LMG1020: Recommendation for amplifier to interface with LMG1020 input

Hi Eric,

I looked, and the ones you found seemed to be some of the best options to me also. I think you will have better luck asking this question to the amplifiers team rather than my team (gate drivers).

The method I am most used to seeing for small pulses, is to generate a larger pulse (say 50ns), then shorten it with an AND gate + RC delay. This avoids the issue of needing high gain and bandwidth amplification, although a part that can do this may exist also. I will move this thread to the team responsible for THS4303, and let them comment.

Thanks,

Alex M.

Hi Alex,

  Thank you for your recommendation and forward!

Hello Eric,

   For a 1ns pulse with no gain, you would need a unity gain bandwidth of at least 350MHz. Looking at the LMG1020 datasheet, the snippet below is I believe the voltage threshold discussed that is required for 5V supply: 

   Most likely will need a gain of at least 10V/V which will require an amplifier specified on the higher end of gain bandwidth products. I would recommend our decompensated amplifiers which require a high gain to be stable, but will give you more bandwidth than a unity-gain stable amplifier. 

   Below is the OPA858:

 And below is the OPA855:

  Let me know if above will work for you application in terms of any other design requirements. These are both very high speed amplifiers which would mean careful PCB layout for optimal performance and avoiding issues with circuit stability. We recommend following the device EVM layout for best performance. 

Thank you,
Sima 

Thanks Sima, I'll give the OPA855/858 a look and see if they will work for my application.  I suspect that they may not be as effective as the THS4302 due to their lower gain-bandwidth product, would you agree with that?