Microstrip transmission lines

Hello,

I have a new question about microstrip simulation and application.
In the past days, I have been working on getting a transmission line using microstrip with duroid material.
Now, here's what happens to me.
I simulate in MWO and after tuning, I get something ready to build.
However, when I actually build it it looks nothing like the sim.
I understand that's why it's called simulation because it's similar.
However, I've done other simulations with LNA's, filter, directional couplers and with most of them I get very close results to what I am looking for.

So, on to my problem.
I wish to build a series-shunt PIN diode SPDT and from there move up to a SP4T switch.
I have my diode model and everything.
I wish to make it broadband from the lowest possible ( f < 500 Mhz) all the way to 18 GHz if possible.
I've tried a few, but I can't seem to get beyond 3 GHz.

Simulation looks beautiful. Reality looks terrible.
And I my guess is how I am doing the simulation that makes it all wrong.
I even made an extract em simulation and even that one looks good.
But again, when I build it....it all goes out the window.
So, two things can be happening:

1.- My simulation is incorrect.
2.- My actual build is not being tested properly.

or maybe both. I don't know at this point. i asked a colleague and he said I should have some sort of matching network in the ports, input and output to level things.

does any of this make any sense at all?

Thank you very much for reading this long message.

That is a lot of bandwidth for a PIN diode switch and PIN diode models are less than reality. Without seeing the problem, it will be difficult to help. Can you attach simulation results and test data?

added:
I just read your other post "Coplanar Waveguide question" and retract my statement above. You need to get a handle on the basics before attempting this. This is a difficult/complex design.

Ah....ok.....I understand the nature of both types...
Microstrip and coplanar.
That's why I tried both.
I understand this is a challenge.
Do you have any suggestions on where I could start?
I understand a few basics but not all of them.

- - - Updated - - -

Ah....ok.....I understand the nature of both types...
Microstrip and coplanar.
That's why I tried both.
I understand this is a challenge.
Do you have any suggestions on where I could start?
I understand a few basics but not all of them.

- - - Updated - - -

Sorry about the double post.

this thing was not allowing me to reply and was acting up.

- - - Updated - - -

Adding a little bit to this...

I've worked with switches that have more than 3 or 4 channels,
that can cover the bandwidth I am talking about.
This is because each channel covers one section of the entire band.
Which makes perfect sense...
But the assignment here for me is the following.
Is it possible to build one that can cover the entire band?
Or is it just a crazy idea to follow?

Thank you very much for your reply, ge.

- - - Updated - - -

Adding a little bit to this...

I've worked with switches that have more than 3 or 4 channels,
that can cover the bandwidth I am talking about.
This is because each channel covers one section of the entire band.
Which makes perfect sense...
But the assignment here for me is the following.
Is it possible to build one that can cover the entire band?
Or is it just a crazy idea to follow?

Thank you very much for your reply, ge.

>> Or is it just a crazy idea to follow?
Crazy Idea.
You are looking at a significant development cost (compared to where you currently are). 1/4 wavelength structures does not work over multi-octaves. Companies such as Krytar spent a bunch of money and time to make those work.

I'm also having issues with double post and edaboard keeps logging me out.

Crazy edaboard.....logged me out a bunch of times as well....

Ok...so if it's such a crazy idea, how can I prove that I can build a switch for such broad band?

Just stick to the regular channels? And knock them down one at a time?

I am curious to know one thing then.

http://www.nardamicrowave.com/east/i...getPdf&id=3728

they advertise as PIN diode switches.
And, it covers a lot of bandwidth.
So, how is it that they were able to build something like that?
I don't want to know what's in there or any schematic or anything.
I am just looking for the mechanics so I can embrace the concept and from there work my way into it.

I thought I did but apparently I am lost.
I've done switches at low frequencies (Fc < 1 GHz) with High power (50W+)
I am trying to up in frequency, no need for power (for now) but I seem to not be fully aware of all I need to know...

Thank you very much for your reply.