EVAL-AD5940ELCZ sensorPal Calibration Resistor

Hi 

I am using the EVAL-AD5940ELCZ with the sensorPal application. I am confused about the functionallity of the Calibration resistor in Cyclic Voltammetry.

I have noted that on the board the RCAL resistor is  value 200 ohms this is also the default value on the sensorPal GUI.

however if I leave it at the default setting I get a much higher current output than I would expect from my sensor  400 uA as opposed to in the range of  0 - 21uA

My settings:

initial potential  0mv

Final potential 1000mv

Step potential 2.148 uV (default)

Scan rate 100 mv/s

current range to 0uA - 21uA (85K RTIA) this is the range i expect.

Calibration resistor 200 ohm

Are the sensorPal examples only applicable for the on board dummy sensor or can custom sensors be used.

Do need to change the RCAL resistor if using a custom sensor.

I have noticed if I increase the calibration resistor value drastically (on the GUI only) the current output will drop to range that I would expect.

Also should step potential units on the GUI be in mV instead of uV  according to AN-1563 "The minimum step size is 537 µV on the AD5940"

Regards

Ian



Edit
[edited by: ian079 at 12:13 PM (GMT 0) on 5 Jun 2019]
Parents
  • +1
    •  Analog Employees 
    on Jun 5, 2019 4:17 PM

    Hi Ian,

    The calibration resistor is used to calibrate the internal Rtia on the LPTIA. SensorPal can be used for your own sensor also. Ensure jumpers JP9, JP10 and JP11 have disconnected the dummy network from the CE0, RE0 and SE0 pins. Are you connecting your sensor via the cables or wired directly to P7?

    RCAL should be chosen to be closer to the value of RTIA you plan on using to maximize accuracy of the calibration routine. I would suggest replacing RCAL with a 10k resistor on the evaluation board. 

    Yes, the units should be in mV. We will update.

  • Thanks for the reply

    yes I am using the cables,

    We are trying to replicate a lab based potentiostsat experiment on an embedded device.

    the sensor used is a  prototype with two electrodes for water sample analysis

    A cyclic voltammetry scan is used to observe oxidization peaks in the samples current response

    I am currently connecting the red crocodile clip (CE0) and blue(RE0) to the counter electrode of the sensor and the green(SE0) to the working electrode as is done using the lab potentiostat.

    I am unsure if this is the correct method using the AD5940 should i use the usb cables or the BNC connector considering it is a water sample i am dealing with.

    From reading the data sheet i think it is possible to route the Cyclic voltammetry waveform signal through the BNC connector by setting the appropriate switches in the programmable matrix but this is probably not possible using sensorPal.

    I am also not sure if i should use the LTC6078 op amp in the circuit in this application or bypass it.

    sensor impedence is 300-400k

  • 0
    •  Analog Employees 
    on Jun 24, 2019 3:13 PM in reply to ian079

    Using the cables is certainly the correct approach. For the cyclic voltammetry  scan, you don't need to use the op-amp. That opamp is for a pH sensor and also for measuring very high impedance sensors where the magnitude is several Mohms.

    Have you been able to get some reasonable data yet?

Reply Children
  • Hi

    I can set and read registers, read temperature sensor etc and I can set the Low power DAC to output 12 bit and 6 bit voltage levels. 

    I am having issues trying to generate a very slow Cyclic voltammetry waveform of scan rate 100mv/s, following the data sheet I need to use the low power DAC. 

    I have set the WAVEGEN bit to 1 in the AFECON register

    and Bit 6 in the LPDACCON register to 1 

    and set up the various registers for the trapazoid generator to achieve the appropriate waveform.

    Where I run into trouble is this statement p92 data sheet :

     " The 32 kHz oscillator must be selected as the system clock when using the waveform generator with the low power DAC"

    if I do this by setting bits the SYSCLKSEL bits to 1,0 in the CLKSEL register, I kill my SPI communication to the AD5940 i assume because SPI 1MHZ is too fast now.

    is there a method of using the 32 khz clock for the low power DAC without affecting the SPI interface 

    Regards

    Ian

  • 0
    •  Analog Employees 
    on Aug 13, 2019 1:55 PM in reply to ian079

    Hi,

    We have an example application in our SDK which carries out a cyclic voltammetry measurement without using the waveform generator. Refer to AD5940_Ramp example for details. The LPDAC is updated through software instead of the waveform generator. It is not possible to communicate with the part when the 32kHz is selected as system clock. 

  • Yes i've seen that example and was wondering why the generator was not used

    Thanks

  • Hi can you confirm that this formula in the data sheet for ADC input voltage is correct 

    in my application i am substituting VBIAS_CAP with  the value of LPTIA_N node as I am conducting a DC current measurement as below

    I have noticed in the sample firmware that the ADC calculation used  is  the following

    float kFactor = 1.835/1.82;

    tmp = ADCcode - 32768;

    fVolt = ((float)(tmp)/32768)*(AppAMPCfg.ADCRefVolt/1)*kFactor;

    here there seems to be no reference to VBIAS_CAP or an alternative negative ADC input maybe this is added else where in the code ? And kFactor what is this 

    Please advise

    Regards 

    Ian

  • 0
    •  Analog Employees 
    on Sep 2, 2019 11:18 AM in reply to ian079

    Adding VBIAS_CAP in the equation is only required when the negative input to the ADC is the 1.11V reference, or ADCMUXN_VSET1P1 in the SDK

    The kFactor is a constant that is required to calculate the correct voltage.