• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 48 subscribers
  • Views 1174 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.

INA196: 4-20mA Receiver - Having Accuracy Problems

Matt Roybal
Intellectual 345 points
Part Number: INA196
Other Parts Discussed in Thread: INA190

I have come across several designs that use the INA196 in a 4-20mA receiver, using a 4.99 ohm resistor and an ADC to read in the voltage. We created a circuit that has 5x of these on the board. Everything works for the most part, the problem I am having is the accuracy between each port seems to be off from one another. 

For instance we are transmitting a 8mA signal, and we are reading in 8.05mA on 2 of the ports, and 1 of the ports reads 8.20mA. They seem to be inconsistent between one another.

I am using a 0.1% tolerance resistor, so this is highly accurate. We are also using a 12-bit ADC. I expected them to at least be consistent with one another, which would allow me to adjust our equation of converting voltage to mA. The voltage for the current loop is 24VDC.

Is there something I am missing here? Are we supposed to be calibrating/trimming each individual port to get the accuracy we want? I tried using the "Current sense amplifier error analysis INA196", but it won't allow me to enter in 4.99 ohms for the resistor, which is common with the INA196.

Thanks and any help is greatly appreciated.

  • 0
    TI__Mastermind 28910 points

    Hello Matt,

    There could be two issues that you are dealing with.  I will start with the cause of the most error.  This would be the offset voltage.  The input offset specification for the INA196 is 2mV.  So for 8mA across the 4.99Ω would give you a voltage of 39.92mV.  The error form the 2mV offset gives range of 37.82mV- 41.92mV.  So only from the error of VOS you could see 8mA measure 7.599mA - 8.401mA.  

    The other error source that is not as large is the input bias current.  When Vsense = 0v the input bias current is about 23uA and this current is not always the same for both input pins.

    For a more complete error source for your application please take a look at this video, but the two above I think will be your biggest concern.

    The reason you did not see the input resistor value input able to go high is that normally current sense amplifiers are meant to measure higher currents.  We do have newer current sense amplifiers that are better equipped with lower IB and lower input offset voltages such as the INA190.  The INA190 has tradeoffs in lower bandwidth and lower input common voltage range but lower offset and lower input bias current.

    The INA293B is a has a higher common mode voltage range but cannot go down to -16 common mode voltage.  The INA293B has the same pinout and much better offset than that of the INA196 but still has about the same IB.

  • 0 in reply to Javier Contreras4
    Intellectual 345 points

    Hi Javier, thanks a lot for that info, very helpful!

    So with these errors, do you think we can overcome them by a calibration process? E.g. are these errors somewhat constant and consistent in which we can calibrate the reading at 4mA, and at 20mA, and this would improve our accuracy?

    I am surprised so many designs use the INA196 in a 4-20mA receiver. It seems unsuitable for this, and it seems there are much better options. But if we could at least get by with a calibration step we will probably be ok.

    Thanks again for the help and I look forward to your opinion on this.

  • 0 in reply to Matt Roybal
    TI__Mastermind 28910 points

    Matt,

    This error can be reduced by doing a calibration process.  If offset is the biggest error source, I believe it is, you can measure at 4mA and calculate the offset of the system.  That may be enough but the other options is better.

    If you do a two point measurement of 4mA and 20mA then you can calibrate out both the offset and the gain error of the system.  The device should be linear for most of the parameters and you would eliminate most of the errors.  The thing you would need to think about is drift or any other changes to the system from the calibration point.

    The INA196 is an older device that may have met the common mode voltage requirements at the time.  This is possibly why the INA196 was used.