What all considerations should I take in account while desiging a Micro-strip line at high frequency (70-85 Ghz)?
I want to design a micro-strip line for a mm wave application. I am using formula mentioned in David Pozar`sMicro-wave Engineering book, Is there any other method or analysis I can use to design for the frequency range suitable for my application. Thank-you in Advance
Hello Khalid Shaikh, Pozars book is a good start. There are multiple books written on the topic, that go into more detail and only focus on MS and coplanar designs. At these frequencies you have to take the surface roughness of your materials into account. Due to the skin effect the electromagnetic waves will hardly penetrate the conductor. I would also recommend that you try to simulate your MS line in a suited EM-simulator. The best suited simulator for this case should be SONNET, but I am sure you can also achieve acceptable results with AWR Microwave Office or Ansys HFSS. Please keep in mind, that not everything that can be simulated can also be manufactured. You should check the manufacturing parameters of your PCB supplier and orient your design according to these requirements. You should also think about how to measure your line once it has been produced. You probably need to create probe pads or use special RF connectors ( I think 1mm is required at these frequencies). Best regards
Microstrip and other lines behave at 85 GHz similar to what they do at lower frequencies, if everything is scaled by the same amount. However, the dielectric may have higher losses and the skin-depth will be thinner, so these things don't quite scale, but impedance and component layout mostly can and should be roughly scaled from lower frequency designs. As Martin Obermaier says, things like surface and edge roughness, and also material variation over a short scale (glass-fibre weave?), have a more significant effect at higher frequencies. You may not want to scale the board thickness, so some things won't be the same, and the track thickness may be larger in proportion to the width, dropping the impedance slightly compared to the same width to height ratio at lower frequencies.
In this frequency range it may be interesting to consider as an alternative also dielectric lines (sort of printed optical fibers) since(metallic) microstrip lines are really getting small then and have considerable losses.