• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 47 subscribers
  • Views 512 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AMC1350: TVS Recommendation for AMC1350 Input Protection with Ultra-Low Leakage

Jorge Ribera
Prodigy 10 points

Part Number: AMC1350
Other Parts Discussed in Thread: TVS0500

Tool/software:

Dear TI Technical Support,

My name is Jorge, and I am a research engineer in motor diagnostics at the Universitat Politècnica de València.

We are currently designing a high-voltage measurement circuit using the AMC1350, and we are looking for a suitable TVS diode to protect the input of the device against overvoltages in the sensed mains voltage.

The system is designed to measure up to 1000V RMS, using a resistive voltage divider with a current of approximately 0.7 mA, which ensures that the dominant measurement error comes from the resistor tolerances, not the input characteristics of the AMC1350.

According to the datasheet, the operating input voltage range of the AMC1350 is ±5V, while the absolute maximum rating is ±15V. Therefore, we need a TVS solution that:

  • Does not clamp within the normal input range (±5V),

  • Clamps safely below ±15V to avoid exceeding the absolute maximum ratings,

  • Has ultra-low leakage current (preferably ≤ 1 μA) to minimize additional measurement error.

We initially considered the SMAJ5.0CA, but its specified leakage current of up to 20 μA is too high for our accuracy requirements.

Could you recommend a suitable TVS diode (or alternative protection strategy) that fulfills these constraints and is compatible with the AMC1350 in this application?

Any suggestions or references to application notes would be greatly appreciated.

Best regards,
Jorge Ribera
Research Engineer – Motor Diagnostics
Universitat Politècnica de València

  • 0
    TI__Mastermind 22200 points

    Hi Jorge,

    Are you trying to clamp a DC overvoltage or a transient overvoltage?

    TVS diodes are spec'd for transient conditions such as ESD/Surge with pulse widths in the nS-mS range. Even a large TVS diode like SMAJ5.0CA will be damaged if constantly conducting in breakdown if there is a large current flow. 

    If you want to clamp a DC overvoltage, you will need a zener diode or a FET based clamp that can sustain constant conduction. 

    Regards,

    Sebastian

  • 0 in reply to Sebastian Muriel
    Prodigy 10 points

    Hi Sebastian,

    Thank you for your reply.

    We are indeed looking to protect the AMC1350 against transient overvoltages, specifically to comply with Category III insulation requirements. For a 1000 V RMS system, this implies the ability to withstand an 8 kV surge (1.2/50 μs waveform) according to applicable standards.

    We use a resistive voltage divider at the AMC1350 input, designed to absorb 1 mA of current under normal operation. This divider reduces the mains voltage (1414 V peak) to around 4.5 V, safely within the AMC1350's input range of ±5 V.

    We initially considered using the SMAJ5.0A/CA for surge protection. However, its leakage current is too high for our accuracy requirements:

    • 0.02 mA at 25 °C

    • 0.05 mA at 85 °C

    This leakage current is variable with temperature and voltage, making it unpredictable. Since it flows into the same node as the voltage divider, it introduces an error of 2–5% relative to the 1 mA divider current. This is unacceptable for our application, where we aim for less than 1% total error using 0.1% precision resistors.

    In summary, we are looking for a protection component that meets the following requirements:

    • Clamps the AMC1350 input to less than ±15 V during a transient surge of 8 kV / 1.2-50 μs

    • Draws no more than 0.005 mA (5 μA) of leakage current under normal conditions (≤5 V)

    • Has stable leakage characteristics over temperature, to preserve the accuracy of voltage measurement

    Could you please recommend a component (TVS, zener, or other) or strategy that fulfills these criteria?

    Thank you once again for your support.

    Best regards,
    Jorge Ribera

  • 0 in reply to Jorge Ribera
    TI__Mastermind 22200 points

    Hi Jorge,

    Thanks for providing all the details.

    This is a very interesting requirement, an 8kV surge will require a very robust protection device. If the 8kV surge is through a 2Ohm network, a chain of a few protection devices in parallel would need to be implemented. 

    These more robust protection devices usually come with the trade off of leakage currents in the uA range as you have seen with the SMAJ5.0A/CA.  

    Would this be an 8kV surge through a 2Ohm or 42Ohm network? 

    If the requirement is an 8kV surge through a 42Ohm network, a possible solution could be 5 TVS0500 in parallel.

    If the requirement is an 8kV surge through a 2Ohm network, I don't really have any solid recommendations that won't violate your leakage current requirements. 

    I couldn't find any reference designs that implement 8kV protection, but maybe this app note is a good starting point: Lightning Surge Considerations for PoE Powered Devices.

    Could you possibly share the standards that are driving the Category III insulation requirements? 

    Regards,

    Sebastian