• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 163 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.

INA296AEVM: Different step responses with different REF voltage setups

Nico Kassner
Prodigy 30 points

Part Number: INA296AEVM

Tool/software:

Hi everyone,

I am currently using the INA296AEVM to measure LED current (0-40mA). @Common mode voltage (VLED) ~4.76V. For this purpose, I have added a small measuring shunt (R1 = 0.3 Ohms) at the high side. Now I would like to measure and amplify the voltage drop using the INA296A5 (factor 200). The low side is controlled by a Melexis uC, f = 500 Hz. I did the following measurements with a PicoScope 3000 Series.

As described in the User Guide under 4.1 Quick start setup, i connected VS = 5V and REF1=REF2=GND. To achieve this, I placed  J6_5 and J5_5 connects REF2 to GND. Unfortunately, when I measure with this setup, I do not get the step response I expected. The output voltage takes approx. 100 us to reach the desired final value. Changing to a different gain factor has made this even worse.

After that i tried setting REF2=GND and REF1=VS. Here I see a perfect step response. Within <1 us it reaches the expected final value. Very similar to what is shown in Figure 6-35 Large Step Response inside the data sheet.

Unfortunately, I have no explanation for this and hope to get an understanding or solution in this forum.

BR Nico


EDIT: I tried to apply several voltages while shorting REF1 and REF2 so VCC2=REF1=REF2. Somehow VCC2 >= 7mV is enough that the step response is looking as expected. everything below that is looking similar to first scope.

EDITEDIT: I set VCC2 = 3,3V and swapped IN+ and IN-. So I'm measuring negative current now. This will work for my application. But nevertheless: I'm curious what the technical background is to my problem. 

  • 0
    TI__Mastermind 33460 points

    Hello Nico,

    I believe this is the recovery from saturation.  If you have a step response from zero voltage at the output it will be delayed due to the output being saturated. I have a few post on this linked below.

     INA296AEVM: Low duty operation 

     INA296A: High side current sensing in the H-bridge and recirculation 

    Regards,

    Javier

  • 0 in reply to Javier Contreras4
    Prodigy 30 points

    Hello Javier, thanks for the quick response. I also found these posts before creating my own. It seemed somehow the same, but i wanted to be sure.
    Is this behaviour documented in the datasheet or is it general knowledge for this kind of sensors? As i said, i found a solution for my case but what is the application then for REF1=REF2=GND? Either having constant current flowing or being fine with the slower step repsonse?

    And a follow-up question: what is the current draw via the Ref pins? or is it nearly zero at REF1=REF2=VCC2? I'm already familiar with input bias current, but this is reffered to the input and not the Ref pins..

    BR Nico

  • +1 in reply to Nico Kassner
    TI__Mastermind 33460 points

    Nico,

    Thanks for your question and viewing the other post.  I understand getting confirmation.

    Nico Kassner said:
    Is this behaviour documented in the datasheet or is it general knowledge for this kind of sensors?

    This behavior is normally on amplifiers in general.  Some just exhibit more extreme than others. Our two stages may also exaggerate the condition.

    Nico Kassner said:
    As i said, i found a solution for my case but what is the application then for REF1=REF2=GND?

    If you have a negative supply you can connect it to the GND pin.  You can treat it as a dual supply as long as you do not go over 20V for the device from the Vs pin and GND pin.  The drawback of this is having the ability for the output to go negative (below GND).  In transients this easily happen and may damage components downstream if they cannot handle that input.  Another option is placing a small voltage on REF pins and calibrate this as an offset.  

    Nico Kassner said:
    Either having constant current flowing or being fine with the slower step repsonse?

    The two options above are the easiest to implement without adding additional error.

    Nico Kassner said:
    what is the current draw via the Ref pins? or is it nearly zero at REF1=REF2=VCC2? I'm already familiar with input bias current, but this is reffered to the input and not the Ref pins..

    There is still in input bias current even when REF1=REF2.  It is about 20µA and does vary about +/-10µA with the change of Vsense votlage.  

    Regards,

    Javier