Brett Cooper@ 6:09pm 07-02-2012

College/Institute and Class

Utah
Hello Dr. Poddar,
Thanks for the kind opportunity to ask my beginner questions.
Best,
Brett

1) A random signal could be viewed as pure sinusoidal carrier that is randomly and simultaneously modulated both in amplitude and frequency, correct?

2) A AM modulated signal has a time-changing envelope while its "instantaneous" frequency remains constant. However, such a signal has a bandwidth of Fourier frequencies. There is a difference between the concept of instantaneous frequency and Fourier frequency: the instantaneous frequency (rate of change of instantaneous phase) ignores possible changes in the envelope of the signal while the Fourier frequency does not.
Are the two different concepts of frequency somehow related?

3) Wavelet transform is useful for both time and frequency localization. The STFT has the same purpose but is less flexible (fixed time resolution).

Why not apply the same theory of FM modulation/demodulation to signals, instead of the wavelet transform, to extract the frequency information and get its instantaneous behavior? That would give even better time and frequency localization than the wavelet transform...

4) Frequency modulation and phase modulation as slightly different but also very related. Which one is better and why? which one is more robust against noise? Which one is harder to implement?

5) What is the most common method for FM or angle demodulation? Detection of interzero crossing?
Replied on: 4:24pm 07-22-2012

(1) Can't say for sure!

(2) The instantaneous frequency of an AM modulated wave is constant and equals the carrier frequency (fc). The Fourier frequencies, should refer to the fundamental and harmonic frequencies present in the Fourier series spectrum of the AM wave. The AM wave then should be periodic with a periodicity of the modulating signal or the envelope of the modulated signal, i.e. 1/fm. Then fc and fm should be correlated through something like fc=(fc-fm) or fc=2*(fc-fm) or fc=3*(fc-fm) etc. so that the bandwith always remains 2fm.

(3) A Phase Locked Loop (PLL) may indeed be used for the purpose.

(4) If the modulating signal predominantly contains low frequencies, as with most practical signals, then the signal to noise ratio or SNR of the PM wave is greater than that of FM wave and as such, PM is superior. On the other hand, if modulating signal predominantly contains high frequencies, FM is superior to PM. For tone modulation (modulating signal being a sinusoid with a single frequency) too FM is superior to PM.

(5) Detection by using op-amp differentiators, balanced slope-detectors, zero-crossing detectors are all used..
The low cost of phase-locked-loops (PLLs) however, should make them the most common FM detectors.

Best wishes..

Biswajit Biswas@ 1:01pm 06-21-2012

College/Institute and Class

surendranath evening collage 1st year
Which day i go to collage for practice of particle class & when?
Replied on: 4:15pm 06-22-2012

You mean you want to practice before the final practical examinations!

Well, the laboratory will have to be set up prior to the final exams next week. Also B.A./B.Sc. final examinations are being held at our college, so I am not sure you will be allowed entry.

I would therefore suggest, you go through all the experiments you have performed in your mind with the help of your laboratory note book and your text-book.

If you have any questions, use the guest-book.

Biswajit Biswas@ 12:16am 06-12-2012

College/Institute and Class

surendranath evening collage 1st year
when our particle exam start? i want to met you.which date i go to collage?
Replied on: 10:16pm 06-12-2012

Where were you all these days?
You may meet meet me in the college (teacher's room) tomorrow or the day after. I would be busy with BA B.Sc. exam invigilation duties and therefore can meet you only when the first half of the exams are over at around 1 pm.
In case you cannot find me, get the exam schedule from the office.

Santanu Das@ 8:51pm 05-27-2012

College/Institute and Class

Rajabazar science college
ami apnr colg ar student chilm,,,, (2008-2011 )batch,,,,
Replied on: 1:59pm 07-01-2012

I think I can now remember you. What exactly are you studying now in Rajabazar Sc. College?
Best wishes.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
With reference to your earlier questions, you may note the following:

In a bulk material, electrons are free to move in all the three dimensions.

In a Quantum Well, electrons are confined in one dimension and as such can move in two dimensions.

In a Quantum Wire, electrons are confined in two dimensions and as such can move in only one dimension.

In a Quantum Dot, the electrons are confined in all three dimensions and can therefore move in zero dimension only.

The reduction in the degree of freedom changes the density of electronic states and leads to the creation of discrete energy levels which bring in changes to the electronic and other properties of the structure.

In a Superlattice, alternate layers of materials like two semiconductors of different band-gaps (thickness of the order of nanometers) form a periodic array of quantum wells leading to novel electronic and other properties.

These nano-structures can be built using MBE (Molecular Beam Epitaxy) or MO-CVD (Metallo Organic Chemical Vapour Deposition) Growth techniques.

For more information, especially regarding applications of such nano-structures, see Wikipedia or use Google search. For in-depth understanding of the physics of quantum wells and hetero-structures, as well as their fabrication, refer to any good book on semiconductors like that of Streetman.

santanu das@ 4:12pm 05-24-2012

College/Institute and Class

Rajabazar science college
Sir r ekta question chilo. What is super lattice?
Replied on: 1:53pm 05-26-2012

Santanu Das@ 4:07pm 05-24-2012

College/Institute and Class

Rajabazar science college
Sir kmn achen? Amr ekta question chilo. "Quantum dot" jinista ki ami bujhte parchina.. Please ektu bolben?
Replied on: 1:58pm 05-26-2012

Santanu, I am afraid I can't remember you! Were you a student of my college? If yes, do specify the year, that might help me to recognize you.
As far as your question is concerned, I feel you should first consult the web, especially wikipedia for answers to such questions. You might as well ask your teachers in the Dept of Electronic Sc. at Rajabazar Sc. College.
If you still are in doubt, use my guestbook.

Debarup Roy@ 8:47am 05-22-2012
Thank you Sir for the reply......
Can you please explain on the requisites of a semiconductor junction to be rectifying?
Replied on: 11:50am 05-22-2012

(I would suggest Streetman's book on Solid State Electronic Devices (Chapter 5) which lists the requirements for a good rectifier).

Keep in mind also that a semiconductor-metal junction could be rectifying or ohmic.

A degenerate semiconductor-degenerate semiconductor p-n junction may not be rectifying at all as is the case with tunnel diodes.

You need to find out how yourself, by consulting your text-books.

Abhijit Poddar
www.abhipod.bravehost.com

Debarup Roy@ 1:00pm 05-16-2012
Sorry Sir, I should have acknowledged your response and not to have made myself guilty.....But to say regarding your response, what about the functions and equations which need to be solved that doesn't are necessarily differential equations? In those cases how can Laplace/Fourier Transforms come handy to us, since we cannot convert them to simple algebraic equations?
Replied on: 10:47pm 05-21-2012

You may refer to Kreyszig's book (Advanced Engineering Mathematics) and go through examples like the one on solving a partial differential equation like the heat equation in an infinite bar by taking the Fourier transform of both sides and converting it to a first order ordinary differential equation which may be then be solved by the method of separation of variables.

Wikipedia also lists many applications of Fourier and Laplace transforms.

The bottom-line is that the integral transforms map a function in one domain or representation into another function in another domain or representation through an integral. For example, in communication electronics, a function representing an aperiodic signal waveform in the time domain may be converted into another function representing the same signal in the frequency domain, in which it becomes easier to perform analysis of the signal as also the system. Likewise, one may use Fourier transform to go from position representation of a quantum mechanical wave function to its momentum representation, in which the study of the quantum mechanical system may become easier.

Debarup Roy@ 12:44pm 05-15-2012
Sir, since the Q-points are the intersection points on the transistor char. curves and load lines, is it good to say that if the circuit is operated in these points then the ckt will be neither linear nor non-linear,or else will be both?
Replied on: 11:34am 05-16-2012

Hey, shouldn't you have acknowledged my response to your earlier question before asking a new one?!!

The load line represents the response of the external circuit (VCC and RC) to which the device (BJT) is connected, whereas the BJT output characteristic is a manifestation of the intrinsic properties of the device itself.
The load line equation is linear: IC = K1*VCE + K2 (K1,K2 are constants) for all values of VCE.
The output characteristic is non-linear over the entire range of VCE: IC = non-linear function(VCE).

The Q point is given by the intersection of the two curves i.e. the solution of the above two equations. Therefore, at these points, IC and VCE obey both the equations simultaneously.

In the active region of the output characteristics, however, the BJT's IC-VCE relationship is linear: IC = K3* VCE + K4 (K3,K4 are different from K1,K2). However K3,K4 change with different values of the base current IB. The transistor functions linearly when operated in this region.

For high values of IB, the Q point is pushed further up towards saturation and into regions where the IC-VCE relationship for the BJT may be non-linear.

The response of the external circuit, dictated by the load line, would however, continue to be linear.

Debarup Roy@ 9:12pm 05-11-2012
Sir, could you please elucidate on the fact that why in case of Laplace transforms/Fourier transforms we use e^-st as the multiplier for all real valued temporal functions and not any other? I mean to say why e^-st and not any other multiplier?
Replied on: 4:52pm 05-13-2012

The e^(-st) term in the expression of Laplace transform makes it possible to express the Laplace transform of derivatives of a function as products of the transform itself with some power of the new independent variable 's' (the 's' in e^(-st)); for e.g.
L{f'(t)}= s*L{f(t)}-f(0) and
L{f''(t)}= s^*L{f(t)}- s*f(0)-f'(0) etc.
This property of the Laplace transform makes it useful as a tool to solve differential equations. To elucidate, if we start with an ordinary differential equation (ODE) of the function f(t), and take the Laplace transform of both sides of the equation, we would then have converted it easily into an algebraic equation involving L{f(t)} (because of the s^(n)*L{f(t)} terms). This may be readily solved and finally the inverse Laplace transform obtained to find f(t). ODEs invoving electrical circuits may be easily solved using the above method.

The e^(-iwt) term in Fourier transform, similarly, helps it to be used in the solution of partial differential equations (PDEs).
F{f'(t)}=(iw)*F{f(t)}
F{f''(t)}=(-w^(2))*F{f(t)} etc.
In this case the PDE is converted to an ODE in the new variable (w) (through the use of the Fourier Transform) and solving it helps us to obtain the original function f(t).

There are other mathematical advantages of the presence of the exponential terms. Maybe you could find them and let us know about them.

BTW, there could also exist integral transforms which do not have exponential kernels. They may have other names and have other utilities.

Goutam Das@ 10:57pm 05-02-2012
Sir, please update our practical exam date and centre for part 2.
Replied on: 12:24pm 05-05-2012

Do check the allotment list that has arrived at the college from Calcutta University.

profdrumasankarsaha@ 2:08pm 04-24-2012
wish u all the best.
Siddhartha Saha@ 10:45am 04-10-2012
Sir, Can u plz snd a short note on Microprocessor Architecture?
Replied on: 4:37pm 04-14-2012

Microprocessor architecture refers to how the different constituents of a particular microprocessor (namely the ALU, the different general and special purpose registers, the address, data and control buses etc) help build the microprocessor and how they work together to make it perform various tasks, be it arithmatic or logical operations or interfacing with memory and peripheral devices.
The architecture of the 8085 microprocessor is different from, say the Z80, so one has to be specific while seeking to know more about the topic.

You must consult your text-book for the particular microprocessor, understand the subject and try to summarize what you have learnt in a short note.

BB@ 6:35pm 04-08-2012
Sir what are the main differences between Windows XP & LINUX?
Replied on: 4:22pm 04-14-2012

I presume you are Biroja Bandhu of 3rd year in our college. If so, I had already discussed it in the class.

Nevertheless, remember that Windows (any version) is proprietary software belonging to Microsoft Corporation and as such the user, apart from being having to buy it, has no access to its source code.

On the other hand, Linux is open-source software that comes with a GNU public license. One may use it for free and also have full read and write access to its source code. One may even redistribute it (for free or otherwise) after modifying the source code, but has to make the modified source code available to the recipient as well.

There are many other differences in features which you may find using google.

Birojananda Bondhu@ 3:27am 03-17-2012
Sir, what would be the c-program for
Newton-Rapshon method
Eular Method
Simpson method
Replied on: 3:23pm 03-19-2012

program to implement numerical integration through Simpson's method
has already been taught to you in the last class:

Read in the limiting x values: x[1] and x[N] from the keyboard (N should be odd)
Calculate the step size
h=(x[N]-x[1])/(double)(N-1);
Create the complete x grid:
for(i=2;i<=N;i++)
x[i]=x[i-1]+h;
calculate the function to be integrated fx[1] to fx[N] at all the x points
for(i=1;i<=N;i++)
f[i]=pow(x[i],3.0) + pow(x[i],2.0) + 1.0; // assuming the function to be x^3+x^2+1

sumeventerms=0.0;
for(i=2;i<=N-1;i+=2)
sumeventerms=sumeventerms+fx[i]);
sumoddterms=0.0
for(i=3;i<=N-2;i+=2)
sumoddterms=sumoddterms+fx[i]);
result=(h/3.0)*(fx[1]+4.0*sumeventerms+2.0*sumoddterms+fx[N]);

/////////////////////////////////////////////////////////////////////

program to implement Newton-Raphson's method to solve an algebric equation in x (say x^3-2=0)

for (i=1;i<=10000;i++)
{
fx=pow(xi,3.0) -2.0;
dfx=3.0*xi*xi;
xiplus1=xi-fx/dfx;
epsilon=fabs(xi-xiplus1);
if(epsilon<=0.001)
break;
xi=xiplus1;
}
printf("root=,%f",xi);

////////////////////////////////////////////////////////////////////

program to implement Euler's method to solve a 1st order differential equation
(say dydx=f(x,y)=x+y given y0(x0)=y(at x0)=0 for x0=0)

// use the relation: y1=y0+h*dydx0 (where y1=y(at x1) and y0=y(at x0) dydx0=dydx at x0
// h = spacing between successive x values, the samller resulting in more accurate solutions