I gotta admit I am rookie in this field. In my current working, I am encounter with some problems, I designed transmission line, grounded coplanar waveguide, return loss is not bad but insertion loss is not what I expected. Then I changed it into conductor backed coplanar waveguide, I designed the space between each grounding vias as quarter wavelength which I read from an article, both the return loss and insertion loss is fine. But due to there is too many grounding vias, the cycling time for manufacturing is a problem. As I changed them half wavelength, the insertion loss is even worse than grounded coplanar wave guide. I wanna ask that is there any solution for that?
Probably, the solution is following recommendations to SIW. When you miss some vias, you create a leaky structure; more pronounced with strict periodicity.
Great place to start is https://www.microwaves101.com/encyclopedias/substrate-integrated-waveguide
When you created an SIW, in the band specified you have your initial channel with PEC approximated walls. Then you can modify it for your kind of coplanar. There are advantages and disadvantages to that... But I guess if there is a backplane, there are more advantages of confining the wave inside, rather than letting the modes going loose along the substrate.
Denis Jaisson
,,I don't really get your question, But if you mean what software I am using. I am using Ansoft HFSS ....
Denis Jaisson
That's why I said I am a rookie dude. I just searched about field calculator in HFSS help, but I couldn't fing anything related to field calculator. It would be great if you could be more specific on that? What is field calculator? why do I need that when I design a transmission line?
HFSS is a field calculator.
The quarter wave spacing is a result of years of practical experience, with a good theoretical reason. It may be possible to achieve low loss in other ways (I don't know) but that is a project that requires significant knowledge in the field and such a solution would probably have repercussions on things like bandwidth, for example, that without experience in the field you won't find out until it doesn't work as you wanted.
The reason for the very poor performance is probably because with half wavelength spacing, you produced a series of coupled resonant structures. Via spacing should preferably be less than .25 wavelength.
don't forget that conductor backed CPW has more complicated modes than simple cpw. And periodic vias provide band-gaps, which both can result in occasional mismatching
I concur with Andrey Porokhnyuk . GCPW can have all sorts of unintended parallel plate (and other) modes if the substrate is too thick for your frequency range. You also run the risk of coupling to unintended resonators either side of your transmission line if you maintain large top and bottom ground planes (the open-circuited board edge is a pretty decent magnetic wall at those aspect ratios; it's the principle behind HMSIW and QMSIW cavities).
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity