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.

LM2904-Q1: The threshold for Vos.

Part Number: LM2904-Q1
Other Parts Discussed in Thread: LM2904B-Q1

Hi team,

I use LM2904-Q1 for small signal amplification.During the test, it was found that LM2904-Q1 did not meet the requirements of virtual short.TI's technical support staff told me that it was due to the existence of Vos that this phenomenon occurred.The LM2904-Q1 has a maximum Vos of 7mv.

But I tested LM2904B-Q1 in the same environment,LM2904B-Q1 met the requirements of virtual short.The LM2904B-Q has a maximum Vos of 3mV.

So when Vos is small to a certain value, it will not cause the amplifier to produce a phenomenon that does not meet the virtual short? So what is the threshold for this Vos?

Best Regards,


  • Hello Ivy,

    There is no threshold (off or on) for virtual short error; it is a continuous range that includes both positive and negative error polarities.

    The amount of DC output voltage error is Vos * NG; where NG is the "noise gain"; which is usually the gain of a signal added to IN+ pin. 

    There are other reasons that could make LM2904B-Q1 preforming better; the sink current capability for VOUT < 0.5V is improved.

    I would need more details to know the actual reason. 

  • Hi Ron,

    This is the circuit topology I was testing.The sampling resistor R1024(4mohm) will flow  current of 1A-10A,then I tested LM2904-Q1 and LM2904B-Q1 respectively, and the results are as follows:

    As shown above,

    1.The positive and negative input voltages of the LM2904-Q1 are different, it is not meet the virtual short.

    2.The theoretical magnification of the circuit is 11,but the magnification of the LM2904-Q1 is much smaller.

    I did not have these two problems when I tested with LM2904B-Q1.I have two questions:

    1.What causes this problem in LM2904-Q1?

    2.What parameters can be paid attention to in amplifier selection to avoid this situation? 

  • Ivy,

    1) That's still a virtual short with an input offset error called Vos

    2) Not correct; gain is m in generic linear equation y=mx+b ; see my chart ; all slopes (gain=m) are the same ; only DC offset 'b' changes

    LM2904B-Q1 will do the same thing, given enough tested samples.

    All of these results must be acceptable , otherwise an op amp with much lower Vos range should be used.