parameters ran immediately one after another.

    Perhaps it would be worth reviewing the User Guide, Fig. 27. p. 12  and the VBA code on pp 13-16. Referring to Fig. 27, at the very bottom of the diagram, after the current sweep is finished, instead of proceeding to "PROGRAM THE AD5933 INTO POWER-DOWN MODE" your code should cycle back to the very beginning and start the next sweep with the next set of start frequency, frequency increment and number of frequency points (block "PROGRAM FREQUENCY SWEEP PARAMETERS INTO RELEVANT REGISTERS (1) START FREQUENCY REGISTER (2) NUMBER OF INCREMENTS REGISTER (3) FREQUENCY INCREMENT REGISTER"). It is possible to run any number of consecutive sweeps this way, the code should exit the loop to "PROGRAM THE AD5933 INTO POWER-DOWN MODE" after all the sweeps are done.

    Hope this helps!

  • Hello!

    Hope everything good from your side.

    I want to clear some doubts related to AD5933 IC w.r.t external clock pin ( pin number - 8 ).

    Actually, I want to do a low-frequency sweep of less than < 5000 Hz. In the eval board datasheet, it is mentioned to reduce the external clock for this. 

    In the UG -364 datasheet, page -20, table-2, Freq's - (4, 2, 1 ) MHz is used for low-frequency sweeps. In the evaluation board, 4 pins SMD external oscillator was used.  

    Question-1:  Can I use a function/waveform generator to drive the pin-8? Is it sinusoidal or square wave? If yes, is the peak-peak voltage = 6.6 V(3.3 V-(-3.3 V) and amplitude =3.3 V setting fine?

    Question 2: Or Do I need to use a 3.3 V crystal oscillator of a specific frequency for this.

    Please help.

    And do you have any idea about the availability of AD5933 IC? For almost a year, it is showing unavailability in stock. I am really worried about it. Do you have any idea about this? Please guide me. It is affecting my work badly. My guide asking me, whether this chip is being manufactured or not? 

Reply
  • Hello!

    Hope everything good from your side.

    I want to clear some doubts related to AD5933 IC w.r.t external clock pin ( pin number - 8 ).

    Actually, I want to do a low-frequency sweep of less than < 5000 Hz. In the eval board datasheet, it is mentioned to reduce the external clock for this. 

    In the UG -364 datasheet, page -20, table-2, Freq's - (4, 2, 1 ) MHz is used for low-frequency sweeps. In the evaluation board, 4 pins SMD external oscillator was used.  

    Question-1:  Can I use a function/waveform generator to drive the pin-8? Is it sinusoidal or square wave? If yes, is the peak-peak voltage = 6.6 V(3.3 V-(-3.3 V) and amplitude =3.3 V setting fine?

    Question 2: Or Do I need to use a 3.3 V crystal oscillator of a specific frequency for this.

    Please help.

    And do you have any idea about the availability of AD5933 IC? For almost a year, it is showing unavailability in stock. I am really worried about it. Do you have any idea about this? Please guide me. It is affecting my work badly. My guide asking me, whether this chip is being manufactured or not? 

Children
  • Question-1

    You most certainly can use the external lab generator to provide variable-frequency MCLK. Per User Guide, Fig. 31. p. 23 you need to remove LK3 while keeping (or putting) LK4 in place and connecting the generator output to connector CLK1. The voltage should be a digital square wave meandering between 0 and 3.3V. Depending on the model, some generators might have such digital output.

    And do you have any idea about the availability of AD5933 IC?

    I am not affiliated with the ADI and cannot comment on the availability. Hopefully somebody from ADI could help you with that.

  • Thank you!

    What connector should I use to connect the external laboratory function generator to the CLK1 of the AD5933 Evaluation board? CLK1 is a SMB socket that I read in UG-364 Datasheet.

  • I found this connector which is SMB to BNC on the Digikey site. Details as below:

    BNC MALE TO SMB FEM, RG174, part no: 24-415-0200-MM500-ND. Can I use this?

  • This should work. A less solid but workable solution would be to get some piece of cable with BNC connector on one end (those tend to float around) and just connect the shield to the Demo Board ground and the signal wire to the exposed part of the LK4. 

  • Hello!

    This is about arduino-AD5933 interfacing!

    I am using arduino-ad5933 code available on , https://github.com/mjmeli/arduino-ad5933.

    I calibrated AD5933 IC , using RFB = 1 K Ω and R_calibration = 10 K Ω. The results of unknown impedance are fine. The problem I am facing are listed below: 

    1. When  RFB = 10 K Ω and R_calibration = 33 K Ω used. The results are fine. But values are negative. Like, for unknown impedance of 50 K Ω, output displayed as  -50 K Ω. Negative values are displayed. Why like this ?

    2. And when I used R_cal values above that,  like, I used R_cal= 43  K Ω or 50 K Ω and anything above. 

    The results are displayed as  -22  K Ω  for every frequency in the entire frequency scan and same for any unknown impedance. 

    I am unable to resolve this problem. Please help me on this. 

  • I am using arduino-ad5933 code available on , https://github.com/mjmeli/arduino-ad5933.

    Do you use this code from the repository you indicated to print values?

    What are the settings you program into the chip: excitation voltage range, PGA gain, MCLK (external or internal?), frequency range, etc.?

    Debugging third-party code remotely is rather difficult, we can mostly look at the symptoms and guess. Can you export the printout into a text file and post it here through Insert->Image/video/file ?

    Particularly useful would be raw Re and Im data sets collected when:
    1. Nothing is connected in place of Z
    2. R_calibration is connected in place of Z
    3. "R_unknown" is connected in place of Z.  

  • Hello Snorlax!

    My board Connection was not appropriate. Now it is solved.

  • Hello ! 

    I have a question related to figure -1 in the datasheet of AD5933 IC and UG-364 user guide. 

    From VOUT to voltage Signal,  just  before GAIN block (inside the AD5933 block), what I understood is VOUT , observes an gain of RFB/Z_calibration. But still Vdd/2 appears along with the amplified sinusoidal signal .  Only thing is Vdd/2 or offset is not amplified. Is this correct ? Won't this Vdd/2 create problem or is it eliminated inside the chip ? 

  • Yes this is correct: the Vdd/2 is not getting amplified. The OPAMPs inside the AD5933 are also referenced to Vdd/2 internally, so this Vdd/2 DC voltage serves as the so-called virtual ground throughout both the external and internal circuits.

  • Hi

    I am currently driving the Eval-Board (AD5933) with 2 MHz external frequency. This is to do frequency scan from 300 Hz to 5 KHz. For fixed unknown resistors measurement accuracy is fine. But when I connected RC combination ( 100 K ohms || 10 nF , 33 K ohms || 10 nF), the impedance measurement shows greater inaccuracy and impedance values changing with every new scan. Is the problem due to settling time ? Default value is 15. I tried changing the settling time  values . some change I could see but again poor repeatability with fresh scans. How to resolve this ?

    Please help.