AD7768 : 8 CHANNEL ADC CAN SUPPORT SINGLE END - BIPOLAR INPUT.

Hi All ,

I have an requirement to convert the 8 channel analog input (Single ended - bipolar) signal with the range of 10Vpp, The waveform will be sin, ramp, square,and Sweep.

Hence I am planing to use AD7768. In this I need a clarification of following points:

1. I have 8 channel single ended bipolar signal inputs (10Vpp, +/- 5 ), So I will connect this signal to the correspoding port of (AIN0+ to AIN7). My question is, May I Ground it the remaining ports of AIN-0 to Ain7- ?

2. My input signal should support (+5V to -5V , 10Vpp), so in REF1- and REF2- may I connect to Ground or -2.5V should give ?

3. If I ground the REF1- and REF2- , The AD7786 support the input signal swing 10Vpp?

4. Even though My Signal is single ended bipolar , Is there I want to make this signal to converted into differential bipolar (True Biploar signal)?

Please clarify the points as soon we are going to finalise the design stage.

Parents
  • Hi NiallM,

    Thank you very much for your inputs !!!

    I am going with AVDDx, REF+x = +2.5V  using  ADP7118 ; and AVSSx , REF-x = -2.5V using  ADP7183ACPZN2.5-R7

    Is this LDO Current is enough for AD7768 Supply ?

    I have an rearranged architecture of my 8 Channel Receiver card.

    Before the architecture is :

     

    Re-arranged architecture :

    Can you give me your clarification of following parts.

    1.        AD826: I am going to attenuate my signal ranges from 10Vpp to 5Vpp (+/-5V to +/-2.5V ). By selecting RF /Rin with inverted output (Example:  RF = 1K ; Rin 2K = -0.5 Gain). This Output is fed to the input of another channel of  AD826. Hence I will get +/-2.5V attenuated signal.

            Is this right way of attenuate my signal range ? or do you have any recommended part for this design?

    1.         ADA4922-1 This attenuated signal is fed to the ADA4922-1 to convert the single ended bipolar signal to   differential bipolar signal and fed to the ADC (AD7768) input.

    This single ended to differential end conversion is there right part to do the conversion instead of you recommended LTC6363.

     

    Please confirm may I go ahead with this part ?

     

    Thanks and Regards,

    Mohamed Sidiq.H

  • 0
    •  Analog Employees 
    on Sep 7, 2018 2:32 PM over 2 years ago in reply to Sidiq

    Mohamed,

    I just stumbled across this post, and wanted make some comments on the signal chain you have posted on August 21.

    You show the ADA4922-1 driving the AD7768 - The ADA4922 will cause performance degradation of the AD7768 at the maximum sample rate.  As the AD7768 samples the output of the ADA4922, it injects a small amount of charge onto that node at the beginning of each acquisition.  The amplifier that drives that node needs to be able to recover from that disruption quickly, and drive that node back to the proper voltage.

    Niall is recommending the LTC6363 or the ADA4940 b/c they are capable of driving the AD7768 at it's full sample rate.

    The ADA4922 *might* be able to drive the AD7768 when using a much lower fmod than the maximum value.  Here's a snippet regarding fmod from the datasheet:

    One of the specs that's critical in this application is GBWP:

    • LTC6363: GBWP = 500MHz
    • ADA4940: GBWP = 130MHz
    • ADA4922: GBWP = 40MHz

    Faster is not necessarily better, but for this ADC, 40MHz GBWP is probably too low for the driver to fully settle between samples.

    On the other hand, if you're willing to use a lower fmod, the ADA4922 might work fine.

    -Anne

Reply
  • 0
    •  Analog Employees 
    on Sep 7, 2018 2:32 PM over 2 years ago in reply to Sidiq

    Mohamed,

    I just stumbled across this post, and wanted make some comments on the signal chain you have posted on August 21.

    You show the ADA4922-1 driving the AD7768 - The ADA4922 will cause performance degradation of the AD7768 at the maximum sample rate.  As the AD7768 samples the output of the ADA4922, it injects a small amount of charge onto that node at the beginning of each acquisition.  The amplifier that drives that node needs to be able to recover from that disruption quickly, and drive that node back to the proper voltage.

    Niall is recommending the LTC6363 or the ADA4940 b/c they are capable of driving the AD7768 at it's full sample rate.

    The ADA4922 *might* be able to drive the AD7768 when using a much lower fmod than the maximum value.  Here's a snippet regarding fmod from the datasheet:

    One of the specs that's critical in this application is GBWP:

    • LTC6363: GBWP = 500MHz
    • ADA4940: GBWP = 130MHz
    • ADA4922: GBWP = 40MHz

    Faster is not necessarily better, but for this ADC, 40MHz GBWP is probably too low for the driver to fully settle between samples.

    On the other hand, if you're willing to use a lower fmod, the ADA4922 might work fine.

    -Anne

Children
  • Hi Anne,

    Greeting for the day !!!

    I made a mistake in the block-diagram , In that my input is differential 10 Vpp signal which is fed to the PGA AD8253 .

    The PGA converts the differential to single end convertion with selectable gain of 20db, 40db, 60db.

    After this PGA I am going to remove high frequency components and allowable range of 0 -30khz Band and remove other range of signals. Hence the Output of PGA will be 20Vpp.

    Due to the Input Constraints of AD7768 (AD7768 – allow only Differential or pseudo- differential Input with maximum allowable Common mode range is 5Vpp).

    hence I want to attenuate this signal 20-Vpp to 10-Vpp and Then ,I want to convert this single ended bipolar signal into differential bipolar as 5Vpp to match the AD7768.

    Can you suggest any other part is available to attenuate and convert single end to differential conversion chip parts ?

    Mr. Niall Suggest to use LTC6363 , In this Common mode input range is 8.8V as Max if 10V supply used. and also maximum power supply is 11V.

    In ADA4940 , Power supply max is 7V and Common mode input will be −VS − 0.2 to +VS − 1.2.

    Due to this constrain I couldn't able to move further,

    In AD4922 have 26V as max Supply and Input common mode range ±10.7 , so that I asked your suggestion may I use this part .

    Please let me know if you have any suggested part for attenuate and single end to differential end conversion.

    Kindle let me know if you need any clarification regarding my inputs.

    Thanks and Regards,

    Mohamed Sidiq.H

  • 0
    •  Analog Employees 
    on Sep 12, 2018 5:07 PM over 2 years ago in reply to Sidiq

    Mohamed,

    Yes - Looking at differential drivers, it looks like there's a limited selection for higher supply voltages.  

    If you are willing to slow down the sample rate of the AD7768, then the ADA4922 will settle fine.  You may run into performance issues at the fastest sample rate.

  • Hi ,

    I am looking for 24 bit ADC. Then I saw this post.

    I can see many things happing in signal conditioning before ADC. I am curious and want to know the Signal conditioning and ADC that is coming for this Post.

    1. ADC7768- Since in the post, we are Differential input in the range of  ± 1.25/2.5V.

    a. It means we are attenuating the signal. So is this the right way to do condition monitoring?

    b. Can we use this ADC for  ±5V range?

    c. Can we use AD7176-2 in place of AD7768 for ±5V range (http://www.analog.com/en/design-center/reference-designs/hardware-reference-design/circuits-from-the-lab/cn0310.html)?

    In Signal conditioning-

    2. We are Passing Differential signal to AD8253 (PGA) that is converting differential signal into Signal ended. Then We are using LPF- LT1697IS (Dual/Quad Low Noise, Rail-to-Rail, Precision Op Amps-) But it an Op-amp. 

    So how can we use this as an LPF? Can we not use differential LPF?

    then We are attenuating the signal using Attenuator (LT1678)- But As per Mohhamad suggested that why not we can go for LT6363 (It can take differential Input (±5V range), work as an LPF and generate Differential Output). So if we use this IC (LT6363) we don't need to add LPF, attenuator, and Single-ended to a differential ended amplifier like- LT1944/ADA4922-1.

    So As we are making the DAQ system we should use minimum component for better performance. Kindly suggest and if I missing any point for the DAQ system. Do we have any other Differential ADC to full fill above criteria? 

  • Hi All,

    I changed my Old architecture which I posted earlier,

    The updated design architecture I mentioned in below:

    Block Diagram:

    Please let me know your comments if anything to be update in the schematic.RECEIVER_CARD_CH0_WITH_ADC7768_1.pdf

    Hi kapiliitd,

    1. It means we are attenuating the signal. So is this the right way to do condition monitoring?

    Mohamed: Now, I am not attenuate the whole range as per the (10 Vpp Differential Input from external world which I mentioned in the Block diagram) . The range I reducing to 5vpp with differential bipolar to matching the ADC differential input.

    2.Can we use this ADC for  ±5V range?

    Mohamed: NO, ADC maximum range of supply AVDD1 to AVSS = 5.5V as max. if you use ±5V as supply, AD7768 will damage due to exceeding the maximum range. Instead of using ±5V you may be us ±2.5V Supply. Which will  be the range of AVDD1 = +2.5V - AVSS = -2.5V = (+2.5 – (- 2.5V )) = 5V.

    3.Can we use AD7176-2 in place of AD7768 for ±5V range 

    Mohamed: I am looking for 8 Channel ADC  which have sampling rate more  than 125Khz , Simultaneous sampling.

    Hence AD7176-2  wont match my requirement.

    Thanks and Regards,

    Mohamed Sidiq.H

  • Hello Mr. Mohammed,

    Thanks for the reply and sharing the details for Analog chain.

    As you replied that we can not use the AD7768 with Input voltage range from +5V to - 5V. But in the datasheet, it has given that We can use the ADC with  +5V to - 5V range. Please see figure 40 and figure 41. 

    In the Figure, it is clearly mentioned that We can use the ADC with +5V to - 5V range. But the Integral Non-linearity (INL) will increase.  Since as per the requirement of Sampling frequency we can go for the median mode.(Fs>108kHz). So please suggest.