This is in relation to the question I asked about deep sleep mode. I'm using PIC18F and it says in the datasheet that in deep sleep it would consume only few nA. But when I configured for deep sleep mode with port set as input, it shows 50uA, which is far more consuming than the limit in data sheet.

It says (under Core Features, page 3):

Power Management with nanoWatt XLP

  • Deep Sleep mode: CPU off, Peripherals off, Currents Down to 13 nA and 850 nA with RTCC
    • Able to wake-up on external triggers, programmable WDT or RTCC alarm
    • Ultra Low-Power Wake-up (ULPWU)
  • Sleep mode: CPU off, Peripherals off, SRAM on, Fast Wake-up, Currents Down to 105 nA Typical
  • Idle: CPU off, Peripherals on, Currents Down to 2.3 μA Typical
  • Run: CPU on, Peripherals on, Currents Down to 6.2 μA Typical
  • Timer1 Oscillator w/RTCC: 1 μA, 32 kHz Typical
  • Watchdog Timer: 0.8 μA, 2V Typical

Why is it consuming much more current than specified in datasheet?

And what all things should I be careful of when using deep sleep mode? I have found one similar question.

I would like to have input from those who have played with deep sleep mode in detail. The controller is PIC18F46J53.

  • \$\begingroup\$ We need more information. Chances are something connected to the PIC is drawing the current. Might even be the programmer. \$\endgroup\$
    – user
    Commented Jun 26, 2014 at 15:14
  • 1
    \$\begingroup\$ An exact circuit diagram and measurement description is the only easy way to solve this. \$\endgroup\$
    – Andy aka
    Commented Jun 26, 2014 at 16:30
  • \$\begingroup\$ Which PIC18F is this, and to what datasheet are you referring for the deep sleep details? \$\endgroup\$
    – JYelton
    Commented Jun 26, 2014 at 17:02
  • 1
    \$\begingroup\$ I have seen similar situations.....Try to see whether you have properly terminated(whether all inputs and Bi directional signals were properly Pulled Up/Down) all the pins. See the quiscent currents of other components also. See for leakage currents. WHen it is specified in the datasheet, yu will get it. we have found problems like this in atmel based board. But the culprits are 2 IO's which were not properly terminate, LDO's quiscent current, OR gate quisccent current. \$\endgroup\$
    – user19579
    Commented Jun 27, 2014 at 5:56
  • 1
    \$\begingroup\$ Generally for all inputs and Bidirectional pins Manufacturers say recommended Pull ups/pull downs to be connected. If any Input or Bidirectional signal is not used. Use pull down/pull up(see for internal Pull ups or pull downs). What other components are there in the board. How the power is dellivered.. \$\endgroup\$
    – user19579
    Commented Jun 27, 2014 at 7:16

1 Answer 1


The datasheet specifies a list of conditions under which those measurements will be valid. These conditions can include a range of things, including supply voltage, clock frequency, temperature, etc. You need to ensure that all of those conditions are met by your setup if you expect to achieve the promised results.

Also, as others have mentioned, make sure you have isolated the PIC in your current measurements. If you have loads attached to any I/O pins or other loads or leakage paths attached to your voltage supply, then those will contribute to your current draw as well.

Finally, make sure that the tool you are using to measure current draw is calibrated and is capable of making the nA measurement.

  • \$\begingroup\$ Thank you @user2635036.I have got my pcb and im testing on that.Initially i have not added any Ic's other than controller.It got connections from pcb to other chips and they were not placed on pcb, when i was testing them.First i got 80nA when in deep sleep mode with no usage of i/o ports,if i/o ports later i got 50ua when i configured the i/o port as input.What would be the ideal mode of testing this circuit.Can i test my consumption with this manner?Why is it showing 50ua? \$\endgroup\$
    – Rookie91
    Commented Jun 27, 2014 at 9:22
  • \$\begingroup\$ I have a resistor of 10k connected to a pin from the 3.3v rail which powers the controller to a pin which i have configured as input.I'll try removing that.If you could find any other traps do let me know.Thank you very much again \$\endgroup\$
    – Rookie91
    Commented Jun 27, 2014 at 9:45
  • 1
    \$\begingroup\$ Surely you have some functional objective beyond minimizing power consumption? We have a saying where I work that goes "test as you fly; fly as you test". After all, if you can't meet your functional requirements, then does it really matter how much power it consumes? So I guess my question back to you is: what is your objective in this activity? \$\endgroup\$
    – kjgregory
    Commented Jun 27, 2014 at 13:36
  • \$\begingroup\$ Thank you very much. I'm using it for entering the pic controller in to deep sleep mode when the normal powering is removed.And i want to use the pic in deep sleep mode that lasts at least one year.This is my functional requirement. \$\endgroup\$
    – Rookie91
    Commented Jun 28, 2014 at 6:17

Your Answer

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

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