• State Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 2 answers
  • Subscribers 49 subscribers
  • Views 8518 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Can I measure 1uA of current with INA138?

jan2809
Mastermind 7850 points
Other Parts Discussed in Thread: INA138, PGA281, INA326, LM7705, TIPD129, TIPD104, INA327, INA333, INA216, INA168, OPA333, ENERGYTRACE

Hi,

I would like to measure a microcontroller current in the range <1uA - 20mA

I build a test circuit based on INA138 (see attached schematic), but the minimum current I can measure is ~100uA.

Did I pick a right device?

Is it achievable?

Jan 

  • 0
    TI__Genius 14825 points

    Jan,

    The INA138 is not really intended to measure Vsense < 10mV.  Most of the specs in the data sheet start at 10mV and go up.  There is a plot in the data sheet on page 4 called "Total Output Error vs Vin" which shows how dramatically the error rises as Vin approaches 0mV, starting at around 10mV to 15mV depending on temperature.  By my calculations, with a 1uA current across a 4.7ohm shunt, you will get 4.7uV.  Even at 100uA, you will get only 470uV, well below 10mV.  In this regard, I don't believe the INA138 is the right current shunt monitor for your application.

    One option to measure currents down to 10uA and up to 100mA is to use http://www.ti.com/tool/TIPD135 which is a circuit designed with a PGA281 to move between ranges and maintain <0.05% error.

    Another option is to use an INA solution.  There is a reference design with simulation that supports 10uA-10mA using an INA326 at http://www.ti.com/tool/tipd104.

    How accurately do you need to measure current below 10uA?

    Thank you,

    Jason Bridgmon

  • 0
    TI__Mastermind 25805 points

    Hello Jan,

    Another TI Design that may interest you is TIPD129 (http://www.ti.com/tool/tipd129).  It is a low-side current sensing solution from 0-1A (requires calibration and the LM7705).  You can adjust the sense resistor for your current range.  I hope this helps.

  • 0 in reply to Pete%20Semig
    Mastermind 7850 points

    Hi,

    Thanks for the responses and links. I should really read the specs more carefully and maybe do the simulation before asking. It is good and easier to see already proven solution.

    I have extra requirements for the project being a very low power consumption and a high-side sensing. I will be happy with a resolution of 0.5uA at lower end and can be higher at higher end (my range is 0-20mA). If unable to cover the dynamic range I might consider using 2 devices with a different gain (or a switchable/programmable gain)

    Do you have any recommendation for a suitable opamp/solution considering the above restrictions?

    Interestingly TI in his Energy Trace Technology build into CCS6 is using another approach, being the measurement of the ripples of switchmode supplying power to the load and there must be a reason for that choice.

    Jan

  • 0 in reply to jan2809
    TI__Mastermind 25805 points

    Hello Jan,

    Given that the supply for your amplifier is 3.2V and your common-mode voltage is 3.0V, the only instrumentation amplifier that could work is the INA326 because it is a RRI/O device. Traditional instrumentation amplifiers will not work because the output swing will be severely limited since your common-mode voltage is so close to your supply voltage.  Therefore I recommend reading TIPD104. However, you will be required to move the shunt to the high-side. TIPD104 has a switch that changes the gain of the device. I recommend switching the gain of the device over the use of multiple shunts or switching shunt resistors. This is because when switching shunt resistors you have to consider the Ron of the switch and the PCB layout can be tricky. Multiple shunts, by definition, mean that you will be driving one of the amplifiers into non-linear operation (and of course you'll have to purchase multiple amplifiers).

    The only drawback given all of your constraints is the power consumption. The INA326's Iq is 3.4mA (max, 25C). If you do not need to measure current constantly, you could consider the INA327 which has an enable pin. When disabled, the INA327's current drops to 5uA.

  • 0 in reply to Pete%20Semig
    Mastermind 7850 points

    Hi Pete,

    Thanks again.

    If I increase the supply voltage to let's say 5V, what would be the recommended opamp for ultra low power?

    Jan

  • 0 in reply to Pete%20Semig
    Mastermind 7850 points

    Hi Pete,

    Thanks for your time and support.

    What can I create?

    Possibly not a ultra low power energy monitor for monitoring very low power devices like WSN nodes.

     

    But, I’m not giving up. If I use a voltage booster to increase the supply voltage to 5V- Vbat will be 5V instead of 3.2V, then I might use

    a conventional opamps to sense current.  Can you recommend a part for low power?

    I will not use much power on efficiency of the booster, but the noise is a concern.

    The switchmode boost converter will be too noisy, but switched capacitor type with some filtering might be OK.

     

    Can you kindly verify if my thinking is correct?

     

    Jan

  • 0 in reply to jan2809
    TI__Mastermind 25805 points

    Hello Jan,

    By increasing the supply to 5V, you could consider using a traditional instrumentation amplifier (INA333 for example). Keep in mind that the output voltage is limited by the input common-mode voltage (see Figure 21 of INA333 data sheet). So, with Vcm=3.0V (and Vref=0V), the output can swing from ~50mV to ~3.8V. However, I recommend designing your solution for an output voltage of 100mV to 3.5V.

    The quiescent current is 75uA (max). I don't know if you consider that 'ultra low power', but it's much less than the INA326 and only 30uA more than the INA138 (your original selection).

    I recommend posting to one of the power E2E forums for the lowest noise 3.2V->5V boost converter.

    I hope this helps.

  • 0 in reply to jan2809
    Mastermind 7850 points

    Thanks Pete,

    I did have a look at INA216 and have a problem how to read datasheet to determine the minimum measurable current.

    Maybe you can help with that as well.

    Jan

  • 0 in reply to jan2809
    TI__Mastermind 25805 points

    Hello Jan,

    The INA216 is a current shunt monitor, which typically have more input bias current than instrumentation amplifiers.  If you're trying to measure 1uA, the INA216 will not work because the input bias current is 3uA.  The INA333, however, has an input bias current of 200pA.

  • 0 in reply to Pete%20Semig
    Mastermind 7850 points

    Hi Pete,

    Thanks, I've learn a lot. Much appreciate your help.

    Do you have a device similar to LTC6102?

    Best regards

    Jan

  • 0 in reply to jan2809
    TI__Mastermind 25805 points

    Hello Jan,

    The closest devices that I'm aware of are the INA138/9 and INA168/9. They are high-side current output devices. However, their input bias current is greater than the LTC device, so they will not work in your application since you're trying to measure 1uA. Note that the LTC device will not work either since the common-mode range is 5V to 105V (see typical application figure on first page of their data sheet). Your common-mode is 3.0V.

  • 0 in reply to Pete%20Semig
    Mastermind 7850 points

    Hi Pete,

    I already decided to use 5V supply for a current shunt monitor circuit, while the DUT supply is 3V.

    If I use one of the conventional opamps with low input bias current I need an extra MOSFET, which otherwise is intergrated inside. Not a big deal, just trying to make it easier.

    The circuit below comes from sbos351d (OPA333)

  • 0 in reply to jan2809
    Prodigy 40 points

    Hi Jan,

        Have you achieved the goal which measures the current ranges from 0 to 20mA? I am doing a similar subject,a battery power management,which has a heavy resistive load about 10M.The input mode voltage is larger than 20V.

        Using OPA333 or other analog components can solve the problem? Or you have had a better solution about this problem already?

    Best regards,

    Hugo

  • 0 in reply to hu shuanglu hu
    Mastermind 7850 points

    Hi Hugo,

    Unfortunately I wasn't able to achieve my goal. If you do not have a restriction of very low power consumption of the power meter itself then you can do it. Have a look at the reference designs from TI mentioned in previous posts.

    See if you can use the new EnergyTrace to monitor your external load. You can use MSP-EXPO430FR5969 (launchpad) and put a micro into very low power mode, negligible compare to external load.

    Good luck

    Jan