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.

LMV651: Calculate the output error lead by input bias current

Part Number: LMV651
Other Parts Discussed in Thread: TLV9061, TINA-TI, TLV9062

Hi Team

As shortage reason I am looking at LMV651 for TLV9061 replacement. However, LMV651 input bias current is much higher than TLV9061.

Here is customer schematic:

It's a very classic differential current sensing circuit, so the positive and negative input impedance is match as 2.5K//10K=2Kohm

I want to know the output error lead by input bias current should be

  1. Ibias*2Kohm*Gain=80nA*2Kohm*4=640uV
  2. Ios*2Kohm*Gain=2.2nA*2Kohm*4=17.6uV

Which one is correct for this schematic? 

Second things I want to konw, As LMV651 is non rail to rail device, below cirucit that bias to 1.65V won't exceed the LMV651 input common mode rang with 3.3V power supply.

Am I right?

  • Hey Gabriel,

    The error from your input bias would be calculated with Ibias, not Ios. Ibias is the current flowing into the inverting or non-inverting terminal needed to bias the circuit. Ios is the difference between these two values (how well these two input pairs match).

    And for your second question I would reference the datasheet for the common mode voltage range at 3V operation. 1.65V Common mode is within the range of valid common modes for 3V operation, so you should be safe in 3.3V operation.

    Best,
    Jerry

  • Jerry

    Thanks for your comments.

    I still not understand why we need to select Ibias rather than Ios to calculate this error.

    I know bias current will generate offset voltage in postive pin: Vos_bias=2Kohm*80nA=640uV

    However, we also expect the negative pin will also have 640uV offset voltage. 

    I am expecting these two offset voltage will eliminate in such structure. Please help me understand why I am wrong. Thanks.

  • Gabriel,

    If you would like, input bias current is something that is simulated in our models, and you may test this there!

    The error presents itself as a DC shift downward.

    I was unsure what the value of the input current and the shunt resistor value was, so I placed those arbitrarily, but the input bias voltage effects should not change based on that. I am attaching the simulations as well in TINA-TI, which you may download here: https://www.ti.com/tool/TINA-TI 

    tlv9062_ib_error.TSCLMV651_ib_error.TSC

    The circuit:

    LMV651:

    TLV9062:

    Best,
    Jerry

  • Hi Jerry

    Thanks for your follow up and suggestion. I understand the input bias current won't be a problem or key factor in this specific circuit no matter with TLV9062 or LMV651. But I just don't understand why matching the negative input and positive input impedance can't eliminate the input bias current affect.

    I have read some APP NOTE indicate we should match the input impedance to minmize the input bias current affect.

    As difference amplifier input impedance is naturally matched it should be also work in here. The voltage shift in negative and positive input caused by input bias current should be counteract and only left the input offset current section base on my understanding.

  • Hi Gabriel,

    the effect of input bias currents and input offset current can very easily be simulated. Just take an ideal OPAmp and two current sources:

    gabriel_lmv651.TSC

    In this simulation a shunt resistance was assumed, which is negligibly small compared to the feedback resistors of OPAmp.

    Kai

  • Gabriel,

    The transfer function will assume an ideal op amp with no input bias current, however the input bias current on both inputs generates essentially a downward DC shift. The input offset current will change how closely IN+ matches IN-, but the input bias current shifts both input voltages down, which will shift the output down.

    This is a DC offset, but will behave differently over temperature, so it may be difficult to calibrate out.

    Best,
    Jerry