0
\$\begingroup\$

I am planning to use a differential amplifier, LTC6409 which is reported to have a GBW of 10GHz. I ran an AC analysis on LTSpice using the Spice model given here. I connected the negative input to ground and the positive input to an AC source. The result shows that the open loop gain reaches 0dB at around 2GHz. The open loop gain at 10GHz is around -27dB.

As far as I understand, I can add feedback and have a flat close loop gain of 0dB up to 2GHz, but it is not possible to have a closed loop gain of more than 0dB beyond 2GHz.

Now I thought of three possibilities

  • My simulation or my understanding of the concept of GBW is incorrect.
  • The Spice model of the op-amp is incorrect
  • The reported GBW is incorrect (I hope not!)

Can anyone please help me understand this? Any suggestion for other differential amplifiers with GBW of at least 5GHz would be great.

\$\endgroup\$
7
  • 4
    \$\begingroup\$ If LT says it has a GBW of 10GHz, and they advertise it as such, it will have a GBW of 10GHz. Why are you not using the LTSpice model of the device in LTSpice? According to the LT website, the model should be in LTSpice already, no need for external models. \$\endgroup\$
    – Joren Vaes
    Commented May 16, 2017 at 20:38
  • \$\begingroup\$ Look at the datasheet conditions to get the 10GHz- It's in the spec table. \$\endgroup\$
    – John D
    Commented May 16, 2017 at 21:00
  • \$\begingroup\$ What is your actual requirement? Not in terms of GBW but what gain up to what frequency. There is a closed loop gain vs frequency graph in the data sheet and they confirm what you said in that at unity gain the gain is only maintained up to 2GHz. This amplifier does have 10GHz GBW at a gain of 100. \$\endgroup\$
    – Kevin White
    Commented May 16, 2017 at 21:06
  • \$\begingroup\$ From Datasheet Page.7, it appears that this IC can't have any meaningful use above 1GHz with gain 1 - 5, and much less if the gain is more than 10-15 dB \$\endgroup\$
    – Ale..chenski
    Commented May 16, 2017 at 21:28
  • \$\begingroup\$ I did not know that the model is in ltspice. Got the same result with that model, but this is gonna be useful in future. Thanks! My understanding was that I would get a unity gain at 10GHz, which turned to be incorrect. Thank you all for the insight. I am looking for an op-amp with reasonable gain around at around 5GHz. Does something like that actually exist? LTC6409 is the opamp I found with highest GBW. \$\endgroup\$
    – Siam
    Commented May 16, 2017 at 21:41

1 Answer 1

Reset to default
2
\$\begingroup\$

I simulated this and got 2G too. Read the paragraphs under applications information in the datasheet. Some specsmanship going on here. This kind of sums it up; enter image description here

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.