• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 378 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.

ADS7961: Problem about driving ADS7961

DeWei Lan
Prodigy 250 points
Part Number: ADS7961
Other Parts Discussed in Thread: OPA172, OPA192, OPA810, OPA320, OPA325

I read ADS7961 datasheet which mentioned input capacitor is 15pF.

My first problem is that 15pF.

Is the 15pF MXO input or AINP?

<br>

Beside, my ADC driver is non-inverting amplifier.

I'm considering about overshoot problem because ADC is capacitive load.

I hope:

* Direct drive ADC without Riso. That means direct short MXO and AINP, but MXO input will be drived by OPA172 or OPA192

* Overshoot < 10Vp

I read OPA172 datasheet

and OPA192 datasheet

I think select OPA192 can direct drive ADS7961, is right?

<br>

My circuit:

Photodiode -> OPA810(I/V Converter) -> OPA172(Integrator) -> OPA192(ADC Driver) -> ADS7961(ADC)

  • 0
    TI__Mastermind 23498 points

    DeWei,

    1. The details on the ESD structure, MUX, and sample-and-hold parasitics are given in figure 62. Reviewing this figure, you can see that the 15pF specification includes all the capacitances.
    2. Do you plan on placing the amplifier between MXO and AINP?
    3. For multiplexed systems, the mux switching will create one transient, and the sample and hold action will create another transient. It is necessary to fully charge the mux parasitic capacitance, as well as the sample and hold capacitance to achieve good accuracy.   If these capacitances are not fully charge you will get what is referred to as a “settling error”.  Using a slow sampling rate allows more time to settle.
    4. The ADS7961 is an 8 bit converter. This means that the LSB = FSR/2^8 = 5V/2^8 = 20mV.  This is the highest accuracy you can achieve with this device.  Normally, settling to 20mV is relatively easy unless your source impedance is quite large.
    5. The amplifiers you mentioned in your question are designed for +/-15V supply rails. Most of these devices can be run at lower supply rails but they are not designed for this purpose.  The input range of the ADS7961 is at most 0V to 5V.  Applying voltage levels beyond this will damage the device.  Normally, a 5V single supply amplifier is used to drive this type of device.  A very common SAR drive amplifier is OPA320.  In case you need a wide supply amplifier for your sensor, you can use a level translation circuit to adjust the high voltage to a safe low voltage. See Noninverting circuit for high-to-low voltage level translation to drive ADC
    6. To directly address your question: normally when an amplifier is used to drive a SAR converter, an external RC filter is used. This helps to achieve better settling and lower noise.  In your case this is probably not necessary as the resolution is quite low (8 bit).  Note that the internal capacitances are quite low and they have an internal resistance in series with the capacitance so you really do not need to worry about overshoot and stability issues (you could consider the switch impedances as a built in Riso).
    7. The process of selecting an amplifier, external filter, and sampling rate is covered in sar-adc-front-end-component-selection-series

    Best regards,

    Art

  • 0 in reply to Art Kay
    Prodigy 250 points

    Many thanks  detailing reply:

    My fully circuit block diagram as here

    (1) Many thanks! I found this useful picture

    (2) Yes, I'd place OPA192 between MXO and AINP as ADC driver

    (3) Actually, >=10KHz sample on per channels is enough for my usage.

        If sample rate is as low as 10KHz, should I need care about “settling error”?

    (4) Integrator also drags detected pulse(width = 10us ~ 500us) to 1 ~ 10 ms

    (5) My system adopt +5V single rail PSU(USB Vbus as PSU).

         Our refer to here

    (6) OPA192 and OPA172 are remaining components in my last project, but OPA320 looking like a good choice if I have not enough  

         OPA192 and OPA172

  • +1 in reply to DeWei Lan
    TI__Mastermind 23498 points

    DeWei,

    1. Thanks for the block diagram. This is helpful.
    2. At 10kHz sampling rate, and 8 bit resolution you will not need to worry about settling.
    3. I would replace the OPA172 and OPA192 with OPA320 or OPA325. The main reason for this selection is to use devices that were optimized for 5V rail.  The OPA172 and OPA192 are great devices, but they are optimized for wider supplies.  The OPA172 for example, has a common mode range limitation on the positive supply of 2V.  So, on a 5V supply an it’s input range is limited to the range of 0V to 3V (see Vcm specification).  You may be able to make these amplifiers work for your application but the OPA320/325 are rail-to-rail input and output with zero crossover distortion on a 5V supply.  Furthermore, they are very commonly use for SAR drive, so this type of application has been tested many times with OPA320.
    4. If you feel that you want to re-use your old components, the OPA192 has rail-to-rail input and output performance, so you might focus on that choice. If you can afford it, I would suggest the opa320 or 325.

    Best regards,

    Art

  • 0 in reply to Art Kay
    Prodigy 250 points

    Oh! My bad!

    I discover Rail-to-Rail of OPA172 only guarantee output instead of both input and output.

    So I'd not adopt OPA172.

    The reason for selecting OPA172 is also because of the cost.

    But for the cost reason, I think OPA325 is a good choice, because OPA325 not only cheap but also provide Rail-to-Rail both I/O.

    I'll also check our stock if OPA192 is enough for replacing OPA172.

    Anyway, thank you very much for your very useful suggestion.

  • 0 in reply to DeWei Lan
    TI__Mastermind 23498 points

    Happy to help!