Loading ...
Sorry, an error occurred while loading the content.

Re: [softrock40] OpAmp question on GBW

Expand Messages
  • Kees & Sandy
    Thanks for the response, Milt. I think that s basically what I was saying...... The GBW is the max frequency for a gain of 1, you can also max out at that
    Message 1 of 10 , Jul 1, 2007
    • 0 Attachment

      Thanks for the response, Milt. I think that's basically what I was saying...... The GBW is the max frequency for a gain of 1, you can also max out at that frequency/100 if your gain is 100, or at that frequency/200 if your gain is 200. If you are at the max, further trying to increase the gain by changing Rfb/Rg does nothing at that frequency ? or does it start automatically trading off frequency vs amplification (loose the Rfb/Rg ratio formula for gain) to keep the GBW which is constant. Or does it just start distorting the amplified waveform. 

      73 Kees K5BCQ

      -- Milt Cram <w8nue@...> wrote:

      Kees & Sandy wrote:

      >
      > As you approach the specified OpAmp Gain Bandwidth, how does the OpAmp
      > react ? Rfb/Rg no longer specifies the OpAmp gain (less) ? The OpAmp
      > becomes less responsive to higher frequencies (and maybe undesired
      > noise) ?
      >
      > 73 Kees K5BCQ
      >
      >
      >
      Kees,

      The OpAmp Gain Bandwidth is the frequency at which the amplifier gain
      drops to unity. No combination of Rfb/Rg resistors can make the gain
      any higher at this frequency. Generally the OpAmp gain falls at
      6dB/octave above the "corner frequency". The slope of the gain curve at
      unity gain will be approximately -6dB/octave, but may be somewhat
      different from that depending on the design of the amplifier. The phase
      shift will also be nominally 90 degrees, but may be different, again
      depending on design (which includes internal compensation to make the
      amplifier stable at unity gain. Some amplifiers are designed which are
      not stable at unity gain, and frequently spec a minimum gain for
      stability (e.g. x2 or x10).

      Milt, W8NUE

    • Tony Langdon
      ... You can t exceed the GBW, it s a physical limitation of the op amp being considered. In fact, as you approach the unity gain frequency, the op amp is no
      Message 2 of 10 , Jul 1, 2007
      • 0 Attachment
        At 12:49 PM 7/2/2007, you wrote:

        >Thanks for the response Bill. I was just wondering what happens if
        >you exceed the GBW ....say you have a frequency range into the OpAmp
        >and a gain which when multiplied by the frequency range is the GBW.
        >If you further increase the gain by changing Rfb what happens ....is
        >the frequency bandwidth automatically reduced or will the OpAmp not
        >respond properly to Rfb/Rg. Or is it just a distorted waveform over
        >the frequency range ?

        You can't exceed the GBW, it's a physical limitation of the op amp
        being considered. In fact, as you approach the unity gain frequency,
        the op amp is no longer acting as an op amp by definition, since one
        part of the definition includes having very high open loop gain, and
        the familiar formulas no longer apply (you have to take the lesser
        gain of the device into consideration).

        73 de VK3JED
        http://vkradio.com
      • Tony Langdon
        ... You ll get less gain and/or bandwidth than you hoped for, for starters. If there are any non linearities in the op amp at high frequencies, these will
        Message 3 of 10 , Jul 1, 2007
        • 0 Attachment
          At 01:03 PM 7/2/2007, you wrote:

          >I should say "try to exceed" the GBW.........

          You'll get less gain and/or bandwidth than you hoped for, for
          starters. If there are any non linearities in the op amp at high
          frequencies, these will also become apparent, because there is less
          negative feedback available (due to the low open loop gain).

          If you build an amplifier with 10x gain (at low frequencies) and
          attempt to run it at the unity gain frequency, then the actual gain
          will be around unity (may be more or less, because there is also like
          to be a phase shift in the output, which affects the feedback
          loop). It certainly won't be 10.

          73 de VK3JED
          http://vkradio.com
        • Tony Langdon
          ... The frequency response of the amplifier is affected by the gain. As you increase the gain (by changing the feedback resistors), you lower the frequency of
          Message 4 of 10 , Jul 1, 2007
          • 0 Attachment
            At 01:02 PM 7/2/2007, you wrote:

            >Thanks for the response, Milt. I think that's basically what I was
            >saying...... The GBW is the max frequency for a gain of 1, you can
            >also max out at that frequency/100 if your gain is 100, or at that
            >frequency/200 if your gain is 200. If you are at the max, further
            >trying to increase the gain by changing Rfb/Rg does nothing at that
            >frequency ? or does it start automatically trading off frequency vs
            >amplification (loose the Rfb/Rg ratio formula for gain) to keep the
            >GBW which is constant. Or does it just start distorting the
            >amplified waveform.

            The frequency response of the amplifier is affected by the gain. As
            you increase the gain (by changing the feedback resistors), you lower
            the frequency of the corner point in the frequency response, so yes
            you could say (in simplistic terms) you are trading gain for
            bandwidth and vice-versa. If you need a higher gain bandwidth
            product, you have to either choose a device with a higher gain
            bandwidth product, or use multiple lower gain stages in cascade (e.g.
            2 x 10dB vs 1 x 20dB should give you a little over 3x the bandwidth
            at 20dB gain, all other things being equal - 10dB is a voltage ratio
            of 3.16 roughly).


            73 de VK3JED
            http://vkradio.com
          • Kees & Sandy
            Thanks Tom, and the other guys. I understand, or at least have a much better idea of what happens. 73 Kees K5BCQ -- Tony Langdon wrote: ...
            Message 5 of 10 , Jul 2, 2007
            • 0 Attachment

              Thanks Tom, and the other guys.

              I understand, or at least have a much better idea of what happens.

              73 Kees K5BCQ

              -- Tony Langdon <vk3jed@...> wrote:

              At 01:02 PM 7/2/2007, you wrote:

              >Thanks for the response, Milt. I think that's basically what I was
              >saying..... . The GBW is the max frequency for a gain of 1, you can
              >also max out at that frequency/100 if your gain is 100, or at that
              >frequency/200 if your gain is 200. If you are at the max, further
              >trying to increase the gain by changing Rfb/Rg does nothing at that
              >frequency ? or does it start automatically trading off frequency vs
              >amplification (loose the Rfb/Rg ratio formula for gain) to keep the
              >GBW which is constant. Or does it just start distorting the
              >amplified waveform.

              The frequency response of the amplifier is affected by the gain. As
              you increase the gain (by changing the feedback resistors), you lower
              the frequency of the corner point in the frequency response, so yes
              you could say (in simplistic terms) you are trading gain for
              bandwidth and vice-versa. If you need a higher gain bandwidth
              product, you have to either choose a device with a higher gain
              bandwidth product, or use multiple lower gain stages in cascade (e.g.
              2 x 10dB vs 1 x 20dB should give you a little over 3x the bandwidth
              at 20dB gain, all other things being equal - 10dB is a voltage ratio
              of 3.16 roughly).

              73 de VK3JED
              http://vkradio. com

            Your message has been successfully submitted and would be delivered to recipients shortly.