Electronics > RF, Microwave, Ham Radio
Antenna reciprocity: directivity/gain
ejeffrey:
--- Quote from: E Kafeman on June 16, 2021, 09:55:04 am --->If we don't take into account material property change due to a high power for TX, the path loss will be identical.
Antenna path loss, |S12|=|S21|, will always remain same in both directions or else do we need a new physic.
--- End quote ---
That's not true. |S12| == |S21| is a consequence of a system that is linear, time invariant, passive, and isotropic. The first two are needed for S-parameters to even be well defined (although in some cases you can hack something that works anyway), active devices like amplifiers are rarely reciprocal, and devices made with magneto optical materials such as circulators are not isotropic. These conditions are so widely applicable and useful that people tend to think of it as a fundamental law but forget that it comes with caveats.
ejeffrey:
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm ---Another example is a transformer with core. When you don't exceed it's specification of the input voltage, it's parameters will be within specification. But when you reach some voltage limit, it leads to a core saturation and transformer parameters will be significantly different.
--- End quote ---
This is a particularly good example because a transformer by itself could be considered to be reciprocal. The S parameters are power dependent but saturation happens the same forward and reverse. But you have violated the rules of the proof which means you have no guarantees. If you then cascade the transformer with an attenuator (which is reciprocal) the cascaded network is longer even approximately reciprocal since a signal from one direction hits the transformer first and saturates it. The same power from the other direction is attenuated first and doesn't saturate the transformer.
E Kafeman:
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm ---When I talked about path loss, I mean power loss from one antenna port to another antenna port.
--- End quote ---
Yes that is how antenna path loss and Friis equation normally is defined. See https://www.antenna-theory.com/basics/friis.php
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm --- Because when you apply high power (for example 1 kW) to Antenna 2 it will change ferrite rod properties (magnetic saturation).
--- End quote ---
This is on the other hand not how antenna path loss i defined. Friis equation assumes a reciprocal system, which it is if measuring from antenna port to antenna port just as you wrote above.
Link budget is a better description for what you is trying to describe.
Only passive isotropic and reciprocal elements exist between antenna ports as we else would cause an unbalance of total amount of energy in universe depending on signal directions. That is not allowed.
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm ---By the way, I think S-parameters is not applicable here, because S-parameters are intended for a DUT with linear behavior, but effects I'm talking about leads to a non-linear behavior.
--- End quote ---
Yes you is talking about active non reciprocal systems, not related to antenna path loss calculation. Please check Friis equation. Not even reflection loss is included as it is happening before antenna port and for same reason do it not neither exist any big Tx rig causing antenna elements to saturate. Just forget your mega power transmitters if it is transmission loss that is calculated.
For same reason in Friis equation is not a Tx modulation or transmission mode factor. For link budget however, assume that your ferrit is saturated to a big resistor due to a lot of power feed to it. It will affect Tx losses exactly just as much as Rx losses if it is duplex traffic.
For good reasons do we not include local malfunction or poor design in Friis equation. Traffic mode will neither be a part of path loss, but may affect link budget.
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm ---Another example, when you apply high enough power into transmission antenna which is placed in the air environment, the air state may be changed to a plasma,
--- End quote ---
Yes, just as I did wrote above.
--- Quote from: radiolistener on June 16, 2021, 03:56:25 pm ---
--- Quote from: E Kafeman on June 16, 2021, 09:55:04 am ---Antenna path loss is not including power as parameter in its unit-less result but it describes a power ratio, how much power is reduced at receiving antenna port from Tx antenna feeding port.
Path loss includes antenna directivity but not losses due to antenna imperfections in opposite to link budget calculation.
--- End quote ---
But when you exceed these limits it may leads to a significant change, include directivity change, gain change and efficiency change. And as result path loss change.
--- End quote ---
Yes, just as I did wrote above. It is NOT a part of antenna loss path and it is not anything included in Friis eqation and neither usable information for calculation of path loss.
It is a result of poor or malfunction in a local radio system if antenna system is degraded by its own Tx system.
Such things are of limited interest when as example, calculating wave propagation for a cell phone system, as an example where antenna path loss calculation is of high importance when calculating needed number of base stations and where to place them.
--- Quote from: ejeffrey on June 16, 2021, 08:44:08 pm ---active devices like amplifiers are rarely reciprocal,
--- End quote ---
That is true, but as these are very seldom found between antenna ports, is it no need to include these when calculating path loss between two antenna ports and it would anyway be a bad idea with active circuits placed between antennas as there is no power supply in free space.
If such amplifiers or any else non-reciprocal element had been normal occurring in free space had Friis equation added a note about that: "look out for amplifiers between the both antenna ports".
Friis equation is very clear at that point. It is a measurement from port to port and not even antenna transmission losses, reflections, diffraction or indirect scattering are included, neither any non reciprocal amplifiers.
It is a real world measurement from antenna port to antenna port just as you can simulate S21 in CST or similar software between two antennas.
Revers same measurement in CST and tell if S12 not have same value.
It is a useful value to calculate as later selecting of needed Tx power, possible bandwidth and modulation type can be critical depending on if satisfying coverage will be reached with proposed antennas.
[ Specified attachment is not available ]
See bottom attached image.
This is an example where path loss is calculated and needed to be known before hardware is constructed and sent away.
It is very big differences between earth an mars parabolic antennas gain and it is almost 20 dB difference in receiver sensitivity for Orbiter and earth equipment but S12 antenna path loss value is equal to S21.
Please define if you is adding statement about how path loss and Friis equation is defined and eventual flaws with its calculation or if you describe something completely else, such as a malfunctioning overpowered transmitter and its poor designed power handling in its non reciprocal components.
We ALL know that such problems exist and most of us do even know that amplifiers seldom are reciprocal between its ports (S12 differ from S21)
but it is NOT a part of antenna path calculation and is not even distant covered by Friis equation as opposite is clearly declared.
Try to understands some of the text written here: https://resources.system-analysis.cadence.com/blog/msa2021-understanding-the-circuit-and-antenna-reciprocity-theorem
For me are these calculation real world, normal thing to calculate and of practical use. Faulty designed transceivers with corona discharges and saturated transformer cores not so much. It is probably more a thing for a repair shop.
T3sl4co1l:
--- Quote from: E Kafeman on June 17, 2021, 01:16:04 am ---That is true, but as these are very seldom found between antenna ports, is it no need to include these when calculating path loss between two antenna ports and it would anyway be a bad idea with active circuits placed between antennas as there is no power supply in free space.
--- End quote ---
For example, pumped laser media has an amplifying effect; but as far as I know, it's reciprocal, acting on whatever wave you give it -- so you still need iso/circulators to do that anyway. The fields equivalent of a tunnel diode of matched (negative) impedance, it just sits there doing its thing.
Tim
Navigation
[0] Message Index
[*] Previous page
Go to full version