Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

help frequencies methods ( FDDF,FDTLM)

Status
Not open for further replies.

as-zu

Newbie level 6
Joined
Dec 24, 2005
Messages
11
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,361
etd finite differencefrequency domain

hello eveybady, could one of you help me

i want know the difference between (FDfD et DFTD) et (TLM et FDTLM)
ie in the algorithm


thanks
 

I "think" the methods you are referring to are:

Finite-Difference Frequency-Domain (FDFD)
Finite-Difference Time-Domain (FDTD)
Transmission Line Method (TLM)
Time-Domain Transmission Line Method (TLM)

All four of these methods typically work by fitting the fields and materials to discrete points on a Cartesion grid. The formulation of equations relating the fields through Maxwell's equations is slightly different in each method, but they are all essentially just enforcing Maxwell's equations across the grid.

Time-domain methods have the ability to simulate a device over a very broad frequency range in just a single simulation. A pulse source is typically used so at all points of interest the impulse response is recorded. Fourier transforming the impulse response gives you the spectral response. FDTD has another big advantage in that it does use linear algebra. This mean that if you double the size of your problem, memory and run time only doubles. Other methods, doubling the size of the problem could increase your memory and run time by 4 to 16 times or more! Time-domain methods, however, are poor for modeling highly resonant devices, or devices where the spectral response changes abruptly. In these cases, the frequency-domain methods tend to be more accurate.

Frequency-domain methods tend to be more efficient and accurate for highly resonant devices or when you are only interested in one or two frequency points. They are based on linear algebra so memory and run time increases exponentially with problem size. Typically you convert Maxwell's equations to either A*x=b for scattering problems, or A*x=a*x for eigen-value problems.

I will also say that FDFD is the simplest numerical method I know to implement. It is not as efficient as the finite element method (FEM), but FEM is more tedious to formulate. I will also say that I do not see TLM or TDTLM used nearly as much as FDTD so there is much more literate available for FDTD. Most of this is also applicable to FDFD, but otherwise FDFD can be difficult to find in the literature.

In my Ph.D. dissertation, I describe in good detail a number of numerical methods including:

Finite-Difference Frequency-Domain (FDFD)
Finite-Difference Time-Domain (FDTD)
Plane Wave Expansion Method (PWEM)
Rigorous-Coupled Wave Analysis (RCWA)
String Method
Level Set Method (LSM)
Fast Marching Method (FMM)

You can download it here:

**broken link removed**


Hope this helps!
-tip
 

thank you very much for your answer

i want to know which is the expression of excitation in frequency method

thanks
 

I don't understand your question. Are you asking how to incorporate a source into a frequency-domain method? That depends on the numerical technique you are choosing. There are many types that are all "frequency-domain," For example there is finite element method, finite-difference frequency-domain, method of lines, etc. Which method are you interested in?

-Tip
 

thank you ,

i'm intersting in FDTLM

thanks
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top