Question about high frequency and high power divider/comb

Hi,all
I need to design a divider/combiner which operates at 15GHz, and the output power of combiner is more then 30W.
1.Can I choose a microstrip hybrid to design the divider/combiner? such as Branch Directional Coupler, Coupled Line Directional Couplers,
Can I choose them to combine the the amplifiers?

2.Must I use waveguide to design the hybrid?

3. If I use wilkinson Power Divider, when the characteristic impedance is 70.7ohm, the width of microstrip is 15mil, is it ok for designing the combiner? Will the waste of power be very much?

Thank you in advance!

Could anyone help me?
THE high frequency and high power divider/comb is used in a 30W PA module.
Now I need to choose the structure of the hybrid immediately !

No one help me?
because the questions are very easy or so difficult?

Re: Question about high frequency and high power divider/com

cause you want to work at 30W.
general few people design at high power.
also i am.

dielectric constant is high
The substrate loss is low
substrate Thickness is high

Re: Question about high frequency and high power divider/com

i am also desprately looking material on high frequency and high power divider/com pls if any anybody help in this regards

hi chaofan, please indicate the dissipation loss of your wilkinson splitter, as well as the material including thickness, loss tangent, etc... One more thing, are you splitting the 30W or you are combining lower power items to get the 30W?.

Re: Question about high frequency and high power divider/com

material Rogers 5880 thickness 31 mil combining 30 watt to get around 40 or 50 watt of power is it possible that combining can be done on microstrip freq is around 14 Ghz

Re: Question about high frequency and high power divider/com

Hi Chaofan,

What materials or dielectric constant you're using again?

Re: Question about high frequency and high power divider/com

dilectric constant of rogers 5880 is er=2.2

hello,

using your data the thermal resistance of your microstrip transmission line is about 1.1K/W, which implies that when handling 30W the temperature increase will be about 30K. If you are working at normal temperature (eg 20degC) that means that the transmission line will reach about 50degC at the input (and approximately 70degC at the output lines if you combine two 30W signals). In principle, duroid withstands more than 150degC, so the transmission line will not impose a problem to your design.

I was assuming substrate thickness 31mil=0.78mm, 50ohm width is 96mil = 2.45mm; epsilon=2.2; TRL loss=3dB/m. However in the first post you mention that 70ohm line = 15mil, which gives to me a thickness of 10mil for the substrate, and about 6db/m for the line, which turns to give about 4K/W for the thermal resistance of the line. assuming 30W this means a temperature increase of 117degC for the line, which is very close to the 150degC limit at the input, and above it at the output assuming 50W !!.


other thing is the loss of the power combiner. Bear in mind that the loss of the combiner translates directly into heat, so you should consider that problem carefully.

furthermore, one other thing is the power handling of the resistors of the design. when both input signals are present, the resistor does not have to handle power, but when only one signal is present, one half its power is routed to the resistor, so be careful!

rgdz

Re: Question about high frequency and high power divider/com

There are numerous PCB based components in this range. Try these suppliers: NARDA, Krytar, Tyco among others.