• State Resolved
  • Replies 2 replies
  • Subscribers 48 subscribers
  • Views 35 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

INA350: Input Resistors Kill Gain in Unipolar Configuration

Evan Cornell
Expert 1400 points
Part Number: INA350
Other Parts Discussed in Thread: TINA-TI, INA351

Tool/software:

Please find attached a small TINA-TI simulation of INA350. When the two input resistors are each 200kOhm, the gain gets killed down to 0dB. When I change to 0ohm, the gain pops back up to the expected 20dB (10V/V) for this model.

Could I please get an explanation for why that is?

The INA350 is for piezo transducer receiver, but 200kOhm resistors are necessary to allow transmitter to operate at several hundred volts. Diodes are there to clamp input pins of INA350 to no more than the diode drops and also bias the inputs to mid-rail VBIAS.

Thanks!

INA350_UNIPOLAR.TSC

  • +1
    TI__Expert 8816 points

    Hey Evan,

    I took a look at the simulation. 

    Good news, I do not think there is a problem with the INA350 model. Bad news, there is there is still a problem. 

    If you remove the INA350 from the schematic, you will still see that the 200k input resistor attenuates the VID (VM1) which would be present at the inputs of the INA:

    0-Ohm:

    200k-Ohm:

    This means that the biasing diodes are likely the culprit. 

    If we do an AC analysis, we can see that you are creating an LPF with the diodes and the source resistance:

    Can you use a different diode, or different bias method?

    Also, if you want to save a little bit on your BOM, we now have the INA351. This is identical to the INA350, just with an integrated reference buffer

    Please let me know if you have any questions,

    Best,

    Jacob

  • 0 in reply to Jacob Nogaj
    Expert 1400 points

    Jacob,

    Much appreciated for the quick response! I will investigate more appropriate diodes for my application. Also, thanks for the tip on INA351, I hadn't seen that part before. Although it's very helpful to have the integrated reference buffer, I suspect I will still need separate op-amp so I have access to the buffered VBIAS to properly bias the inputs to the INA. 

    Thanks,
    Evan