Post Go back to editing

Can anyone recommend the best circuit for skin to electrode impedance measurement using the AD5933.

Hello,

I am trying to build a portable system to measure skin to electrode impedance using the AD5933. Is there a recommend circuit i should follow?

Do i need to be reducing the current to make it safe to connect to a person?

Any circuit diagrams, tips or suggestions would be greatly appreciated.

Kind Regards,

James

  • Hi James,

    There is a good deal of papers etc on Impedance Spectroscopy with the AD5933 if you look it up on any search engine.

    You will need to meet IEC 60601 standards for it to be safe to measure human skin.

    You will need to add a dedicated AFE to the AD5933 to convert from voltage excitation to a current excitation configuration.

    Regards,

    Liam

  • Thanks Liam for your help.

    I have read through a few papers using the AD5933 with an Analogue Front End (AFE). The usual configuration is high pass filter, a voltage controlled current source, unknown impedance, and another voltage controlled current source.

    I was wondering if there was a standard circuit i could follow? Or if not a circuit perhaps some tips on how to create the circuit.

    Regards,

    James

  • Hi James,

    If helps, one of the electrodes is connected to a Howland current source (controlled by the AD5933) and the other extreme to a bias voltage , typically Vdd/2.

    An In-amp is used to convert from current to voltage and feedback the AD5933.

    The electrode impedance is  quite high so, ensure that the current source has high output impedance.

    Regards,

    Miguel

  • Based on Miguel’s recommendation, would the below circuits be appropriate for measuring skin to electrode impedance?

    A Howland Voltage Controlled Current Source (VCCS).

    (Found at http://www2.hawaii.edu/~krystina/ee496/PrecisionCCS.pdf)

    IL = (1/R)*Vi – (1/Ro)*VL

    Vi = Input Voltage

    VL = Voltage across load

    R = (R3*R2B*(R1 + R2A))  /  ( R3*(R2A + R2B) + R2A*R4)

    Ro = (R2B*(1+R2A/R1) )  /  ((R2A+R2B)/R1-(R4/R3))

    However, for some reason I need to take the voltage across the load to perform the calculation. Is there a way around this? or a way of automatically doing it?

    Current to voltage.

    I couldn’t find an Instrumentation Amplifier (IA) version of a current to voltage converter, there is a IA voltage to current converter and op-amp voltage to current converter, as well as an op-amp current to voltage converter. A simple op-amp current to voltage converter is shown below (from Electronic Devices by Thomas Floyd).

    Circuit

    So if I combine both circuits and  I get:

    Is this correct?  If you could direct me to any resources (or ideally examples) that would help me determine component values, or make any suggestions on changing or improving the circuit, I would be very grateful.

    Liam, would the circuit you provided in http://ez.analog.com/message/25531#25531 be able to determine the skin to electrode impedance? Would it require much altering?

    Kind Regards,

    James

  • Hi James,

    I was suggesting a circuit like that,

    Not sure if its clear the circuit....

    Regards,

    Miguel

  • Hi James,

    I dont have too much experience in bioelectric impedance... I just suggested this type of circuit bc  operates around Vdd/2.

    why you will like to operate at Vdd/2?

    Bc the circuit operates in the linear region using a single supply. In your circuit, even using a R/R amplifier, the signals wont operate in the lineal region so you will have distortion due to the components...

    Not sure if is clear my explanation...

    regards,

    Miguel

  • Hi Miguel,

    So you are showing a Howland Current Source, which i guess in detail would look the same as the one i mentioned.

     

    The instrumentation amplifier would amplify the difference between both electrodes, which is understandable. However, wouldn't having VDD/2 be a problem? Surely that wouldn't meet the IEC 60601 standards for it to be safe to measure human skin, as you cannot confirm the current being injected into the user? Do you need VDD/2? You are already injecting current into the skin, why is it necessary to inject VDD/2 into the skin?

    I like the simplicity of the design, but can you clarify the theory behind it? Do you have any references i can read? Why is the circuit not commonly used in bioelectric impedance?

    Many thanks for your help.

    James

  • Hi Miguel,

    I now understand why you suggested vdd/2, however i believe that this circuit would not pass the IEC 60601 safety standards.

    Could the circuit that Liam provided in http://ez.analog.com/message/25531#25531 be adapted to determine the skin to electrode impedance? Would it require much altering? An explanation of how it works would be very helpful.

    Kind Regards,

    James

  • I wouls think the App note that describes "measuring Grounded impedances" would work well,

    You need to make sure that the Return voltage is correctly biased to match the imput circuit of the Ad5933

    The IA does this nicely with the ref cct. You can set it to the right level,

    AC coupling the output will keep the DC bias at zero irrespective of the output selected.

    The only thing left is t use a Isolator such as the 4Ch isolator with built in Pwr reg. I forgot th part.

    This gives 2W and Isolation is good enough for you i think. If you need more you could use  Te ADUM2401 and a max845 plus halo transformer.

    I have used this for medical devices and it will pass regs

  • Hello nicsky,

    Sorry for late reply, but thank you very much for your suggestions. The measuring grounded impedances App note provides a lot of useful information - http://www.analog.com/static/imported-files/application_notes/AN-847.pdf

    Kind Regards,

    James