Electronic Projects, forums and more.

Go Back   Electronic Circuits Projects Diagrams Free > Other Forums > Math and Physics


Math and Physics Discuss the complex nature of mathmatics and physics relating to electronic circuitry.

Reply
 
Thread Tools Display Modes
Old 9th November 2007, 06:51 AM   (permalink)
Default Inverse Laplace Transform

I need to inverse Laplace transform this:
2s/((s^2+4)(s+2))

I know the answer is cos(t)^2+sin(t)*cos(t)-1/2-1/2*exp(-2*t).
I can inverse Laplace transform stuff, but I'm a little unsure with this specific type of problem. Can you explain how to do it, please?

Last edited by drolex; 10th November 2007 at 12:20 AM.
drolex is offline  
Reply With Quote
Old 9th November 2007, 10:55 AM   (permalink)
Default

I think your answer is wrong. I get:
f(t)=.5(sin(2t)+cos(2t))-.5e^(-2t)

Start with partial fraction expansion for complex roots:
F(s)=(As+B)/(s^2+4) + C/(s+2)
2s=(A+C)s^2 + (2A+B)s + 2B+4C

set coefficients equal
0=A+C
2=2A+B
0=2B+4C

I get A=.5, B=1, C=-.5

The -.5/(s+2) section is straightforward

Manipulate the complex section to get in a recognizable form
(Mw+N(s+a))/((s+a)^2+w^2)=(.5s+1)/(s^2+4)
where w=2 and a=0 as determined by the roots of (s^2+4)
solve for the amplitudes M and N by setting coefficients equal
2M+Ns=.5s+1
M=.5, N=.5

F(s)= (.5*2+.5(s+0))/((s+0)^2+2^2)-.5/(s+2)
Use tables for ILT
f(t)=.5(sin(2t)+cos(2t))-.5e^-2t

I skipped some of the details, but these links should fill them in:
www.tcc.edu/faculty/webpages/pgordy/PEReview/Partial.doc
http://mathworld.wolfram.com/LaplaceTransform.html
jeg223 is offline  
Reply With Quote
Old 9th November 2007, 11:10 AM   (permalink)
Default

Well personally I have no idea about the whole transform part, but have you tried splitting the fraction, "partial fractions" style, in order to make your life easier?

2s/[(s^2+4)(s+2)]


A/(s+2) + (Bs+C)/(s^2+4) = 2s/[(s^2+4)(s+2)]

A(s^2+4) + (Bs+C)(s+2) = 2s
let s = -2
8A = -4
A = -1/2

-1/2*(s^2+4) + (Bs+C)(s+2) = 2s
let s=0
-2+2C=0
C = 1

-1/2*(s^2+4) + (Bs+1)(s+2) = 2s
let s=1
-5/2+3(B+1)=2
(B+1)=3/2
B=1/2

Sooo, after all that;
2s/[(s^2+4)(s+2)] = -1/[2(s+2)] + (1/2s+1)/(s^2+4)
Tidy it up a bit
=(s+2)/(2s^2+8) - 1/(2s+4)

I hope that helps you cause, sorry I am TeX n00b, i'll make sure I learn it at some stage. As I said, I'll leave you to the transform part...

EDIT: jeg beat me to it typing it out, but he atleast finished it for you... lol, also learned something new from him... excellent lol.

Last edited by erosennin; 9th November 2007 at 11:13 AM.
erosennin is offline  
Reply With Quote
Old 9th November 2007, 11:27 AM   (permalink)
Default

At least we got the same answers for the coefficients. Thats a good sign
jeg223 is offline  
Reply With Quote
Old 9th November 2007, 01:33 PM   (permalink)
Default

Quote:
Originally Posted by jeg223
I think your answer is wrong. I get:
f(t)=.5(sin(2t)+cos(2t))-.5e^(-2t)

Start with partial fraction expansion for complex roots:
F(s)=(As+B)/(s^2+4) + C/(s+2)
2s=(A+C)s^2 + (2A+B)s + 2B+4C

set coefficients equal
0=A+C
2=2A+B
0=2B+4C

I get A=.5, B=1, C=-.5
Perfect solution I just want to point out that it's exactly what the OP expected.

sin(2x) = 2sin(x)cos(x)
cos(2x) = 2cosē(x)-1

Hence, f(t) = .5(sin(2t)+cos(2t))-.5e^(-2t) = cos(t)^2+sin(t)*cos(t)-1/2-1/2*exp(-2*t)
eng1 is offline  
Reply With Quote
Old 9th November 2007, 10:41 PM   (permalink)
Default

I never could remember those trig identities
jeg223 is offline  
Reply With Quote
Old 9th November 2007, 11:56 PM   (permalink)
Default

Quote:
Originally Posted by jeg223
Manipulate the complex section to get in a recognizable form
(Mw+N(s+a))/((s+a)^2+w^2)=(.5s+1)/(s^2+4)
where w=2 and a=0 as determined by the roots of (s^2+4)
solve for the amplitudes M and N by setting coefficients equal
2M+Ns=.5s+1
M=.5, N=.5
Thanks!
Where is the table that you found that transform on?
drolex is offline  
Reply With Quote
Old 10th November 2007, 04:24 PM   (permalink)
Default

It's actually the sum of the sin(wt) and cos(wt) transforms with a scalar multiplier and a frequency shift.
The mathworld link above has a table, but if you look you should be able to find a more extensive table with the (Mw+N(s+a))/((s+a)^2+w^2)=e^-(at)(Msin(wt)+Ncos(wt)) transform.

Last edited by jeg223; 10th November 2007 at 04:32 PM.
jeg223 is offline  
Reply With Quote
Reply

Bookmarks

Thread Tools
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Latest
Inverse of DC offset going down instead of up walters General Electronics Chat 25 1st December 2005 06:43 PM
Laplace transform help... elec123 General Electronics Chat 1 16th October 2005 01:27 PM
LAPLACE TRANSFORM cyprio7 General Electronics Chat 6 20th April 2005 09:48 PM
Enquiry Of Fourier Transform shermaine General Electronics Chat 2 25th August 2004 07:14 PM
Fourier Transform shermaine General Electronics Chat 2 20th March 2004 01:49 PM



All times are GMT. The time now is 11:21 PM.


Electronic Circuits  |  Electronics Wiki
Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2008, Jelsoft Enterprises Ltd.