patch antenna microstrip feed line question

[Hellfir€]

Hi all,

i ve got a question regarding to using of 50 ohm microstrip feed line for mstrip patch antenna.

lets say we have an antenna like this:

https://www.antenna-theory.com/tutorial/txline/quarterwave4.jpg

Why are we using a 50 ohm feed line ? I know why we are using transformer between antenna and feed line but dont know why we are using feed line after connector to transformer? Cant we use only lambda/4 transformer directly between antenna and 50 ohm connector as a single microstrip line feed?

another question is lump port has something to do with it?


thanks all.

 

[WimRFP]

Your feeling is right.

The length of the 50 Ohms transmission line between the connector and the quarter wave transformer is not important, as there is no or little reflection in the 50 Ohms line. So the length can be zero.

There is advantage of simulating a piece of transmission line. In real world the source may not be at the input of the quarter wave transformer and there will be a transmission line between the transformer and the source. The discontinuity has some parasitic effects (inductance and capacitance). By simulating with a 50 Ohms trace (or trace width other impedance), one can tweak the antenna to take this discontinuity into account.

Another thing is that many simulators have de-embedding schemes to enhance accuracy. The simulator adds some transmission line segments in order to calculate the current and voltage at the actual position of the port. The extra segments in front of the actual port position assures that the desired propagation mode is present only (for example quasi TEM in case of microstrip lines). Using de-embedding with the trace width of the quarter wave transformer is not useful in my opinion.

If you would simulate your antenna + quarter wave section without 50 Ohms line, you may experiment with a localized/lumped port. You will notice that the length of the lumped port section has influence on simulation results. Normally spoken I would use a lumped port only when I can't use a port scheme with de-embedding. I frequently have situations that I need to use a lumped port (for example non-planar antennas and antenna feeds that are not above a ground plane).

 

[Hellfir€]

Your feeling is right.

The length of the 50 Ohms transmission line between the connector and the quarter wave transformer is not important, as there is no or little reflection in the 50 Ohms line. So the length can be zero.

There is advantage of simulating a piece of transmission line. In real world the source may not be at the input of the quarter wave transformer and there will be a transmission line between the transformer and the source. The discontinuity has some parasitic effects (inductance and capacitance). By simulating with a 50 Ohms trace (or trace width other impedance), one can tweak the antenna to take this discontinuity into account.

Another thing is that many simulators have de-embedding schemes to enhance accuracy. The simulator adds some transmission line segments in order to calculate the current and voltage at the actual position of the port. The extra segments in front of the actual port position assures that the desired propagation mode is present only (for example quasi TEM in case of microstrip lines). Using de-embedding with the trace width of the quarter wave transformer is not useful in my opinion.

If you would simulate your antenna + quarter wave section without 50 Ohms line, you may experiment with a localized/lumped port. You will notice that the length of the lumped port section has influence on simulation results. Normally spoken I would use a lumped port only when I can't use a port scheme with de-embedding. I frequently have situations that I need to use a lumped port (for example non-planar antennas and antenna feeds that are not above a ground plane).

Thanks WimRFP, it was really helpfull.

and what do you mean with " discontinuity"? Does it refer to feed line?

additionally, can you tell me the difference between lump port and wave port simply? I read something about it in the forums but it was little bit complicated.

 

[WimRFP]

Hello,

A discontinuity is a change in width or diameter in a transmission line or a bend Generally spoken, it is a change in te normally continuous structure of a transmission line and may cause some reflection. In your case it it the change in trace width at the "feed line - quarter wave transformer " transition. At such a discontinuity the (quasi) TEM propagation mode is distorted. A transition from coaxial feed (for example an edge mounted SMA connector on a PCB) into a microstrip is also a discontinuity. Careful design is required to keep reflections to a minimum in case of UHF/SHF.

wave / lumped port
Assuming you have a port that provides energy in positive X direction into a transmission line. The port is at the left side, the transmission line with antenna is on the right side. A wave port extends the trace in negative X direction (to the left of the actual port position). So that the port itself doesn't behave a as discontinuiity. In other words, at the port's position, there is a (quasi) TEM wave.

A lumped or localized port just converts a metallic segment into a voltage or current source. So in case of our example, at the position of the port, there is a voltage or current source. The transmission line is not extended into negative X direction, so it is no longer a continuous structure. This causes field distortion at the position of the port. Many simulators alllow both vertical and horizontal lumped/localized ports.

If you want to simulate a dipole antenna in free space (without modelling the feed line), you use a lumped/localized port in the middle of the dipole to feed the dipole. I hope your simulator's documentation have some examples in it, so that you can experiment with them.

 

[rodgerwxh]

Agreed with WinRFP. Wave port solves a 2D eigen value problem, any shape in a cross sectional cut only support a certain amount of wave-modes that can propagate through. For example, a Rectangular waveguide with width and height given, then it only support TE10 or TM01 ...... mode (Check Balanie's book chapter 6). Such a mode is a 2D spatial distribution of the field on the cross section of your structure. Waveguide, coaxial, TL can all be feeded by such called waveport. It guarantees that the field you put in to X+ will never get a reflection from the X- direction due to the mode-matching.

However, everything depends on your real problem, if you are feeding your antenna (e.g. horn antenna) with a rectangular waveguide, then of course when you design your horn, you want to use a waveport as the feeding since in reality you did use waveguide to feed. If you are designing a microstrip antenna, at least in my best knowledge, most of the time the TL is fed with SMA with only the pin touching the TL and outer touching the ground plane, then in such case, lumped port is good enough and obviously better than waveport to model this situation, since SMA simply induce a voltage/current at one point on the TL which is best described by lumped port behavior. Of course, in such a case, you should expect some reflections from your simulation due to the discontinuity, but such reflection is in reality as well.

 

[WimRFP]

I put an example of a horizontal lumped port for CPW here:
https://www.edaboard.com/threads/290830/ posting: #7.