[SOLVED] AC analysis(linear) for folded cascode opamp?

[petelee]

Dear Analog experts,

I'm looking at this file in the web.

https://www.eit.uni-kl.de/koenig/deutsch/TESYS_Lutgen_09.pdf

It's a bit slow to upload. I'm trying to understand how to get the graph in page 25. It says "AC analysis(linear) in differential mode" and the lefthand side graph's y-axis is some voltage level with Mag(V). Could you anybody know what kind of simulation input stimulus in ac analysis I need to use to get the graph? I'm only family with voltage gain in ac analysis.

Best regards,
Pete

 

[erikl]

... voltage level with Mag(V). Could you anybody know what kind of simulation input stimulus in ac analysis I need to use to get the graph?

It's an AC analysis with a std stimulus of unit size (1V), which means a DC gain of 16.5Meg ≙ 144dB .
For AC analysis, the circuit is linearized around its operating point, so the absolute values of voltages don't care.

 

[LvW]

Could you anybody know what kind of simulation input stimulus in ac analysis I need to use to get the graph? I'm only family with voltage gain in ac analysis.

Hi pete, as mentioned above, the ac analysis is an analysis that shows the gain around the linearized transfer curve.
The input stimulus is an ac voltage. Normally, you select Vac=1V and the output voltage - given as magnitude and phase - is identical to the gain (Vout/Vin).
However, of course you can use another voltage, in this case you must display the ratio Vout/Vin (often also in dB).
In your example the frequency is swept not linearly but in log steps with at least 10 points/dec.

 

[petelee]

It's an AC analysis with a std stimulus of unit size (1V), which means a DC gain of 16.5Meg ≙ 144dB .
For AC analysis, the circuit is linearized around its operating point, so the absolute values of voltages don't care.

Hi Erikl,

Thanks for commenting~~

Y-axis in the graph of page 25 is labeled as "Mag(V)". Does "Mag" mean 10e6 scale?

And also, in page 26, the gain is around 90dB, not 144dB. They're not matching. I'm still confused.

Regards,
Pete

 

[LvW]

Hi pete,

it is not easy to read the graph - but on page 25 the voltage is 16.5 V. That is the MAGNITUDE (mag), because the ac analysis gives mag and phase.
Because the input voltage is unknown, you cannot derive a gain value from this graph.
However, we can calculate "back".
The gain is app. 84 dB (not 90 dB).
For a gain of 84 dB (1.6*1E4) the input voltage (page 25) would be 16.5V/1.6*1E4 which is approx. 1mV.
This sounds realistic.

 

[petelee]

Hi Eric,

In page 21, in the diagram, va=1. So I couldn't understand the 16.5v magnitude. I think the diagram is just for illustration, and the author must have changed it to 1mV for his real simulation later.

Now everything makes sense to me.

I appreciate your kind explanation. Have a great day~~

Sincerely,
Pete

 

[LvW]

In page 21, in the diagram, va=1. So I couldn't understand the 16.5v magnitude. I think the diagram is just for illustration, and the author must have changed it to 1mV for his real simulation later.

Yes - on page 21 the test setup is explained in detail. However, for actual simulations an input of Vac=1mV was used.

As another information: The SLOPE of dc transfer curve on page 25 and 26 (right side) gives you also the dc gain. If you are interested (to complete the picture and your understanding), try to determine the slope and compare it with the max. gain value of 16000 (84 dB).

- - - Updated - - -

 

[erikl]

In page 21, in the diagram, va=1. So I couldn't understand the 16.5v magnitude. I think the diagram is just for illustration, and the author must have changed it to 1mV for his real simulation later.

Hi petelee,

yes, I had mixed up Mag with Meg, sorry! As one can see from page 23 (green marked line) the DC gain actually is Av≈16000, so they must have stimulated with 1mV, not 1V as indicated on page 21 - as LvW agreed, too. And LvW's hint concerning the slope of the DC transfer curve also supports this Av value.

Best regards, erikl

 

[petelee]

Hi LvW and Ericl,

I just noticed that I've communicated with two gentlemen. I thought I have talked to only erikl.

LvW, I appreciate your valuable inputs. EricL, thanks a lot for coming back and confirming LvW's comment~~

Regards,
Pete