Discussion Forum

Hi

try a search on the forum, this has been discussed a few times, in v4 you have the waveform functions too

--
Good luck
Ivar

Hi

thanks for reply but i found square wavefroms in functions only


actually, i want to apply a pulse train with specified rise and fall time

i have searched in forum but i don't understand what is mod function? and heaviside function?

please help me in detailed steps ( i am new comsol user)

Hi

"mod()" is the "modulus" operator check the web or your math books, or even COMSOL help "operators"

The COMSOL waveform "square pulse" train, has smoothed rise and fall times (the same value for both) so if you need different rise and fall times you can combine two step() operators in an analytical COMSOL function (and you can plot them easily to check the values)

--
Good luck
Ivar

Dear Mr.Ivar

thank you for taking your time but i am really very confused in comsol. i open analytical function to make pulse train but i don't know what can i write in ( Expression field) and (Argument) !!!!
and what about lower limit and upper limit !!!



i want to built pulse train of amplitude= 16
pulse width=500 e-6 sec
time period =1 ms
rise time = 100 ns
fall time = 28 e-6 s


i am very for your help

yomna

Hi

take a closer look to the model library examples:

arguments are typically x, or t, or ... its a local definition and replace later with the expression/variable you enter
expression is the equation you want dependning on the "argument" i.e. sin(2*pi*x)

min max values are the limits of the argument for the plot.

try it out on a simple known case, once you feel familiar, then you might go on with complex expressiosn, and perhaps multiple arguments ...

You could define two step functions one from 0 to 1 and another from 1 to 0, with transition durations of your respective rise and fall times, with time offset of 1/2 the rise time for tthe 0 to 1 step1() function, and 1/2 fall time + pulse duration for the 1 to 0 step2() function, then yo combine tem in an analytical expression of the type 16*(step1(t)+step2(t))

Pls check it carefully I'm not by my COMSOL WS so I cannot check it out

--
Good luck
Ivar

Dear Ivar,

I have read the thread about creation of pulse train. From there I was able to generate one. For test purpose, I went ahead with a single pulse and found that it works until pulse is down. As soon as the pulse is down, the solution do not converge. What might be the problem, please help me understand my mistake. I am giving the specifications below.

pulse start time : 0.5s
pulse width : 1s
analytical function : rect1(t) (Although when I use rect1(t[1/s]) it is not accepting in v5.2)
variable --> Vol : 10*an1(t[1/s])

My aim is to generate a pulsed voltage train for a period of 10 seconds. I am using AC/DC module. I have used the appropriate boundary conditions. FYI, model works fine for continuous voltage.

I would say you can do two things to be successful.

One thing is to increase your "size of transition zone" in the rectangle definition.

Another thing that is important is the the time stepping method in your solver settings.
Set the "steps taken by solver" (BDF Method) to "intermediate" and try again.

Those things help for me.

Good luck
Ingo

Dear Ingo,

Thanks for your early reply. I already have my size of transition zone much greater than my time step. That didn't work. I have tried taking intermediate BDF method and ended up with the same problem, i.e the solution do not converge after the pulse fall off. An other suggestions in solving this are greatly welcome.

Thank you.