1
\$\begingroup\$

I just came about an op-amp spec where they say, "the Op-amp X is unity-gain stable with a small-signal unity gain bandwidth of 105MHz.." As you continue to read, it then says, "with a gain-bandwidth product of 70MHz, ...."
What is the difference between these specs? How can its unity gain be specified for two different bandwidths (I understand it says small signal, but I still don't know how I'd use the information). It seems to me the meaningful spec is the 70MHz GBP and the 105MHz is just "specmanship". But, if anyone can explain what they mean by that small signal BW, and how that info is useful, I'd appreciate it. Thank you.

Edit: I didn’t want to malign it, but at your requests, here’s a link to the OPA2810 data sheet. https://www.ti.com/lit/ds/sbos789c/sbos789c.pdf?ts=1605745386167&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FOPA2810

\$\endgroup\$
3
  • 4
    \$\begingroup\$ Include a link to that opamp's datasheet. "Small signal" means that the signal is such that the (undesired) effects of "large" signals are not limiting the performance. One common large signal performance limit is slewing, see: electronics-notes.com/articles/analogue_circuits/…. \$\endgroup\$ – Bimpelrekkie Nov 18 '20 at 21:16
  • 1
    \$\begingroup\$ We definitely need a link to the datasheet. The two specs may be under different conditions like supply voltage or load or test circuit. One may be "typical" and the other may be minimum. It could even be an error in the datasheet. \$\endgroup\$ – John D Nov 18 '20 at 21:32
  • \$\begingroup\$ I believe the point is that GBW=gain times bandwidth only holds true in the dominant pole approximation, i.e. 90 degrees crossover. Open loop gain plot instead (fig. 44) shows phase changing before 0dB modulus, there must be a second pole around. \$\endgroup\$ – carloc Nov 19 '20 at 2:02
3
\$\begingroup\$

Not sure if this is the op-amp you are referring to, but the OPA2810 typical characteristics contain similar numbers.

Interestingly, perhaps, the typical unity-gain bandwidth of the amplifier increases from 75MHz to 105MHz (typ) when the capacitive loading is increased from 4.7pF to 33pF as it makes the amplifier less stable and there is some gain peaking near the cutoff.

enter image description here

At higher gains (+11), with 4.7pF loading, the typical gain-bandwidth product is 70MHz, which is pretty close to the 75MHz unity gain bandwidth with 4.7pF loading.

\$\endgroup\$
2
  • \$\begingroup\$ Yes, this was the opamp I was talking about. I still don’t follow, though. The 2 plots show frequency response for 20mVpp input at multiple gains (left) -unity gain being the red curve- while the rhs plot shows unity gain at different common mode output, ie, a 20mVpp signal centered at various DC levels (and CL=33pf).. I see that the BW increases, but I don’t understand how this is useful info. How would I use this seemingly very specific info? \$\endgroup\$ – jrive Nov 19 '20 at 1:18
  • \$\begingroup\$ You're looking at the light grey curve in the RH graph vs. the red curve in the left graph. Notice also that the -1 response is quite different from the +1. How do you use it? Depends entirely on what you are trying to do. If you need flat response the bandwidth is going to be much less than if you just care about when you lose half the power gain. \$\endgroup\$ – Spehro Pefhany Nov 19 '20 at 3:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.