• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 64 subscribers
  • Views 346 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.

REF102: Designing with REF102

Shibin Varghese
Intellectual 835 points
Part Number: REF102

Hello Team,

We are planning to use the IC REF102 for a 1mA precise current generation.
The drift of the current source must be below 100nA.
The same current is used as the excitation current of PT1000 and PT100 temperature sensors.
Can we achieve the 100nA precision with the IC?

Initially, we were using a Howland current source, and the drift and the noise from the Howland current pump were very high since it has many components, and all those components contribute to the temperature drift and noise.
Can we achieve the 100nA precision with the IC?

Also, we need a 10uA current as the excitation current of some other sensors like the diode.


The R in the above circuit can be varied using an analog switch.
But the analog switch will also have resistance, so we need to adjust the voltage to the 4th pin of the REF102 to get the correct 1mA or 10uA (We also need calibration to nullify the offset due to aging).
So we are planning to use a DAC to inject a voltage to the 4th pin pf the REF102 IC.
Can you please comment on this?
Is there any good approach other than this to get a precise current source within our tolerance limit?

  • 0
    TI__Expert 5350 points

    Hello,

    REF102 should have no issue staying within 100nA precision. I would like to further understand your application so I can accurately comment on your design description.

    Can you provide more information or a schematic on how you plan on using an analog switch to vary R?

    Additionally, what resistance values are you considering? Do you plan on varying the reference current by changing R? 

    Are temperature drift and noise the most important specifications that you are considering in this design?

    Thanks,

    Jackson

  • 0 in reply to Jackson Wightman
    Intellectual 835 points

    Hello Jackson,
    Thank you for your reply.
    I have attached a block diagram explaining my requirements.
    Hope this will be enough for you.


    We need 3 sets of current sources.
    1mA, 100uA and 10uA.
    For 1 mA (+/-0.1uA), 10uA (+/-1nA)& 100uA (10nA)

    The resistors to the mux are selected in such a way that
    For 1mA,
    R=10V/(1mA) = 10K

    For 100uA,
    R=10V/(100uA) = 100K

    For 10uA,
    R=10V/(10uA) = 1000K

    The OPAMP, resistors, and the reference IC may contribute drift in the current generated so we are planning to use a DAC (for voltage injection) to calibrate the current source over time.

    Can you please comment on this?
    The value for the current sense resistor can be up to 2K ohms (connected to the current sense amplifier INA190A3IDCKR ) for 10uA current.


    For 10uA current, we need a precision of 1nA.
    Is it achievable with this circuit?


    Are temperature drift and noise the most important specifications that you are considering in this design?
    Yes, we need a clean supply with low noise as possible and almost no drift.


  • 0 in reply to Shibin Varghese
    TI__Expert 5350 points

    Hello,

    The highest initial accuracy of REF102 would be REF102C, which is ±0.025%. Not including long-term drift, temperature drift, resistor tolerances, or drift from the amplifier, this would give a precision of 2.5nA. This is higher than the 1nA specification. The DAC could be used to compensate for some of this. 

    Is the current sense amp and the ADC being used to validate the current source design? If so, what is the application of the precision current source? What is giving the 1nA precision requirement?

    If low noise and low temperature drift are the most important, we do have lower noise and lower temperature drift options available, and I can assist and selecting the best voltage reference for your design, depending on the application. 

    Thanks,

    Jackson

  • 0 in reply to Jackson Wightman
    Intellectual 835 points

    Hello Jackson,

    Thank you for your reply.

    Ask I mentioned earlier we need a precise current source as the excitation current source for sensors like PT100 and PT1000.

    Any change in the excitation current is going to affect the temparature measured.

    The current sense amp and ADC are used to validate the current source.

    The output of the DAC is adjusted after a period to get the same current output.

    The current sense amp is used to measure the current so as to adjust the DAC value.

    Low noise, low drift are more critical in our application.

    Can you please suggest some voltage ref ICS 

  • 0 in reply to Shibin Varghese
    TI__Expert 5350 points

    Understood. In this case, wanting to vary the source current by varying R, the configuration with REF102 is most likely the one of the best accuracy specs you will be able to get from a voltage reference IC. I don't see any issues with your block diagram. 

    If you have any other questions please let me know.

  • 0 in reply to Jackson Wightman
    Intellectual 835 points

    Hello Jackson,
    Thank you for your reply.
    From the datasheet of REF102,


    The output varies drift, and input voltage.

    Is there any other reference IC or reference circuit with extremely low drift and low noise so that we can use it for very accurate current generation?
    Also, what value is to be added to the DAC to get compensation for the current?

  • 0 in reply to Shibin Varghese
    TI__Expert 5350 points

    For this application, I would suggest the REF102. If you want to explore other options, our precision voltage reference portfolio can be found here:

    https://www.ti.com/power-management/voltage-reference/series-voltage-reference/overview.html

    How you approach calibration will depend on a few factors that are specific to your application and device, such as temperature range, voltage reference output, and other factors. More information on how to calculate the error and calibrate accordingly specific to your system can be found here:

    https://www.ti.com/lit/an/snaa320a/snaa320a.pdf

    Sections 4 and 5 are particularly helpful in this case.

    All the best,

    Jackson