• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 49 subscribers
  • Views 437 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.

LMH6554: DC Coupled Driver for ADS42JB69 ADC?

Trent Allison
Prodigy 20 points
Part Number: LMH6554
Other Parts Discussed in Thread: ADS42JB69,

Hello,

We purchased a DC-coupled, single-ended ADC card with a ADS42JB69 ADC that is being driven by a LMH6554.  The driver circuit follows the typical application schematic on the LMH6554 datasheet, minus the AC coupling caps.  When we apply a square pulse we get droop on the top of the pulse and undershoot when the pulse ends, as if something is AC coupled.  Can the amplifier be used in this DC-coupled configuration?  If not, is there a pin-compatible DC-coupled option?

Thanks,

Trent

  

  • 0
    TI__Mastermind 19486 points

    Hello Trent,

    I found a TI document providing a bit more clarity on using the LMH6554 in DC-coupled configuration.  It may provide additional justification or perspective on using the LMH6554 in a DC-coupled configuration.  In the meantime, I will look over your circuit and think about what could cause the droop & undershoot.

    https://www.ti.com/lit/an/snoa565a/snoa565a.pdf

    Best,

    Alec

  • 0
    Guru 48395 points

    Hey Trent, you show TINA file, but the pulse looks measured? What is out there after the series output R's - is that the ADC and if so what is the full network into it. I opened the schematic for that EVM, looks to be all transformer interface, 

  • 0 in reply to Michael Steffes
    Prodigy 20 points

    Hi Alec and Michael,

    Thank you for the app note and looking into this for us. The pulse is measured data at 100 Msps, we had a square pulse (1 usec, ~400 mV) into a scope (from an op-amp circuit with a 49.9 output resistor) then plugged it into the ADC card and got the plot. The ADC card we purchased came with a data sheet that provides the ADC buffer circuit. I took that circuit and put it in TINA (the figure posted) and it simulates as expected. From what I can tell, the output from the TINA circuit goes directly into the ADC and the 1.9 V VCM comes directly from the ADC (the schematic only shows from the input at the dashed line to the output after the two 49.9 ohm resistors, and that the VCM is an input with a 0.1 uF bypass cap, there also seems to be plenty of bypassing on the LMH power). There do not seem to be any transformers.

    I also posted this question under the ADC forum and they sent be app note SBAA220 May 2016 for the ADC.  It shows overshoot and undershoot due to a gain=0.05 path with a high pass in parallel with the gain=0.95 main path. The plots in the app note do not show the droop that we are seeing. Do you think the 0.05 gain path with the high pass explains the behavior we are seeing in the blue plot from before? We will probably try adding a digital filter to compensate for the high pass (as mentioned in the ADC app note) but I doubt that will remove the droop on the top of the pulse.

    Thanks again for the help,

    Trent

      

  • 0 in reply to Trent Allison
    TI__Mastermind 19486 points

    Hello Trent,

    I will take a look at your post and the described behavior; I appreciate you providing this extra detail and description.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Prodigy 20 points

    Hi Alec,

    We put in a longer pulse and the response looks like the plot in the ADC app note. So, I think we have tracked the AC-coupled response we are seeing down to the internal G = 0.05 high-pass filter path inside the ADC. We tried the digital compensation filter in the ADC app note and that seems to help. Looks like the amplifier circuit is working fine in this DC coupled configuration but we need a different ADC, unless we have good luck optimizing the compensation filter. Thanks again for looking at the circuit for us.

    Trent