Hey guys, Whats the best way to determine the bandwidth of an op amp apart from looking it up in the datasheet?
Use a signal generator to create a sinewave and measure the output. Find the corner frequency by changing the frequency until the output is 3 dB down
Many opamps have "slew rate limiting" where the max output level drops above a certain frequency because the output waveform becomes a triangle-wave. A lousy old 741 opamp has a frequency response to 1MHz when its gain is 1 and its output level is low but its max output level drops above 9kHz due to slew rate limiting.
Unfortunately not. So do I just take it that the bandwidth of the op amp is the frequency at which the gain drops 3dB below the output peak? Also did I read somewhere that at the bandwidth frequency the output should be roughly half the output peak, or did i just imagine that?
3 dB down is the same as 1/SQRT(2) As Audioguru's graphic illustrates there is the "bandwidth" of the output being 3dB down and there is the "unity gain bandwidth". Two entirely different animals.
Thanks for the help with the bandwidth guys, got it sorted. One other question, I want to measure the step response of the op amp. To do this I am inputting a 1Vp-p square wave at a low frequency, but I can't see any overshoot on the ouput?
This certainly is not a bad thing. Minimal overshoot with no capacitive load would suggest that the amplifier has plenty of phase margin >45°. What amplifier are you looking at? You can determine phase margin from the open-loop gain vs. phase curve in the datasheet.
Its the AD8651 , the graph your talking about is on page 9 of the data sheet. Im working with a gain of 100, so by my calculation the phase margin is approx. 67°?
A closed loop gain of 100 is 40dB. If you drew in a closed loop curve across 40dB, it would intersect the open loop curve ~600kHz and the phase there is -90°. So you have 90 degress of phase margin. Beautiful, textbook curve IMO.
Sorry, I was reading the value of the phase axis where the two intersect, my mistake, thanks. So with so much phase margin whats the best way to measure the step response?
The way you are doing it is fine. Voltage follower, 1kHz square wave. You can do small signal step response, <500mVp, and large signal step response, >1Vp to observe the overshoot and ringing.