ac and dc

Discussion in 'Electronic Basics' started by jason, Mar 9, 2005.

1. jasonGuest

Hello All,

I am beginner in electronics
I wish to know why do we need to do DC analysis and ac analysis of a
circuit?
As a rule , what component or node we must short to gnd for ac
equivalent circuit?
Then for DC analysis, what are the rule of thumb?
Please let me know in step by step. Or any online document you can
share

Also if input is voltage source or current source which I do not know
if it is ac or dc type, what should I do with them when I draw an ac or

dc equivalent circuit?

Kindly help
thanks

Jason

2. John PopelishGuest

The common names for the two kinds of analysis are transient and AC.
Transient analysis follows the response of each component through
time, as all the surrounding conditions change. It makes no
assumptions about linearity, but simply follows the models given for
each component. AC analysis assumes a single set of operating
conditions and assumes that every part acts in a linear fashion, and
solves for the gain and phase shift of the circuit over a range of
frequencies. Transient analysis works in the time domain (where one
instant follows another, and sine waves are a foreign language) and
AC analysis works in the frequency domain (where all signals are waves
that exist for all time).
None. But before you do AC analysis, you have to figure out what bias
point all the nonlinear parts (diodes, transistors, etc.) are
operating at so you can pick the linear value you will use for the AC
analysis. This bias point calculation assumes all capacitors are open
circuit, and all inductances are shorts, so that time does not have to
get involved in this calculation, since these are the time dependent
parts.
DC analysis (and the bias point calculation that precedes AC analysis)
requires detailed models of all the components.
I have a shelf of books on the details of how each kind of analysis is
both types and gives you graphical and numerical results.
http://www.linear.com/company/software.jsp
It also exports schematic files in ASCII form, so you can post them in
places like this, for help with design questions. They also have a
Yahoo forum you can join that deals with operation of the simulator,
itself. But before you can understand how it works, you need to do
quite a bit of study of the analysis methods, themselves.
Draw a squiggle (one cycle of AC ~ ) after the voltage label on the
voltage or current source.
That is the one with all the capacitors open (removed) and inductances
shorted (replaced by their winding resistances).

3. Andrew HolmeGuest

DC analysis is about bias point and quiescant (quiet) state i.e. what
the voltages and currents are when there is no signal.

AC analysis is about gain and frequency response to signals.
Anything that is at a constant potential e.g. because it is decoupled
to ground by a capacitor; or it's a power supply rail ...
Well, for bipolar transistors, you can often get away with:

1. Ie = Ic+Ib
2. Vbe = 0.7V
3. Ic = beta * Ib

If you set the base voltage with a potential divider, the quiescant
emitter voltage will be fixed about 0.7V below that. You can then set
the currents with a resistor from the emitter to ground. It's a big
subject, but that should give you a flavour. There are many books
Well, you can usually guess from the context. If it's a signal, it's
probably AC. If it's a capacitively-coupled input, you would omit it

4. Kitchen ManGuest

The math is different. For DC (steady state), you can make do with
algebra. AC and transient requires calculus and imaginary numbers.

You may find answers to this and your other questions on sites like this
one:

5. jasonGuest

Thank you everyone for spending time and effort to help.
I really appreciate it.
I would like some confirmation from you for the following;
1) For ac analysis, current sources from coponent like a current mirror
or current sink must be shorted? Am I right?

2) For ac analysis, voltage source such as battery must be shorted.
While anything that is represented by a circle with ~ in it (~) is the
alternating voltage which must be kept for ac analysis. Am I right?

Kindly let me know if the above statement is right

Then is ac equivalent circuit another name for small signal equivalent
circuit?
Are they the same ?

In small signal circuit, I do see capacitors still being kept in the
circuit. So does it mean small signal circuit is not ac equivalent
circuit?
If they are the same, why capacitors are remain at small signal
circuit?
Another thing is that I read that capacitor is considered shorted at
high frequency. So how will this affect any general ac signal or small
signal model?

Kindly enligthen.

Thank you

rgds and thanks
Jason

6. Andrew HolmeGuest

No. These constant current sources/sinks bias the transistors
high-impedance to AC signals so, in your AC analysis, you replace them
with an *open* circuit i.e. remove them.
It is not black and white / open or closed curcuit. There are degrees.
The reactance of a capacitor depends on frequency. Big capacitors are
almost a short circuit to high frequencies. It depends on the circuit
designer's intention. If he just wants to couple a signal and block
DC, he selects an appropriately large capacitor. Other times, he
chooses an in-between value that does not approximate an open or a
closed curcuit. These in-between values are there for various reasons:
impedance matching, phase shifting, filtering .... and, yes, you need
to leave them in your AC equivalent circuit. What constitutes
"in-between" depends on the reactance of the capacitor compared to the
impedance levels in the surrounding circuitry. It's relative.

7. Guest

Hi Thanks a lot for the suggestion
But may I know what does the archieves function do for us?

Jason

8. Kitchen ManGuest

Jason, my best advice is to invest in some formal training, or at the
very least get some good textbooks and make sure you have the math
skills to follow the analyses. Circuit analysis is not trivial. I came
out of the USAF with an excellent background as a RADAR technician, and
when I started on my Electronics Engineering degree, the circuit
analysis classes kicked my butt.

9. Andrew HolmeGuest

Imagine someone connects a voltage source to the constant current
source; now imagine the voltage changes by a small amount dv. The
current does not change - because it's constant - so:

di/dv = 0

r = dv/di = infinity

So, to AC, the constant current source "looks" like a high resistance.
It's never infinite in real life because no current source is pefect.

10. jasonGuest

Thanks a lot Joseph, Andrew and Kitchen
It is really great to see such an excellent answer in this forum. I
believe some of them cant never be found on books. Actually I am
following some text, else I wont have so many questions as the more I
read , the more I would ask and the more I will know . Of course thanks
to all who have guided me here.

Cheers and thank you all

Jason