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.

INA240: Unbalanced midsupply voltage

Part Number: INA240

Hello,

I created a current sensing circuit with INA240A1D recently. The electronic circuit is below.

I set the output to be midsupply voltage, but the wrong voltage is outputted.
Vcc = 5.0 [V]
(IN+) - (IN-) = 0.0[mV]
Out = 1.53 [V]

Before this phenomenon occurs, Out was the correct value, 2.5 [V] with 0.0 [mV] (IN+) - (IN-). Does this unbalanced midsupply voltage often occur as one of the failure modes? Do you come up with main causes?

  • The image did not attach; please ensure to use the "Insert Media" button. See [FAQ] How to insert an image in your forum post.

  • Dear user,

    As discussed in section 8.4.3 of the datasheet, the reference pins internal to the INA240 are resistive in nature, and require a low impedance drive to achieve accuracy, but 1V certainly seems pretty far off the mark. 

    To begin, could you depopulate C12 and C10 and see if there is any improvement? I am curious if the capacitors are interfering with the value of the network here.

    Second, how are you achieving a VSENSE = 0V configuration here? Can you share a bit more about your input structure (size of shunt, planned current, etc.?)

  • Dear Carolus,

    Thank you for your response.

    I removed C12 and C10 like below, but there was no change.



    Vcc = 5.0 [V]
    (IN+) - (IN-) = 0.0[mV]
    Out = 1.54 [V]

    Regarding how to satisfy VSENSE=0.0 [mV], input circuit is below. OUT1 and OUT2 of U1 don't output voltage in this test. R6 is 150 [mOhm] shunt resistor.



    I suspect INA240A1D has some failure.

  • Dear Carolus,

    Thank you for your response.

    I removed C12 and C10 like below, but there was no change.



    Vcc = 5.0 [V]
    (IN+) - (IN-) = 0.0[mV]
    Out = 1.54 [V]

    Regarding how to satisfy VSENSE=0.0 [mV], input circuit is below. OUT1 and OUT2 of U1 don't output voltage in this test. R6 is 150 [mOhm] shunt resistor.



    I suspect INA240A1D has some failure.

  • When U1 is not outputting a voltage, then very probably the input bias currents of INA240 cannot properly flow. This would result in a malfunction of INA240.

    It's impossible to give further help with so sparse information. What is U1? And what is the load connected to the shunt?

    Also, the solder joints look terrible. There's a risk of short-circuits or unwanted leakage currents. Remove the excessive solder, clean the solder pads with denaturated alcohol and solder again :-)

    Kai

  • Dear user, 

    I agree with Kai regarding the bias currents. Typically for a test case where VSENSE = 0V, the device needs to be maintained with bias currents allowed to enter the pins of the device. If the device isn't biased properly, there could be issues. Have you probed the shunt resistor to see if any voltage is present on the shunt? 

    Below is a typical test setup used for a VSENSE = 0V condition. A voltage from the same supply is fed to both inputs of the device, effectively making this a VCM of the chosen voltage, and Vdiff of 0V. VMID here is simply mid-supply as you have it configured. 

    Is it possible to perform such a test on your board? I would recommend isolating U1 for the purposes of this test, but this would help us understand what is going on on your board. As Kai also mentioned, it would be useful to understand what U1 is. I'm assuming some form of converter?

    Also agree that there could be issues stemming from flux residue left on the board. I would recommending cleaning with ispropyl alcohol to ensure that any flux leakage paths are removed the board. Reflow of solder looks like it could be useful as well. 

  • U1 is TA7291P, a full bridge driver IC for DC motor control. It switches the motor power voltage (+12V) for OUT1 and OUT2 depending on IN1 and IN2. In my previous test case,
    +12V: about 0 [V] (Since this power is off)

    and keep CONN6 open. Regarding more details about TA7291P, please see https://www.mouser.com/ProductDetail/Toshiba/TA7291PO?qs=rG6OSLlFOsrQGpHmzDlyqg%3D%3D.
    Also, I probed the shunt resistor and found 0.0 [mV].



    As for the new test, I will isolate U1 as much as possible. But I can't remove U3 or cut the pattern to U3. "Out" of INA240 is connected to ADC input pin fo PIC16F1459. I programmed it to use the pin as AN3. Please see https://www.microchip.com/wwwproducts/en/PIC16F1459 for details. The test circuit is below.

  • Dear user,

    You say the 12V source is off, but is there any voltage present on the OUT1 line? 

    Also, another test I thought of: is it possible to isolate and drive the REF pins with a voltage directly? This could help us see if there are any impedance problems between the sources and reference causing this issue. 

  • When +12V power is off,
    OUT1 (=IN+): 0.0 [mV]


    I performed the test you proposed after I remove flux residue. Please see the below schematics.

    Be careful with the below:

    - I disconnected "out" of INA240 - "RA4/AN3" of U1 connection.
    - REF1 pin is connected C2 in Vcc (5V) power supply circuit. Impedance between C2 and REF1 is only PCB pattern.

    I probed INA240 pins:
    IN1 - GND: 13.04 [V]
    IN2 - GND: 13.04 [V]
    out - GND: 1.53 [V]
    REF1 - GND: 4.98 [V]
    REF2 - GND: 0.0 [mV]

    The voltage, out - GND is still not half of REF1, so I believe INA240 failure cause it.

  • Do you have any additional INA240 units? This could be easily confirmed by changing out the device and seeing if this alleviates the issue. If it would help, DM me and I can see about getting some additional samples to you to verify this?

  • No. I don't have any additional samples. I'd like to get a sample, but I don't know how to DM you.

  • Dear user,

    I've sent you a friend request to see if we can arrange this for you.