Is this a right approximation- for High speed opamps?

[analogartist]

Wondering what causes High speed in an opamp(closed loop).. I derived the step input for an closed loop single pole system to be
V0(1-e(-wf))
where wf is the closed loop bandwidth= wo(1+βAo)

where wo is open loop BW , Ao is open loop gain and β is feedback factor..

Is it good to approximate this wf to wo*β*Ao... considering wo << wo*β*Ao.

If I do this then the speed of the system depends on GBW and not BW.? But most of the books say speed depends on the BW(Rise time =0.35/BW).

Put it in another way
If I have an opamp1 which has say a BW of 300MHz and GBW of 1MHz and
and another opamp2 which has a BW of 500Mhz and GBW of 800Hz. Which is is faster pamp1 or opamp2?

Would appreciate enlightment on a simple (bothersome ) question?

Thanks

 

[pixel]

Rise time depends on bandwidth and for two pole response is 0.3-0.5/Trise.

 

[elbadry]

I think a faster Opamp means a higher GBW Opamp.
A faster performance for a given feedback circuit involving an Opamp means a higher BW of the overall circuit.

This can be thought of as follows:

GBW is a characteristic of the Opamp ( intrinsic property indepent on any feedback network ).

So, for the same feedback factor β, the opamp with higher GBW will have a larger BW and hence faster performance.

Hope I didn't confuse you!!

 

[Miguel Gaspar]

You wrote:

Is it good to approximate this wf to wo*β*Ao... considering wo << wo*β*Ao.

This is a question.

Usually in manufacturers catalog the term High speed refers to GBW≥5MHz, thats the case of Texas Instrument, and the slew rate grows according to the GBW. And for National Semiconductors, GBW ≥ 4MHz.

Added after 2 minutes:

you can reed the following article of Texas Instruments:

https://focus.ti.com/lit/an/sloa051a/sloa051a.pdf

Added after 1 minutes:

Where you can find that the term high speed refers to bandwidth and slew rate.

 

[analogartist]

Miguel,

I am not sure if I got it right.. so here I am stating it again.....

So what you are saying is the approximation is right for wo > 5Mhz?

But otherwise is it primarily BW that is of concern not GBW

Added after 26 seconds:

Miguel,

I am not sure if I got it right.. so here I am stating it again.....

So what you are saying is the approximation is right for wo > 5Mhz?

But otherwise is it primarily BW that is of concern not GBW

 

[Miguel Gaspar]

Well that information is from the manufacturers of opamps.
About w0 >5MHz, it is from the catalog of texas instrument.
In the article that I gave you, it is of high speed opamps, where they say that the two most important parameters are BW and SR. And, yes they don't mentioned GBW.

Check the catalog of National or Texas or any other manufacturer to verify that information.

 

[jerryzhao]

I agree with elbadry.
u compare the two op's speed, they must have the same β.(all the op assume single pole).
U should better read some book about the op, such as the Razavi's Design of Analog CMOS integrated circuit.
The feedback of op like a system, you must understand it clearly. otherwise our answer will confuse you.
Give you a sample:
A car and a bike witch run faster?
Maybe you say car.
But I say the bike run faster if in the country road .
So u can think over the elbadry' answer.
I hope you understand, and don't confuse you too.

 

[eldo]

Of cause speed is defined by GBW.
For the same β, higher GBW, higher speed.

By the way, GBW is always larger than BW. It is impossible to have a BW of 300MHz and GBW of 1MHz !