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: two questions of INA240

Part Number: INA240

Hi dear supporting team,

here we have two questions on INA240:

1. For what kind of case we will need add the input filter?  there are two conflict description in the d/s(one place suggest adding filter, the other place do not ,  if customer do not add the filter, what's the consequence? 

2.  when customer need detect bidirectional current,  and they want to keep the flexibility for tuning the bias voltage,  they will need use below circuit, while the ADC only accept single ended input, while in d/s it mentioned do not suggest single ended connection.  so how to deal with it?  and it is hard to understand why in fig 27~29, it can have single ended output.   tks!

  • Thank you for using the TI forum, we will get back to you as soon as we can.
  • Hi Vera,

    Thank you for your question on the E2E Foruns.

    1)

    Generally, no input filtering is required or desired on our devices. Filtering reduces the available bandwidth and impacts the slew rate of the device, so you should aim to use it as a last resource. There is internal PWM rejection circuitry for the INA240 and external filtering impacts this to some extent.

    With that said, there are applications where input filtering or external ESD clamps are desirable. Say a very noisy input or situations dealing with ESD discharges or other large transients that would otherwise cause damage to the device.

    TI, gives recommendations on the range of values and layout considerations that can be used for the filter components, with the aim to offer the least possible impact  to the device accuracy. In certain situations it is possible to go outside these values, but given the large amount of applications we cannot test or guarantee outside what is shown on the datasheet. So in short, you should start your application without filtering, but leave the place for filter components, just in case any tweak is required at a later stage. Look at our EVM layout, such as the INA240EVM  for reference.

    2)

    The internal resistors are matched perfectly, and if using these to set a middle point reference voltage, with respect to the supply, the offset values are those tested by TI and shown on the electrical characteristics section of the datasheet.

    When introducing an external voltage divider, internal bias currents will create a shift to the reference voltage, which will cause a change in offset. Furthermore, the internal device resistors, while matched to 0.1%, can have a variation on the absolute value of up to 20%. This adds up to the offset error.

    So how to go on about this?

    One simple approach, is to use a differential measurement, since the voltage across the ADC+ and ADC- connections on figure shown above will always represent the absolute value of the output, despite any shifts caused by tolerances.

    For single ended applications, using a second ADC channel is possible to perform two measurements at the ADC+ and ADC- points. The absolute Vout = {(ADC+) - (ADC-)}

    Where none of the above is possible, a last approach is to reduce, as much as possible, the impedance seen at the inputs REF1 and REF2. This can be accomplished by either using an external voltage reference*, a buffer (such as a voltage follower op-amp) or very low impedance values for R1 and R2 (for example using 220Ohm will result in a lower offset error than 10K, but at the expense of a higher power consumption). Note however that, this approach will still result in a best case offset error up to 20% (+ any error induced by the bias currents and external resistor tolerances) of that shown on the datasheet due to internal process variations, as explained above.


    Will this affect you?

    The default offset for the INA240 is +/-25uV, therefore, adding a 20% error will result in a new offset of +/-30uV. In most cases this will not be an issue, but you should be able to best know based on your application requirements.

    *) It is important to make wise selection when using an absolute external reference voltage. While, in principle, this may seem like an attractive situation, it may actually increase the error. The reason is, most MCU's and ADC's use the supply voltage as a reference and therefore, for the same absolute value of the measured voltage, a variation in VCC will result in a change in the output reading. You should only use an absolute reference voltage in situations where the MCU is also using one.

    I hope this has helped on clarifying your questions.

  • Hi Carlos,

    thank you very much for the detail answer! :)

  • Hi Carlos,

    below are two additional questions:

    1. for pin1, it is NC, but d/s ask connect to GND,  could they float it?

    2. could you help provide the formula on how to calculate the output voltage for below connection?  it seems the output voltage will be related with each resistor, even the input offset voltage of the amp is 0, the output voltage will not be Vref exactly.  tks!

  • Hi Vera,

    1) Not recommended. Pin 1 is used for factory trimming. Noise could theoretically cause the device to misbehave.

    2)The output voltage for the diagram shown below should be VREF +/-25uV. There is no other formula.

    As explained above, where a voltage divider is used, there could be internal currents as you mention. While I am not able to calculate those (only the error) I can tell you these currents will cause the output of the resistive voltage divider to change slightly and this can be measured directly with a voltmeter. An external reference voltage does not have this problem as it can source whatever current is required to keep it's output stable.

  • Hi Carlos,

    tks for the reply!

    so how about below two circuit, do we have the output voltage calculation formula? tks!

  • Hi Vera,

    On circuit, in figure 25, since the reference in at an extreme, in this case ground, the circuit can only measure positive flow currents. These will show as Vout = Vin*Gain (+/-Offset).

    Similarly, for the circuit in figure 25, since the reference in at an extreme, in this case VS, the circuit can only measure negative flow currents, which in this case, will be subtracted from VS. This could be useful to measure the discharge rate of a non rechargeable battery, for example. Vout = {VS-(Vin*Gain)} (+/-Offset).