best 45 degree RF bend

[yefj]

Hello, I have designed a coupler 32-38Ghz however i need to bend the legs in an optimal way.
My bend is 45 degrees. Is there a logic i could use to find the optimal distance of the bended leg an the coupler?
maybe the the legs input leg has some logic too i need to use?
Thanks.

 

[volker@muehlhaus]

The bend itself must be optimized, not feed line length

See section "Microstrip miter compensation"

Microwaves101 | Mitered Bends

Microwaves101 | Mitered Bends

www.microwaves101.com

 

[yefj]

Hello Volker, i saw a much improvement when i have shortened the distance of the bend to the coupler.
The smith charts before an after show it clearly.
What could cause this?
Thanks.

 

[volker@muehlhaus]

It is possible that some parasitics compensate, depending on phase offset. It is also possible that your feed line impedance is not exactly 50 Ohm.

But why don't you follow the microwave design rules that work well for everyone since 50 years? You always ask very basic stuff that is known for very long time.

 

[yefj]

Hello Volker, i think there is another thing I remmember.Where the distance of the bend from the coupler has some resonance logic.
it needs to me multiple of quarter wavelength.
Is this sound familiar to some theory maybe?
Thanks.

 

[volker@muehlhaus]

Optimize the right thing, don't try to hide one issue with another. Your Smith chart shows that your S11 impedance is too inductive, for whatever reason.

Distance matters if the bend itself is not matched well, or if the line is not matched well.

 

[yefj]

Hello Volker,Yes i know that for a resonance we need the inductive part be equal to the capacitive part .
Could you please recommend me a manual to see how in my situation I can increase the capacitive part?
Thanks.

 

[volker@muehlhaus]

I give up, you do not read the answers and ignore all advice.

 

[yefj]

Sorry , so you say its purely the bend
Ok i will try to work on it.

 

[volker@muehlhaus]

so you say its purely the bend

The bend and/or line impedance of the lines from coupler to the ports, see #4