• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 47 subscribers
  • Views 1806 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

INA826: INA826 circuit measurement unstable

Fantastic
Prodigy 165 points
Part Number: INA826
Other Parts Discussed in Thread: TINA-TI

I got problem when using INA826 on LED current measurement as below.

It is a simple LED current measurement circuit. There are two LED strings with 20mA current go through both strings. The current translated to voltage by 0.1R resistors and there's a MUX followed by the current sense. INA826 is used to amplify the small current difference. The ground of LED strings are different from the measurement ground. 

The experiment setup as below:

1. Set C173 & C176 to NC

2. Setup MUX(U53) to connect BL2_IS+ & BL2_IS- pair to INA826 input.

3. When BL1 LED string is on, the output of INA826 is normal (~1.8V because no current sense pass through BL2 string)

4. When BL2 LED string is on after BL1 LED string on, the output of INA826 is normal (~1.86V) for a couple second and then go down to 1.8V (No current sensed).

5. When both BL1 & BL2 string on at same time, the results is same as 3.

6. After changing the C173 & C176 to 100nF, the readout becomes stable without degradation after both strings on.

Here's the question:

Why the INA826 output becomes abnormal after a couple seconds when C173, C176 is NC. Is it likely to be over-voltage protected? 

Thank you!

Brian

  • 0
    TI__Guru 68521 points

    Brian,
    What you describe has all the signs of settling time issue. Placing 2.2uF cap between the INA826 inputs together with 1k input resistors creates a long time constant which may take a while to settle and adding 100nF caps stabilize the circuit. Also, you indicate that the ground of LED strings are different from the measurement ground; do you mean the supply ground for INA826 is different than the input common-mode (LED) ground and if so what is the potential difference between the two grounds? Have you used the INA826 Vcm vs Vout Tina-TI model available under TI website to make sure that you still operate within the valid linear input voltage range?

  • 0 in reply to Marek Lis
    Prodigy 165 points

    Dear Marek:


    I've taken off the CAP and the result is the same. The ground level is different between LED and INA826. I tried to put 100kohm to replace C173, C176 and

    it is solved. I doubt that is because the Ibias current need a return path. But even if I did not do the replacement, C173, C176 placed with 100nF (Must be greater than 100nF) can solve this as well. Here I attach the waveform of experiment as below.

    1. LED OFF->ON with INA826 (C173:NC, C176:NC, C175:NC) -> Measurement current down to zero after couple of second

    2. LED OFF->ON with INA826 (C173:100NF, C176:100NF, C175:NC) -> Problem solved

    3. LED OFF->ON with INA826 (C173:100Kohm, C176:100Kohm, C175:NC) -> Problem solved

    Brian