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.
Any Microwave Handbook or a more basic electronics handbook describes the topic. But if you have not tried before I would advise not to try building these components. It is really difficult if you do not have many years of experience.
It depends on the application how successful you will be. If you can stand a bit of loss in your system then it's OK and you will not spend 6 months trying to optimize it. I have had some success printing ferrite loaded pastes onto ceramic.
If you need best performance then I agree with other posts in this topic - buy one!!!
My thesis was on that topic. What is your target frequency, power, bandwidth, loss and size requirement. First at all, you need to know how much you would pay for each one. The military class isolator/circulator can be as much as $200. If you just need a few, you can buy from some company, MA/COM is a good one. Usually, circulator is substrate dependent. Even you want to make circulator yourself, you need to find a good sourse of substrate, such as YIG.
Could you share more information on printing the ferrite paste on ceramics? The idea seems good, if there's enough accuracy.
Other thoughts: if you're going to make an isolator or a circulator, you'll need a huge bunch of expensive machinery in addition to the simulators and a nework analyzer: an accurate CNC mill (for the waveguide isolators and general mechanics), a good diamond saw and grinder lathe (for shaping the ferrite pieces, round or triangular) and a magnetizer for fine-tuning the magnets. Not a one-weekend project.
OK, most of these tools can be made in a (hard-core?) hobbyist-level as well, but it's still not going to be simple. Very educating, though, and you might be able to make a small business out of that experiment as well.
There is nothing especially magic about circulator design.. you just have to understand what it going on.
I have desiged many waveguide isolators, mostly Faraday rotation type but also Y junction, with low loss in the 100GHz region.
Yes, machining ferrite is nasty (but not impossible.. it grinds easily and you do not necessarily need diamond tools)
No, you do not need to tune the magnetic field. If the design is done correctly, and optimised for the chosen ferrite magnetic saturation, you simply need enough field to saturate the ferrite, and it is then also not sensitive to external fields.
A well equipped micro machine shop is all you need, and a good designer