small signal bandwidth and large signal bandwidth

[lhlbluesky]

in some datasheet, there is two parameter, ssbw and lsbw. and lsbw may be much smaller than ssbw.
my question is:
1, what is the definition of ssbw and lsbw? as i konw, lsbw=SR/(2*pi*Vmax)
2, how to simulate the ssbw and lsbw of opamp in simulation tool, such as hspice? how to add the stimulus ? is it a AC stimulus? and is the opamp connected in open loop or closed loop (unit-gain)?
3, why is there such a big difference between ssbw and lsbw (ten times difference sometimes)? just because the input amplitude different?
i have read some related items, but don't understand very clearly.
thanks.:grin:

 

[LvW]

The answer is quite logical:
* the small-signal bandwidth can be measured with a small-signal analysis (ac simulation)
* the large-signal BW cannot be measured with a small-signal analysis (surprise?). You must use the Tran analysis. The cause of this effect is the limited slew rate.

 

[FvM]

While small signal bandwidth is a clearly defined term that doesn't need additional explanation, "large signal bandwidth" is one of several terms used to described large signal effects. It can be e.g. found in Razavi Design of Analog CMOS ICs. Power bandwidth or full-power bandwidth is another, and in my view more common, term. It's mainly referring to slew rate effects.

In a short, small signal bandwidth is determined in AC analysis, full power bandwidth in transient analysis with sine source. The parameters will be determined for a defined (closed loop) amplifier configuration.

 

[crutschow]

Answer to question 3: The large-signal bandwidth can be much smaller than the small-signal bandwidth because it requires additional power (current) to slew the amplifier internal capacitances at high signal levels (particularly the op amp compensation capacitance) and this is limited by the op amps internal bias circuits, especially for low power op amps. Thus higher power op amps usually have a faster slew rate.

 

[lhlbluesky]

thanks for reply.

 

[VirtuosoDracula]

Hi,
I want to ask a question about CS amplifier with current load. It would be really great if any of you answers this basic question of mine.
My question is, that the voltage gain that comes out for this configuration is -gm1(ro1||ro2). So why don't our gain comes as -gm1(1/gm2||ro1||ro2) like in the case of diode connected load despite of the fact that the load in both the configs is in saturation,,so why such discrepancy??
I'm sorry if you guys feel it too basic to answer..but it has really caused a lot of confusion to me. I've tried it by solving ss model as well but I'm really not able to understand the absence of 1/gm2 term in current source load..

Thanks in Advance.

 

[crutschow]

VD, please don't hijack this thread with an unrelated question. Start a new thread of your own.

 

[godfreyl]

...and please show your circuit so we know what you're talking about.