Hi, i've to make a Rf amplfier up to 2GHz and i need to compare two type of board for the PCB: the lowest in FR4 and the highest in PTFE. When i put the dielectric constant for the two type of material FR4(er=4.4) and PTFE (er=2.02 ) the S21 give me the same result. (i also make the right wire impedence adjustment) I think that the software use a costant er for all the frequency and this isn't correct for the FR4 board! Let me know if i'm wrong.
Ok, i try to explain: ads is schematic version and the frequency sweep (S-parameter) is 1MHz to 2GHz, my problem is that ads use a constand tand and er,i think.
Then you are probably using MLIN. If this is the case, then there are three dispersive models available. You can get more information from the documentation. The footnotes in the documentation reference the model papers.
Note that in practice, PCB laminates usually do not have a uniform er. This is very difficult to model. I would imagine the tand is not very well controlled and neither is er unless the vendor puts effort into it.
That being said, if you really want to include a frequency dependent er and tand, you will need to write an equation on frequency instead of using single values. I doubt this will work for time domain, but should work fine for s-parameters.
I don't know if Momentum or EMDS allow frequency dependent er and tand.
