Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Hi,
I'm also trying to design this kind of structure! Do you have any documents now? Please share them if possible!
My email: tung_manh2000@yahoo.com
Thanks a lot!
Yes, you did not provide enough information. CPW is mostly used on high frequencies. May be Wilkinson combiner/divider can be of choice, but please provide more details for your question.
Yes, you can do it with a several stages, but there is some caveats. The required isolation for low end and for high end arguable will be achieved with multistage Wilkinson. Secondly, did you check the group delay numbers? The 5-50MHz part is dedicated to the reverse link for cable modems. There S21GD is very important but this is the area when Wilkinson will have a big spike. One more point: did you count the DC path connection to one leg of the splitter and its effect? All these questions are not easy even for you, but how less experienced guys will solve them? I heard about one 10 stages Wilkinson for the band of 50 - 6000 MHz. The size was about 10 inches and it was designed for lab tests, not for real high speed networks.
It is interesting. If you designed Wilkinson for CATV and it match the CATV specs it is very good. My congratulations! I remember that I designed Wilkinson about 3 years ago for CATV but for 5 - 5000MHz range. Even one stage on FR4 was able to work so-so for insertion loss and was better than any off-the-shelf splitter. But isolation problems still exists and S21GD issue at low end too. Later I switch to transformer based splitter, but still keep in mind Wilkinson and going to make it again when time allows. Usually CATV itself is not a big problem, the cable high speed (1GBs) network has stronger spec.
I think that FR4 will provide almost the same result as Rogers. The only problem that FR4 have is unstable Er. In this case Nelco 4000-13 is good material with very stable parameters and practically the same price as FR4. If it is not secret why you extend CATV to 3 GHz? As far as I know TV stops at 2150MHz and high speed network starts at 3300 MHz. Area between 2150 and 3300 MHz is empty or I am not right?
What about wifi, bluetooth, wimax they located between 2-3G, but all this I meantion
is wireless (not by cable like CATV).
and about the pcb I am not recognized vs Nelco 4000-13 (Can u upload some details),
but currently i am using EM280: tje same price as FR4 but stable ε (4,3-4,5) and tanδ is lower.
I do not know about EM280, could you please upload the datasheet or other information? I included Nelco datasheet here. I used this material for about a year for frequencies up to 6 GHz. It works well and so far there are no problems at all.
Regarding CATV spectrum between 2.15 and 3.3 GHz my concern is serious. I design high speed networking systems which are sitting above 3.3 GHz. Today we have no concern what is below 3.3 Ghz but it may be compatibility problem in near future if another company wants to use this spectrum. So, if you can tell me about the usage of 2.15-3.3 Ghz band it will be nice to know and consider in our designs future compatibility. If you are doing design for 2.15-3.3 Ghz it is beneficial for you.
Thanks a lot for this information. Unfortunately I have no ability to download due to exceeding download limits long time ago. May be they will fix this mistake, but today I cannot see your file. I will try to find this material by its name.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
