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.

TINA/Spice/LMH6629: Calculated voltage gain is not as the simulated value for LMH6629

Part Number: LMH6629
Other Parts Discussed in Thread: TINA-TI,

Tool/software: TINA-TI or Spice Models

Hi, 

While designing a noninverting op amp shown in following circuit, I used Rg=13 ohm, Rf=240 ohm, which will should give me a voltage gain of 1+Rf/Rf=19.5. However, from transient simulation, the voltage between input and output only present 116.03/10.69=10.85. This is not match to the calculated value. Are there any places that I did wrong here?

  • Hi

    It seems the signal you put is so high frequency. It seems the signal is 200MHz, and the gain is 19.5. The GBP of LMH6629 is 900MHz.
    In Op Amp design, we have a very simple estimation that GBP > 20*fs*Gain.
    In your design, it could not meet. I suggest you change the signal frequency to 200kHz and do the simulation again.

    Thanks!
  • Hi,

    Yun is right. But there's still another issue: The "COMP" pin of LMH6629 is floating! This furtherly slows down the LMH6629:

    lmh6629.TSC

    Kai

  • Thank for the reply. But I thought LMH6629 has a GBP of 4 Ghz but not 900 Mhz. This is why I thought it may be usable at 200 Mhz.
  • Hi Kai,

    I do not connect com pin because I am using SOT23-5 package. The reality shows me that S21 is somehow -37dB which confused me a lot. That's why I am modeling here. But the modeling also shows me that it is not perform as I expected. But I am still appreciate your feedback.
  • Hello user5365154,

    The comp pin is internally set to high for the SOT23-5 package. For simulation purposes please power this pin. Like Kai mentioned this should improve the bandwidth of your simulations, and provide the estimated GBW of 4GHz that you expected. While your signal will now be within the bandwidth of your device (above the -3dB point), it will still not achieve the expected gain of 19.5 because high frequency of your input. This is shown below:

    Best,

    Hasan Babiker

  • Hi

    I don't think LMH6629 has 4GHz GBP. Even it has 4GHz, refer to table 1 in sloa035. The safety margin is not enough.

  • Hi Yun,

    please look at the header of the table in section 6.5 of datasheet. The specified bandwidths are for a gain of 10 and COMP = "1":

    So, a "small signal bandwidth" of 1000MHz at a gain of 10 gives a "gain bandwidth product" of 10GHz. And a "large signal bandwidth" of 380MHz at a gain of 10 gives a "gain bandwidth product" of 3.8GHz.

    Kai

  • Thank you, Kai. That really makes more sense. However, I did use same parameters for hardware. It seems that bandwidth still not enough as simulated results at 200 mhz. Thus, I decreased Av down to 10 and S21 performs as quite similar for both hardware and simulation.

    In addition, I also would like to measure noise figure for unterminated and terminated cases by using VNA with gain method. However, only terminated case allows me to get reasonable value which is about 7 dB (datasheet specifies 8 dB). But unterminated case, the measured value is 21 dB which is quite far away from simulated results 1.2 dB. I am guess this may due to ports are not matched to VNA resulting in huge reflection. I will appreciate that if anyone has better measurement methods for unmatched case. (PS: I do not have noise figure equipment nor noise source)

  • Hi,

    When you set gain=25.76dB, the -3dB bandwidth could get up to about 315MHz. But at 200MHz point, the gain decreased to 24.34db (16.48V/V). 

    Here is the response when a 200mVpp 200MHz signal input. The gain is about 16V/V.

    From Kai's reply. We know that the 0.1dB bandwidth at gain=10V/V is 190MHz. So you couldn't get a 25.76dB gain at 200MHz frequency。