Horn with septum for RT

Pablo_UDEC

Newbie level 6
Joined
Oct 1, 2024
Messages
11
Helped
0
Reputation
0
Reaction score
1
Trophy points
3
Visit site
Activity points
155
Hello everyone, I am currently designing a septum horn in a dual-polarization rectangular waveguide. I would like to resolve some questions that have come up, and I would appreciate any recommendations and advice.

First of all, I have derived the septum calculation from documentation. One of the requirements is that it has a 200 MHz bandwidth. I have read several documents on how to implement them, but I don’t understand how to find the impedances for each step, for example Z_{n+1}. If anyone has information on this that could clarify it, I would appreciate it, because it is important that the septum is designed with a specific bandwidth of interest. As I mentioned earlier, I used a known calculator for those who have designed this type of OMTs, but it does not specify for which BW in particular it is made. I understand that the mathematical theory is of a stepped impedance transformer, and this can be designed with objective parameters such as wide bandwidth and low reflection coefficient. I am attaching the link to the calculator: https://docs.google.com/spreadsheet...WUUmURkmXs/edit?gid=1682293646#gid=1682293646.

What I did in my HFSS project is a waveguide analyzed for 3 ports. The output port has Ex and Ey with a 90-degree phase shift (I don’t understand how to view this), and both input ports are set to 1W, with one of them having a 180° phase shift. From what I’ve read in different sources, this should happen. Now, I’m not sure if this occurs in the two frequency bands of interest (1.42 and 1.667 GHz). Currently, I have two input ports, but later I will add coaxials to extract both right and left circular polarizations at quarter wavelengths from a short circuit, so I’m not sure if it’s correct that I placed these ports at the end of my OMT. I would like to know if I am extracting values that should be important for the analysis of this instrument. What do you recommend I change, modify, or what other data are important, for example, the modes? Thanks in advance.

[Moderator action]
Embeeded to Edaboard relevant file hosted to external server
 

Attachments

  • 1.png
    32.5 KB · Views: 51
  • 2.png
    99.2 KB · Views: 48
  • Downloads.rar
    243.7 KB · Views: 36
Last edited by a moderator:

how to find the impedances

two frequency bands of interest (1.42 and 1.667 GHz).
Though I have no experience with your construction, general rules could apply:

* Impedance goes hand in hand with the voltage and Ampere level, that is, the amounts generated or amounts available. L:C ratio becomes important.

* Greater inductance L is associated with lower frequency, lower Amperes.
* Greater capacitance C is associated with lower frequency, greater Amperes.

* Bandwidth is related to Q. Q is damped by increased ohmic resistance. Bandwidth is narrowed as you eliminate ohmic resistance.
 
Hi Pablo, first thing is to make sure you have the correct size of square waveguide, you don't want any higher order modes propagating --> they should all be under the cutoff frequency. It is correct to set the two waveguide ports and do the coaxial transition later. At the COM port (square WG) make sure to add at least two modes (X, Y). Even better is to set it to five so you can monitor the level of higher order modes. The 90deg phase difference will be between the two fundamental modes at the COM port from one of the waveguide inputs (doesn't matter which one). If your waveguide is properly dimensioned the septum will equally divided the power and the optimization of the septum steps will be to have as close to 90deg phase over the band will keeping a good return loss and port-port isolation.
 

Cookies are required to use this site. You must accept them to continue using the site. Learn more…