• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 50 subscribers
  • Views 989 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.

THS3217: Query on High speed DAC-ADC Opamp circuit design : THS3217, THS4541

Shyamsundar Hegde
Prodigy 120 points
Part Number: THS3217
Other Parts Discussed in Thread: DAC3484, THS4541, ADS42LB69, TINA-TI, DAC38J84, THS4509

Hi All,

I've been working on analog link based on high speed DAC-ADC

My DAC board has : connecting differential outputs from DAC3484 to Op-amp THS3217  which converts differential to single ended o/p, and this single ended o/p is terminated on the connector.

ADC board has: connecting single ended output from DAC board to THS4541 which converts single ended signal to differential and in turn this differential o/p is connected to 16-bit ADC inputs.

We used 50 Ohm RF coaxial cable to connect single ended o/p from DAC to ADC board.

Please suggest , are we using correct Op-amp parts for signal conversion in the design? any design concerns/considerations?

Attached the scheme here.

Thanks&Regards,

Shyamsundar Hegde

  • 0
    TI__Genius 13935 points
    Hello Shyamsundar,

    What is your signal bandwidth? What is your desired amplitude at the ADC? Are you using a reconstruction filter on the DAC? Are you using an anti alias filter on the ADC?

    Is it possible to use a differential transmission line instead of 50 Ohm single ended?

    Regards,
    Loren
  • 0 in reply to Loren Siebert 1
    Prodigy 120 points
    Thanks Loren,

    Signal bandwidth is upto 100MHz, Desired amplitude at ADC is 2Vp-p, As of now we are not using reconstruction filter and anti-aliasing filters any concerns on this? Currently we are considering ADC ADS42LB69, and exploring more higher speed ADCs like JESD204B standard.

    We cannot run the differential signals form the DAC board to ADC board because of the impedance stuff as well as cabling(i.e number of cables). That is the reason we are converting DAC differential signals to single ended on the DAC Board, connecting this to ADC board through 50 Ohm RF coaxial cable and on the ADC board again converting it to differential.

    I'm tryng to simulate the same on TINA-TI . Looking forward for your valuable suggestions.

    Thanks&Regards,
    Shyamsundar Hegde
  • 0 in reply to Shyamsundar Hegde
    Prodigy 120 points
    Hi Loren,

    Waiting for your reply.

    Thanks&Regards,
    Shyamsundar Hegde
  • 0 in reply to Shyamsundar Hegde
    TI__Genius 12631 points

    Hello Shyam,

     Lets start first at the transmit side (DAC + THS3217) and  then make our way to the receive side (ADC).

    - What is the desired output amplitude and signal frequency from the Tx side?

    - Please check out the following reference design which has some test data and circuits for a DAC38J84 + THS3217. This should help you in optimizing your design quickly.

    - Are you planning on achieving a certain distortion/IMD target with this design? If so what are your desired targets?

    -Samir

  • 0 in reply to Samir Cherian
    Prodigy 120 points
    Thanks Samir,

    - Desired output amplitude from TX side(i.e DAC + THS3217) is 2Vp-p, and signal frequency is 100MHz

    We are trying to achieve 7 bit encoding in this scheme. Please suggest

    Thanks&Regards,
    Shyamsundar
  • 0 in reply to Shyamsundar Hegde
    TI__Genius 12631 points
    Hello Shyam,
    You should be able to get a 2Vpp, 100 MHz signal from the Tx side with no issue. I would add some filtering between the DAC and D2S and also between the D2S and OPS stage in order to minimize Harmonic Distortion. I would configure the OPS in a gain of 2V/V to 2.5V/V. Lower the gain the better for distortion performance but I really wouldn't go to much below 2V/V.

    For 7-bits, the SNR should = 6.02*7+1.76= 44 dB. I would add some guard band and say we need 50dB performance.

    I haven't looked at the DACs performance at 100 MHz so please check that.

    For the THS3217, the HD2 (Figure3 of datasheet) performance for a 2Vpp signal at 100MHz is 50dB and the HD3 (Figure 4 of the datasheet) is much better so I wouldn't worry about that. The noise is around 8nV/rtHz which output referred in a gain of say 4 becomes 32nV/rTHz. In a bandwidth of 100MHz the total noise is 32*SQRT(100MHz) = 320uV (rms). For a 2Vpp signal, 60dB of noise performance is 2mV (rms) so noise isn't an issue either. Bottom line I think we are good on the Tx side.

    Do you expect a lot of external interference between the Tx and Rx? How much will this degrade the signal?

    On the Rx side, I would recommend again checking the ADC to ensure it has sufficient performance.

    If you now use a 50ohm system between Tx and Rx you are only going to have a 1Vpp signal at the input to the THS4541. I will assume that you will be configuring the THS4541 in a gain of 2 which means its output will be a 2Vpp signal. Looking at Figure 13 of the THS4541 datasheet, the HD performance at 50MHz is around 50dB which doesn't leave much room for other errors. I would recommend using a faster amplifier. Look at the THS4509. It will be sufficient to achieve your end performance needs.

    There will be some performance degradation because when driving an ADC the amplifier usually drives a capacitive load which degrades the HD performance. However, the base performance of the THS4509 is good enough that this shouldn't be a problem.

    Also, like Loren mentioned use anti-aliasing filters to minimize out of band perturbations. Look at Figure 86 of the THS4509 datasheet. That circuit will act as an anti-aliasing filter. Its not much of a charge-bucket filter but it should meet your needs. The design can be tweaked later by changing component values, if needed.

    -Samir