• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 3 answers
  • Subscribers 47 subscribers
  • Views 177 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TRF37D73: Is the output of the TRF37D73 inverted? / Propergation delay

Shinichi Inoue
Guru 21566 points
Part Number: TRF37D73

Tool/software:

Dear Specialists,

My customer is considering TRF37D73 and has two questions.

I would be grateful if you could advise.

---

I am considering TRS37D73 for 1ns pulse amplification,

Could you please answer the question?

(1) Is the output of the TRF37D73 inverted?
When I checked the TINA TI reference design, it was inverted.
Is it correct to understand it is inverted?

TRF37D73.TSC

(2) What are the specifications for the propagation delay of the TRF37D73?
Could you please let me know typical and maximum values

---

 I appreciate your great help in advance.

Best regards,

Shinichi

  • 0
    TI__Intellectual 1030 points

    Hi Shinichi,

    Thanks for your question. The most accurate model for the time delay is going to be the S-parameters, as these are measured. The spice model is going to be a fitted approximation of device performance. I have plotted the phase of the power gain from the input to the output, and you can see there is a 180 degree phase shift at low frequency, which looks like an inversion in the time domain.

    You can take the phase to a time delay by dividing the unwrapped phase value in radians by 2*pi*f. Plotted, this gives the following:

    Looks like there is between 140 pS and 80 pS of delay.

    Let me know if any more questions come up,
    Gideon

  • 0 in reply to Gideon Townsend
    Guru 21566 points

    Hi Gideon,

    Thank you for your reply.

    I understand that the most reliable way to check the delay characteristics is to use S parameter.

    I have some questions.

    Could you please advise?

    ｰｰｰQuestions

    (1) Have you used a pulse step response to check the delay characteristics in the SPICE model?

    (2) The phase characteristics show a phase shift of 180° in the low frequency range.

    I understand that at this time, the operation is like inversion.

    (3) As the frequency increases, the phase shift decreases, and at 5.3GHz the phase shift is 0°.

    Is it correct to understand that at this time, the input and output phases are the same (non-inversion)?

    (3) Furthermore, at a frequency of 6GHz, the phase advances by about 20°.

    Should we consider the phase delay to be 340° when using a single pulse?

    In that case, the delay time is 340/(360×6G)＝157ps. Is this correct?

    ---

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi

  • 0 in reply to Shinichi Inoue
    TI__Intellectual 1030 points

    Hi there Shinichi,

    1. I was just working with the S-parameters. The spice model is a fitted behavioural model, and so to guarantee performance I was using the S-parameters, which are measured.

    2. Yes that is correct, at low frequency, there is a 180 degree phase shift, which looks like an inversion.

    3. Yes that is correct.

    4. Yes good point, since TI has yet to build a non-causal IC haha. So at a phase advance of 20 degrees, then the delay can be thought of as 340 degrees, which at 6 GHz, corresponds to 157 pS. But after you account for the inversion (add 180, not 360), this corresponds to roughly 74 pS. I was incorrect in my original plot, it looks like I subtracted 180 instead of added. The corrected plot is below. Note that I add 180 not 360 because I am accounting for the inversion.

    I also added a pulse input to see the response. The time domain plot is shown below:

    The red curve is the input, and the blue curve is the output. Zoomed in and scaled is on the right. As you can see there is about a 63 pS delay, which more or less matches with our frequency domain plot.

    Thanks,
    Gideon

  • 0 in reply to Gideon Townsend
    Guru 21566 points

    Hi Gideon,

    Thank you for your reply.

    I could understand more deeply the response of the TRS37D73.

    I share with the customer.

    When the customer has an additional question, I consult you again.

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi