• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 49 subscribers
  • Views 1561 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

INA219 with input filtering

Laszlo Nemeti
Prodigy 85 points
Other Parts Discussed in Thread: INA219EVM, INA219, INA226, INA230, INA220, INA231

I have a circuit with INA 219. I used the INA219EVM to create my schematic. Rshunt=0.01, R1=R2=10, C1=100n, default config register value. I have no load current.

If i read the bus voltage register I get correct results, but reading the shunt voltage reg I get strange values.

With 0V bus voltage -> shunt 'voltage' (with the values above directly represents current in mA)  0mA (result: OK)

With 5V bus voltage -> shunt 'voltage ~15mA (it should be 0mA with no load)

With 10V bus voltage -> shunt 'voltage' ~30mA (it should be 0mA with no load)

Because I was not satisfied the results, I shorted R1 and R2 and then i got good results (0mA). (Before that I checked R1 and R2: both have 10.4 ohm resistance, so they are mached)

Reading the INA219 datasheet I found that INA219 has NOT the same input inpedance:

Vin+= 20uA; Vin- =20uA || 320k. This additional 320k generates current  which current generates voltage on R2(10) which voltage generates differential voltage on INA219 input. At 5V bus voltage this voltage diff is ~156uV that represents 15.6mA current. At 10V@312.5uV@31.25mA. Naturally with R1=R2=0 this diff voltage is 0.

Is my succession correct? If i dont want to remove the input filter resistors(filtering+ESD functions) how can I eliminate this problem? I will try to add 320k parallel with R1 too.

Thanks in advance.

Laszlo.

  • 0
    TI__Genius 14825 points

    Laszlo,


    You are correct, the external filter caps mixed with the VIN+/VIN- imbalance does create a voltage drop across the shunt.  Adding 320k to the VIN+ side works in simulation to neutralize the imbalance though I don't know how well it will work in real life.

    There are other devices, the INA220, INA226, INA230 and INA231 all have a separate VBUS pin for measurement that doesn't imbalance the VIN+/VIN- inputs.  You might consider changing to one of those devices, or eliminating the filter.

    I set up a TINA simulation to play with this, modeling the inputs, and it validates what you observed and calculated.  Please find it attached.

    4527.Vin_neg_impedance.TSC

    Jason Bridgmon

  • 0 in reply to Jason Bridgmon
    Prodigy 85 points

    Hi Jason,

    Thank you for your answer. I have played with your simulation file. 

    I will try to add a 320k to the VIN+ side and share my experience about how well it work in real life.

  • 0 in reply to Laszlo Nemeti
    Prodigy 85 points

    Here are my experience:

    1. Adding 460k:

    current error decrased by ~3 (example 0Vin:-2mA->-2mA; 5Vin:15mA->5mA; 10Vin:32mA ->10mA)

    2. Adding 330k (I could not find 320k):

    0Vin:-2mA->-2mA; 5Vin:15mA->-3mA; 10Vin:32mA ->0mA; 16V:51mA->8mA

    It seemed good under 10V, but above 1x V the current started to increase.

    3. Replacing 10 ohm filter resistors with 0 ohms:

    0Vin:-2mA->-2mA; 5Vin:15mA->-1mA; 10Vin: 32mA ->-1mA; 16V: 51mA->0mA; 18Vin: 58mA->0mA

    So adding a ~330k resistor significantly decreses the error, but finally I ended up replacing the 10ohm filter resistors with 0ohm resistors.