ISO122 Defibrillator Isolation

My aim is to measure the leakage current through a patient from different positions while undergoing defibrillation.

We are resistively attenuating the biphasic(bipolar) signal from up to 4000V to a +-10V range for input into the DAQ module yet we require additional safeguarding to protect the DAQ module if the attenuation may fail. We are looking at using an ISO122 for galvanic isolation yet think we may need something like a gas discharge tube or a transient voltage suppressor as well.

The following are my questions:


  • Is it possible to use a series of bi-directional GDT's or TVS's  to shunt the voltage and if so can i ground the excess voltage thought the com terminal?
  • My attenuation will be a simple 1000:1 voltage divider network. Will i need a separate network for  both the input+ and input- terminal?


Attached is a picture of my proposed circuit.



5 Replies

  • Matthew,

    Sorry, but I do not see an attached schematic.

    Based on your description, it seems that you are potentially creating a path for current to flow during defibrilation. This seems counter to the purpose of galvanic isolation, but without a schematic, I'm not sure I fully understand the plan. Also, my understanding is that galvanic isolation in this type of equipment should be capable of withstanding an 8kV open-circuit defibulator pulse. The ISO122 is not rated for this voltage level.

    Please understand that there may be many safefy standards that govern such equipment. We are not experts in these standards and cannot assure that our advice would lead to a design that meets all requirements. Still, you are welcome to try again to attach a schematic and I will do my best to provide advice.

    Regards, Bruce

  • In reply to Bruce Trump:

    Sorry, here is the attachment, this is a rough illustration  of what the setup.

    The purpose of the device is to measure leakage voltage from the defibrillation yet have a high impedance front end so as to not draw the current away from the patient. Newer, biphasic defibrillators have a peak voltage of up to 4kV but isolation amplifiers of that caliber out of by budget. This is why i am suggesting to use some kind of voltage clamping device to dissipate the energy if the resistive attenuation may fail. The TVS or GDT will clamp the voltage at below the isolation rating of the amplifier.

    Under normal operating conditions, the isolation amplifier will only see a voltage range from ~20mV-4V

    I hope this better helps you understand my situation.

    P.S. with regards to safety standards, i have also taken them into consideration.


  • In reply to Matthew OEding:


    The input impedance of the ISO122 is approximately 200k but is not accurately controlled. This would affect your voltage divider and give an uncertain ratio. So far as the protection devices across the barrier, I think it's a potentially dangerous approach. I think it could cause damage to the DAQ circuitry and/or people.

    Regards, Bruce.

  • In reply to Bruce Trump:

    OK, i seem to have been confused between barrier impedance and input resistance. I guess the ISO122 will not be suitable for my application after all...

    The aim was not to have the voltage suppression across the barrier but between the resistive network and the input of the isolation amplifier and have the voltage shunted into the input common terminal of the isolation amplifier. Can the isolation amplifier input common terminal handle this shunt discharge?

    Also, can u recommend any other TI isolation amplifiers that you think may be suitable for this application?

    Your help is much appreciated!

  • In reply to Matthew OEding:


    The issue with input impedance and input protection could be solved with a buffer and additional clamp components, as you proposed. The greater issue in my mind is the isolation voltage rating of the device. Its design and rating is not sufficient for defibulator applications. The AMC1200 is rated somewhat higher but still not sufficient to meet standards (to my understanding).

    If you have questions on the AMC1200, it is supported on our precision data converter forum.

    Regards, Bruce.