I'm trying to develop a cut wire detector that will be used in order to detect if a specific wire is broken or not (a.k.a fake bomb).

I'm using an ATMEGA328P running on batteries (3V) at the internal 8 Mhz clock. The external circuit used to detect the tamper loop, has this configuration:


In order to detect the state changes of the tamper loop, I'm using the ISR INTERRUPTS of the atmega MCU, but with this configuration the resistor draws around 3/4 mA, also, when putting in a higher resistor (R1 = 100kΩ) the power consumption is too high (30/40 μA).

What do I need? I need to find a solution that allows me to monitor if the wire is cut (the wire lenght is around one meter) without using any additional power from the circuit, and keep the circuit running on battery power without the need to worry too much about the power consumption of the tamper monitoring circuit.

The target power consumption for the tamper circuit has to be lower than 100 nA.

Any suggestions?

  • 1
    \$\begingroup\$ What time do you need on the detection? 100ms? 500ms? If it isn't something you need "real-time", consider putting the processor to sleep and waking up every-so-often and check the input rather than use an ISR to wake the processor... \$\endgroup\$
    – Ron Beyer
    Jan 3, 2022 at 16:20
  • 1
    \$\begingroup\$ The detection is not needed in "real-time", i say that some seconds of delay can be okay, i've tried the "watchdog looping" strategy eliminating R1 and switching the GPIO as INPUT (when in deep sleep) and INPUT_PULLUP (when reading from deep-sleep wakeup), i've seen that the overall circuit consumption is much higher (fews mA) than using the ISR_INTERRUPTS and putting in deep-sleep the entire MCU (1.5uA) \$\endgroup\$
    – VirtApp
    Jan 3, 2022 at 16:32
  • 3
    \$\begingroup\$ connect the loop between two GPIO pins \$\endgroup\$
    – jsotola
    Jan 3, 2022 at 16:56

1 Answer 1


You can connect the tamper wire between two microcontroller I/O pins then only wake the microcontroller infrequently to sample the tamper connection between the two.

This answer describes the principle of doing this for a switch with the same low power goals. The actual final average current achieved will depend on your tamper connection sample rate and any other microcontroller activity.

If you tamper connection is prone to ESD, such as by human contact, you may also need to add TVS diodes across the tamper connection ends and low value series resistors to the I/O pins. But that depends on the actual implementation, so I'll leave that to you as you don't detail it in your question.

  • \$\begingroup\$ Thanks for that Tony. About the TVs diodes and resistors in the tamper line, are you sure that they don't produce any current leakage/consumption during work? \$\endgroup\$
    – VirtApp
    Jan 3, 2022 at 18:55
  • \$\begingroup\$ @VirtApp, yes, I'm sure. Can you see them causing a current flow? \$\endgroup\$
    – TonyM
    Jan 3, 2022 at 19:03
  • \$\begingroup\$ Maybe I'm wrong, but I remember that BAT54 has about 0.5-1µA leakage, so they leaks current \$\endgroup\$
    – VirtApp
    Jan 3, 2022 at 20:12
  • \$\begingroup\$ @VirtApp, a BAT54 is not a TVS diode. But anyway, the answer above and the referenced answer describe the principle of a solution. You need to take your original requirements and these principles, research parts as needed and produce your own schematic. You can then plan out your software behaviour and estimate your resultant average current. That's all work for yourself, as the site can't be a free design house or multi-comment personal tutorial. I can clarify the principles detailed above further but not do your final design for you - I'm afraid that's not how the site works. \$\endgroup\$
    – TonyM
    Jan 4, 2022 at 11:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.