• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 48 subscribers
  • Views 1874 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.

analog switch for INA128 offset correction

aubril julien
Prodigy 70 points
Other Parts Discussed in Thread: INA128, ADS1251

Hello,

For my thermocouple acquisition, I use an INA128 (gain=1000) and ADS1251 for A/D conversion. I want to add a switch between thermocouple and INA128, to wire alternatively the ground or the signal to the amp (like the first stage of a chopper amplifier). Unlike the chopper, I don't want to demodulate the signal. I would like to measure (with ADS1251) alternatively the thermocouple signal and the ground. With this, I hope correct numerically the INA128 offset.

Is it reallistic, or just a bad idea ? What switch should I use, or at least what parameters should I pay attention for ??

Thanks in advance,

Regards

  • 0
    TI__Mastermind 44720 points
    Hi Aubril,

    Welcome to the TI E2E Forums!

    I'm not exactly sure what you mean to do from your description. Might you have a schematic you could share?

    Typically to correct for offset error, you would first short the inputs of the INA128 together (and to an acceptable mid-supply common-mode voltage) and measure the offset. Then you would save this result and subtract it from all of the following conversion results (while measuring the thermocouple).

    Other ADC's can perform the offset (and gain) calibration mathematical operations for you. Even though the ADS1251 does not have this function, you can still implement it in a MCU or DSP in your system.

    The particular switch you use is not very critical. The switch RON and the mismatching of the INA128 input bias currents will cause an additional offset error that will not be accounted for in the calibration.

    Best Regards,
    Chris
  • 0 in reply to Christopher Hall
    Prodigy 70 points

    Dear Christopher,

    Maybe this schematic could help ?

    On the first position of the switch (see picture), the output value is V1 = gain*(thermocouple voltage + offset)

    On the second position of the switch, the output value is V2 = gain*(offset)

    So it would be easy to calculate the real thermocouple voltage, (V1-V2)/gain.

    The switch will be set by a microcontroller. So, during time, I want to measure V1, switch, measure V2, switch, measure V1... and so on.

    I already shorted inputs to correct offset. I used the "Offset Trimming" circuit proposed in the INA128 datasheet. But I'm afraid some deviation will occur with time, temperature... So I would prefer to measure offset during measurement.

    The mathematical operations will be performed on an Arduino, a microcontroller based commercial card.

    The additionnal offset is the critical point, I think. If I add more offsetwith switches, my idea is bad. So I searched switches that could imply only very little voltages, but I don't really know how to find it...

    regards,

    Julien

  • 0 in reply to aubril julien
    TI__Mastermind 25865 points

    Hello Julien,

    Shorting the input pins to the mid-supply voltage (ground in this case, assuming your're using a dual, balanced supply) and measuring the output will enable you to cancel out the offset voltage of the INA. I recommend using switches with low on-resistance since they will be in series with your inputs. Note that there will still be some error based on the impedance matching of the resistors and the input offset current of the INA, though it is going to be very small with respect to the initial offset voltage of the device.  I would also place them close to the INA so as to minimize PCB parasitics.

    You may want to consider performing a 2 point calibration by disconnecting the thermocouple and inputting two known voltages. Such a method will also calibrate out the gain error of your circuitry.

  • 0 in reply to Pete%20Semig
    Prodigy 70 points

    Dear Pete and Christopher,

    thank you for your answers.Your help is really appreciated.

    I will test switches with low Ron.

    Julien