Better op-amp buffer configuration?

[billcalley]

Hi All,

I realize that using a single-supply op-amp as a buffer may be one
of the easiest circuit configurations in the world; just tie the
chip's output back to the inverting input, and insert the signal into
its non-inverting input. That's how the books show it anyway, but
are their any real-life op-amp issues that I have to look out for with
this basic circuit? Anything I can do to make this circuit operate
better under temperature, along with the unavoidable op-amp variations?
Perhaps to help with any stability or voltage offset issues? Or is
this circuit simply the way op-amp buffers are done, even in real life?

Thank You,

Bill

Cross posted to sci.electronics.design and sci.electronics.basics

 

[Glenn Gundlach]

Hi All,

I realize that using a single-supply op-amp as a buffer may be one
of the easiest circuit configurations in the world; just tie the
chip's output back to the inverting input, and insert the signal into
its non-inverting input. That's how the books show it anyway, but
are their any real-life op-amp issues that I have to look out for with
this basic circuit? Anything I can do to make this circuit operate
better under temperature, along with the unavoidable op-amp variations?
Perhaps to help with any stability or voltage offset issues? Or is
this circuit simply the way op-amp buffers are done, even in real life?

Thank You,

Bill

Cross posted to sci.electronics.design and sci.electronics.basics

For many (but not all) opamps that configuration will work fine. Quirk
I had with LM324 as a follower, fast rise or fall times exceeded speed
of the amp and cross coupled (through the bias system?) to other amps
in the package. TL084s fixed that. Current feedback opamps can be used
this way but would require a feedback resistor to prevent oscillations.
Best bet, check manufacturers data sheets to verify usage.
GG

 

[Adrian Tuddenham]

Hi All,

I realize that using a single-supply op-amp as a buffer may be one
of the easiest circuit configurations in the world; just tie the
chip's output back to the inverting input, and insert the signal into
its non-inverting input. That's how the books show it anyway, but
are their any real-life op-amp issues that I have to look out for with
this basic circuit? Anything I can do to make this circuit operate
better under temperature, along with the unavoidable op-amp variations?
Perhaps to help with any stability or voltage offset issues? Or is
this circuit simply the way op-amp buffers are done, even in real life?

Check whether the chosen op-amp needs external compensation to be
unity-gain stable (the NE5534 does).

Check whether a resistor may be needed in series with the output to
prevent problems if you are driving a capacitive load such as a long
screened cable.

 

[martin griffith]

Hi All,

I realize that using a single-supply op-amp as a buffer may be one
of the easiest circuit configurations in the world; just tie the
chip's output back to the inverting input, and insert the signal into
its non-inverting input. That's how the books show it anyway, but
are their any real-life op-amp issues that I have to look out for with
this basic circuit? Anything I can do to make this circuit operate
better under temperature, along with the unavoidable op-amp variations?
Perhaps to help with any stability or voltage offset issues? Or is
this circuit simply the way op-amp buffers are done, even in real life?

Check whether the chosen op-amp needs external compensation to be
unity-gain stable (the NE5534 does).

Check whether a resistor may be needed in series with the output to
prevent problems if you are driving a capacitive load such as a long
screened cable.[/QUOTE]
Put the isolating R in series with the output of the opamp, and take
the feedback to the -i/p After the resistor. Generally works, and
compensates for attenuation that you would get when you drive a lowish
Z load


martin

 

[billcalley]

Thanks for all of your help guys!

Cheers,

Bill

 

[martin griffith]

snip
Actually, if you take the feedback from after the resistor, the stability
will normally be even worse than without the resistor at all. If you need
to take feedback from after the resistor for accuracy reasons, then put a
capacitor directly from the op-amp o/p pin to the inverting input, and a
resistor between the load and the inverting input, and of course the
original low value between the op-amp output pin and the load.

http://www.analog.com/library/analogDialogue/archives/31-2/appleng.html

Chris

excellent article, thanks


martin

 

[Chris Jones]

Check whether the chosen op-amp needs external compensation to be
unity-gain stable (the NE5534 does).

Check whether a resistor may be needed in series with the output to
prevent problems if you are driving a capacitive load such as a long
screened cable.

Put the isolating R in series with the output of the opamp, and take
the feedback to the -i/p After the resistor. Generally works, and
compensates for attenuation that you would get when you drive a lowish
Z load


martin[/QUOTE]

Actually, if you take the feedback from after the resistor, the stability
will normally be even worse than without the resistor at all. If you need
to take feedback from after the resistor for accuracy reasons, then put a
capacitor directly from the op-amp o/p pin to the inverting input, and a
resistor between the load and the inverting input, and of course the
original low value between the op-amp output pin and the load.

http://www.analog.com/library/analogDialogue/archives/31-2/appleng.html

Chris

 

[Ol' Duffer]

Anything I can do to make this circuit operate
better under temperature, along with the unavoidable op-amp variations?
Perhaps to help with any stability or voltage offset issues?

Instead of just connecting output to inverting input, use a resistor
to approximate impedance of input to minimize offset drift due to
input bias current.