Post Go back to editing

Driving analog data convertor AD7823 with an op-amp


I'm interested in using an op-amp to drive an AD7823 data convertor.

Maximum gain required is about ~30, since I can use a ref voltage as low as 1.2V on the AD7823. Required bandwidth operation is 100KHz.

I have no requirements on the op-amp except that the supply range should be between 3v to 5v. However, I do prefer to use a single supply op-amp. I'm not sure whether I should choose an op-amp with a differential output, which according to my understanding, will be more accurate (Will the AD7823 be able to work with it?) or a single ended output with a common ground to the ADC Vin- input?

After reading some application notes I became a little bit confused with all the configurations out there and would appreciate some help...

Thanks in advance,



  • Hi-

    I have moved this question to the ADC Drivers community.  Someone here should be able to assist you.



    EngineerZone Community Manager 

  • Thank you Dan,

    My main concern about using a single ended op-amp is noise or cross talk from other ground loops using the same ground reference point. Are there any differential-output amplifiers you can recommend that will work well with the AD7823?


  • Hello Eyal,

    On the question of using a single-ended vs. a differential-output amplifier, it generally depends on the level of precision you are trying to obtain. Since the AD7823 is an 8-bit ADC, it's LSB size with a  1.2V reference, one-half of an LSB is 2.3mV and its SINAD is 48 dB, so you won’t have any issues finding an amp that meets the noise/precision limits of the ADC.

    Below is a list of op amps that you may want to consider, depending on your total set of application requirements. Since the AD7823 is a low-power ADC, I focused on quoting some low-power op amps for you.

    Please let us know if you have any further questions.

    Best regards,


    Special Features Amps
    Temp Range Vsupply
    CMRR @ Min Gain
    Ib Vos @ Min Gain Rail-to-Rail
    Vnoise Density
    @ Min Gain (RTI)
    0.1-10 Hz
    Bandwidth Slew
    AD8601 1 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -80 200 fA 500 uV Yes 18 nV/rtHz 6 V/us 1 0.36
    AD8602 2 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -80 200 fA 500 uV Yes 18 nV/rtHz 6 V/us 1 0.44
    AD8604 4 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -80 200 fA 500 uV Yes 18 nV/rtHz 6 V/us 1 0.90
    AD8605 1 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -100 200 fA 20 uV Yes 6.5 nV/rtHz 2.3 uV p-p 5 V/us 1 0.68
    AD8606 2 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -100 200 fA 20 uV Yes 6.5 nV/rtHz 2.3 uV p-p 5 V/us 1 1.19
    AD8608 4 Voltage -40 to 125°C 2.7 6 1.2 mA Yes -100 200 fA 20 uV Yes 6.5 nV/rtHz 2.3 uV p-p 5 V/us 1 1.58
    AD8615 1 Voltage -40 to 125°C 2.7 6 1.7 mA Yes -100 200 fA 80 uV Yes 7 nV/rtHz 2.4 uV p-p 12 V/us 1 0.76
    AD8616 2 Voltage -40 to 125°C 2.7 6 2 mA Yes -100 200 fA 23 uV Yes 6 nV/rtHz 2.4 uV p-p 12 V/us 1 1.29
    AD8618 4 Voltage -40 to 125°C 2.7 6 2 mA Yes -100 200 fA 23 uV Yes 6 nV/rtHz 2.4 uV p-p 12 V/us 1 2.29
    AD8657 Micro Power, Precision, RRIO 2 Voltage -40 to 125°C 2.7 18 22 uA Yes -95 20 pA 350 uV Yes 50 nV/rtHz 5 uV p-p 70 mV/us 1 0.95
  • Eyal, I think we need to know a little more about your application, in particular your input signal source. In particular:

    A) What is the signal range (positive and negative) and impedance?

    B) What is the signal referenced to i.e. is it floating, is it ground referenced or does it have a large offset.

    The large LSB step size of an 8 bit converter should help you, so you need to focus on ensuring that the error sources for the amplifier are managed.

  • First of all, thanks for all the responses.

    I'm interested in amplifying a positive signal (between 0-50mv) from a similar photoreceiver to this one and would like to be able to utilize as much as I can of the AD7823 resolution with only a single stage amplifier.

    The signal I am interested to amplify is the photoreceiver's output which I believe is referenced to its internal ground (It is a double ended output). Unlike the device linked above, my device has a frequency bandwidth of 100KHz, but a similar output impedance of 16ohm.

    The AD8601 seems very appropriate from the frequency point view, GBWP of 8MHz. A simulation also shows that a non inverting single supply configuration with a gain of 10 (500mv = utilizing ~45% of the AD7823 resolution) works reasonably well.

    I originally had 2 concerns:

    1. At a simple non inverting gain configuration of 20 (1v = utilizing 83% of the AD7823 resolution), the  
        amplifier was further away from the ground rail, 9mv @input of 0mv. I know that this is the downside of
        single supply and will apply a positive offset voltage to the input of the op-amp together with a digital offset at the digital receiver which
        interfaces with the ADC.

    2. I am concerned that noise or cross talk from the controller of the ADC on the ground plane will interfere
       with the analog input to the op-amp if I will connect the reference of the photoreceiver to the common
       ground. A possible solution maybe using a digital and an analog ground planes connected by a trace on
       the PCB. Or, as I mentioned before, using a differential output op-amp, separating the analog signal from
       the common ground.

    Thanks again for trying to help,

    I would appreciate any technical advice and will be glad to provide more information if required.


    Message was edited by: Eyal Niv

  • Eyal,

    I agree with Oliver. Can you tell us more about the input? 

    We can certainly recommend a differential receiver amp like the AD8129/30 but you may not need that.

    (Oliver: Thanks for your advice)


  • I would look at the AD8031, MAX4124 or MAX4125, LT6233, MAX4488, ADA4851 or the LT6202. They will all comfortable provide x20 amplification at 100kHz with at least a factor of 10 excess loop gain.

    Instead of making all your signals 0V referenced you could generate a virtual ground reference somewhere between 0.1V and 0.5V above 0V and then configure the front end amp as a differential amplifer rather than a simple non-inverting amplifier. This would raise the opamp output up but a small amount, bypassing the problem of not swinging right down to ground. The ADC negative input would obviously connect to the virtual ground.

    Using a differential amplifier arrangement also provides an easy way of removing ground noise between different parts of your system.

  • Thank you Oliver!

    Right now I have this setup, where I use an offset to solve the voltage swing problem:

    This works well in simulation. However, if understand correctly, in this setup I am not generating a virtual ground/rail splitter and as such do not have a ground noise clean reference point for the AD7823.

    Thanks a lot for mentioning this. I'll do some reading on the differential op-amp configuration, which I am less familiar with, and would really appreciate if you can elaborate little bit more.



  • This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin