The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Vdd tied to ground in AC mode
https://designers-guide.org/forum/YaBB.pl?num=1319181152

Message started by haykp on Oct 21^st, 2011, 12:12am

Title: Vdd tied to ground in AC mode
Post by haykp on Oct 21^st, 2011, 12:12am

Dear Forum,

As you know while doing small-signal analyses (AC) we consider the VDD tied to ground.
Please help me understand why?

thanks,

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 21^st, 2011, 12:35am

haykp wrote on Oct 21^st, 2011, 12:12am:

Dear Forum,
As you know while doing small-signal analyses (AC) we consider the VDD tied to ground.
Please help me understand why?
thanks,

VDD is considered as an ideal voltage source having an internal source resistance of zero. Thus, ac-wise it is grounded.

Title: Re: Vdd tied to ground in AC mode
Post by loose-electron on Oct 21^st, 2011, 5:53am

What is the value of AC signals present in a DC bias line?
Answer = Zero

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 21^st, 2011, 6:18am

loose-electron wrote on Oct 21^st, 2011, 5:53am:

What is the value of AC signals present in a DC bias line?
Answer = Zero

What do you mean with "signal"?
I think the dc line can carry an ac current, can it not?

Title: Re: Vdd tied to ground in AC mode
Post by Garrett.Neaves on Oct 21^st, 2011, 6:54am

haykp wrote on Oct 21^st, 2011, 12:12am:

Dear Forum,

As you know while doing small-signal analyses (AC) we consider the VDD tied to ground.
Please help me understand why?

thanks,

You set vdd to ac ground when you are doing an ac analysis in which your ac input signal is some signal other than vdd.

Typically, an ac analysis is concerned with the ac response due to a single ac signal source. That is, all ac signal sources are set to zero except for the one of interest.

If the ac response to the ac component of vdd is of interest, then vdd would not be tied to ground. Instead, all other ac signal sources would be set to zero. The ac response due to signal on a power supply such as vdd is generally referred to as PSRR for power supply rejection ratio or power supply ripple rejection.

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 21^st, 2011, 6:58am

I am afraid, there is a misunderstanding.
For my opinion the content of the original question is:
Why is the source Vdd replaced by signal ground during ac analysis?

Title: Re: Vdd tied to ground in AC mode
Post by haykp on Oct 21^st, 2011, 7:15am

Thanks all for the comments!

buddypoor many thanks for your more precise explanation.
But one question regarding yourcomment: if the internal source resistance is zero, does it mean we can tied it to ground?

Will you agree with following proof:
In ac mode there is cap between Vdd and Vss, hence we can assume that Vdd is tied to ground.

Title: Re: Vdd tied to ground in AC mode
Post by raja.cedt on Oct 21^st, 2011, 7:34am

hello haykp,

Let me explain what is small signal to understand this bit clearly.

Small signal means how your ckt parameters are changing with applied signal(don't confuse with the ac or dc, it's just change in the bias forget about the frequency for now ). So for example take a transistor with 100uA bias current for .5v vgs, now if you change vgs by 1mv then your current changes by 10uA, so in small signal model we only represent changes (not fixed quantities), now if you draw small signal model vgs and ids would be 1mV and 10uA. Think about an ideal voltage source, no change in the voltage hence short ckt, if it has some internal resistance, instead of short ckt, add resister from that node to gnd and same applicable to current source.

NOW Coming to your last post, we generally keep big decap between supply and gnd to support large amounts of switching currents. This acts as ideal voltage source, so you can represent it by short ckt.

SORRY FOR THE BIG POST.

Thanks,
raj.

Title: Re: Vdd tied to ground in AC mode
Post by raja.cedt on Oct 21^st, 2011, 7:38am

hey haykp,
forgot to tell you, please understand small signal model is not ac, where as small signal model will be used to do ac analysis.

Thanks,
raj.

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 21^st, 2011, 9:05am

In addition to Raja's explanations:
In reality, the supply voltages are NOT ideal (with zero internal resistance) - and, therefore, we put a large capacitor across it (sometimes in parallel with a small one). This ensures a small and negligible internal ac resistance.

Title: Re: Vdd tied to ground in AC mode
Post by loose-electron on Oct 21^st, 2011, 6:50pm

buddypoor wrote on Oct 21^st, 2011, 6:18am:

loose-electron wrote on Oct 21^st, 2011, 5:53am:

What is the value of AC signals present in a DC bias line?
Answer = Zero

What do you mean with "signal"?
I think the dc line can carry an ac current, can it not?

Agreed, however now you are analyzing the power supply loads, which probably is not what you want in a signal processing analysis.

Title: Re: Vdd tied to ground in AC mode
Post by haykp on Oct 21^st, 2011, 11:32pm

Dear raja.cedit and buddypoor,

Many thanks for your explanations.

Finally I understood he reason of replacing Vdd signal by ground in small signal analyses.

Raj, yes I was mixing small signal analyze with AC analyze but now, after your comments I wont :))

Thanks again.

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 22^nd, 2011, 1:18am

haykp wrote on Oct 21^st, 2011, 11:32pm:

Raj, yes I was mixing small signal analyze with AC analyze but now, after your comments I wont :))

Hi Haykp, just for a correct understanding:
An ac analysis ALWAYS is a small-signal analysis. That means: All transfer characteristics are linearized around the chosen operating point (determined through bias conditions).

Title: Re: Vdd tied to ground in AC mode
Post by haykp on Oct 23^rd, 2011, 10:34am

Thanks buddypoor,

As we are speaking about AC analyze, at this stage my assumption is that there is some difference between AC analyze and small-signal analyze.
Though I understand that AC analyze uses small-signal analyze for calculations, but it somehow different. So what is the difference between AC analyze and small-signal analyze?
( please sorry if I am mixing anything)

Title: Re: Vdd tied to ground in AC mode
Post by loose-electron on Oct 23^rd, 2011, 7:33pm

AC analysis tends to mean looking at the frequency response of a system.
Small signal tends to mean looking at the signal processing characterisitics while ignoring DC bias issues.

That is not precise, but the term AC analysis comes from simulator origins, and small signal term comes from hand analysis methods.

thats the history of it...

They overlap and are somewhat similar.

Title: Re: Vdd tied to ground in AC mode
Post by buddypoor on Oct 24^th, 2011, 1:08am

loose-electron wrote on Oct 23^rd, 2011, 7:33pm:

AC analysis tends to mean looking at the frequency response of a system.
Small signal tends to mean looking at the signal processing characterisitics while ignoring DC bias issues.
That is not precise, but the term AC analysis comes from simulator origins, and small signal term comes from hand analysis methods.
thats the history of it...
They overlap and are somewhat similar.

I do not completely agree with loose-electron.
*At first, the term "ac analysis" is much older than computer-based simulations. It is a set of rules used to calculate the frequency characteristics of passive and/or active networks.
Example: The transfer function of a simple passive RC lowpass 1/(1+sRC) is based on these rules: In this case, the impedance Xc=1/sC. This expression - and, thus, the transfer function - applies for sinusoidal signals (that is the meaning of "ac") only!.
*Secondly, both expressions are not "similar" and "do not overlap".
They cannot be compared because they express different principles, which are however closely related.

- A small signal analysis assumes that all relations between the state variables (currents and voltages within a circuit) are linear.
That means: Linearizing all functions around the dc values (bias point).
This leads to the (funny) result that you can use kilovolts as input voltages during small-signal analyses. Such a small-signal analysis can be performed for any kind of input sgnal (squarewave, for example).

- The ac analysis uses this principle and calculates all frequency-dependent properties of a network with sinusoidal excitation - also based on linear equations only (which certainly is always an approximation - with acceptable deviations in most cases).

Thank you

Title: Re: Vdd tied to ground in AC mode
Post by Alexandar on Oct 24^th, 2011, 2:25am

I would like to add to the previous post that besides linear, the system should be time invariant as well. In such a way it's possible to apply LTI system theory to circuits.

Title: Re: Vdd tied to ground in AC mode
Post by loose-electron on Oct 24^th, 2011, 3:11pm

Lex wrote on Oct 24^th, 2011, 2:25am:

I would like to add to the previous post that besides linear, the system should be time invariant as well. In such a way it's possible to apply LTI system theory to circuits.

They are both a linear estimation methodology.

Looking at a few of my older textbooks from the era prior to simulators, shows that it was referred to as "Network Analysis" in 1976 - In another book from 1986 the term "transfer function" gets used. The term "AC analysis" started getting used after it was applied to the function of spectral analysis using simulators.

The term "phasor analysis" also gets used in some earlier texts.