How does bandwidth pertain to op-amps?

Thread Starter

StToonz

Joined Nov 30, 2012
14
Hi,

I am studying for the CET and I come across the amplifier section. Specifically the two different feedback types. Positive and negative feedback. It says negative feedback reduces gain and increases bandwidth. Ok I have studied it yet but I understand what is meant by it reduces gain I just don't know how yet. Anyway not the point, I know what bandwidth means but how does that pertain to and op-amp that's where the confusion comes in.

Thanks for any help, this website has always been helpful
-conrad
 

crutschow

Joined Mar 14, 2008
34,285
Negative feedback reduces gain since some of the output is fed back and subtracts from the input voltage. In a op amp the open loop gain is so high the circuit gain is typically determined by the value of the feedback elements.

That being said, the op amp internal compensation network, which provides stability for the op amp with negative feedback, generates a "gain-bandwidth" product for the frequency response, which is specified in the op amp data sheet (sometimes as the unity gain frequency response). Thus the bandwidth of the op amp circuit is basically determined by the GBW divided by the closed loop gain. For example, an op amp with a GBW of 1MHz would have an operating bandwidth of 100KHz in a feedback circuit that has a closed-loop gain of 10.
 

#12

Joined Nov 30, 2010
18,224
One of the things I have noticed is that, "increases the bandwidth" seems misleading. It needs to be pronounced, "increases the bandwidth that has the same gain as the lower frequencies".

You see, all opamps have decreasing gain ability as the frequency goes up. You see it as a slope on the graph of gain bandwidth. When you use negative feedback to limit the gain, the bandwidth that has the gain you want becomes more. That doesn't mean it is more than the opamp originally had, it just means you are using the opamp in a way that the declining slope of the response curve doesn't start affecting the gain until the frequency is higher than it would be if you ran the amp at its highest gain.
 
Last edited:

antonv

Joined Nov 27, 2012
148
Does this help?: Attached is a chart that shows an OPA211's ac gain curve. The shape of the curve is that of any voltage feedback op-amp's open loop ac gain vs. frequency. Differences between part numbers will be in open loop gain and at what frequency the gain starts dropping. The shape of this curve (the red line) is due to the amp's internal design (do you need to go into that in detail?). In this case the amp's bandwidth is only about a 100 Hz and below a 100 Hz its gain is somewhere around a billion (120 dB). So narrow bandwidth, huge gain.

If you add feedback resistors (now you are using closed loop gain) to set the gain at say 45 dB, then the amplifier's ac gain curve looks like the blue line, you don't change the op-amp's ac characteristics you only chop the top off the curve, and so now its bandwidth is pushed out to 600 kHz.

Much less gain traded for much more bandwidth.

The bandwidth is totally controlled by where you set your gain. This is not true for current feedback op-amps.
 

Attachments

ramancini8

Joined Jul 18, 2012
473
We sure are beating a simple fact to death. In a voltage feedback op amp the gain falls off with increasing frequency as shown in OPA211 graph, at a fixed 20dB/decade. First, notice that beyond 100 Hz the open loop gain decreases. If the op amp is configured for a closed loop gain of 100 (40dB), the open loop gain is approximately 1000 at 100KHz. The error at 100KHz is reduced by the excess gain of 10 to 10%, and at 10Khz the error is reduced by the excess gain to 1%. At approximately 700KHz the gain is -3dB or .707 times the calculated gain.
 
Top