Some new insights on non-reciprocal propagation ("the diode effect")

[AB2EZ]

I finally found some useful information that can explain the empirically-observed lack of reciprocity ("I can hear you, but you can't hear me") that we often experience during the daytime hours (and sometimes during the early evening hours) on 75 meters.

Reminder: The Law of Reciprocity is a theorem, based on the laws of physics, of the form: "If a,b, and c are true... then the communication path will satisfy reciprocity".

For long haul communication, where the signal hits the ionosphere are a fairly shallow angle, and is bent by the ionosphere the produce a reflection, the "if" part of the Law of Reciprocity holds pretty well... so therefore such paths exhibit reciprocity.

At the other extreme, we know that we can buy microwave devices called circulators... which do not exhibit reciprocity. I.e., you put a signal into port A, and it comes out of port B. You put a signal into port B and it comes out of port C. You put a signal into port C and it comes out of port A.

What about the "if" part of the Law of Reciprocity is not satisfied (intentionally) by a microwave circulator? Well, a microwave circulator includes a permanent magnet, and a material (dielectric) that produces a phenomenon called Faraday rotation when it is placed in a fixed magnetic field. This means that an rf signal passing through the dielectric material will have its polarization rotated clockwise when the rf signal is traveling in one direction, and counterclockwise when traveling in the opposite direction. This (non-reciprocal) Faraday rotation phenomenon is, in turn, caused by time-varying properties of the dielectric (magnetic moments that are precessing like tops), as a result of the presence of the fixed magnetic field of the permanent magnet. A time varying material (i.e., one that induces Faraday rotation) does not satisfy the "if" part of the Law of Reciprocity... i.e., the Law of Reciprocity states: If a propagation path has properties which are not varying with time, and.... (other "if" items)... then the propagation path will exhibit reciprocity.

So what does this have to do with "the diode effect" on AM?

Well, for typical paths between NJ and New England (for example) the mode of propagation is Near Vertical Incidence Skywave propagation... in which the transmitted signal goes (nearly) straight up and, as a result of various physcial mechanisms, is reflected (nearly) straight down.

When a signal passes upward (or downward) through the ionosphere, it experiences Faraday rotation... an effect that is well known in satellite communication.

From:
    http://esa-spaceweather.net/spweather/Alpbach2002/Arbesser%20communications%20and%20navigation.pdf

I obtained the following formula:

θ = 236 x B x NT x f **–2

Where: Theta is the angle of Faraday rotation (radians); B is the average value of the earth's magnetic field (in Webers per square meter); NT is a constant whose value depends upon the state of the ionosphere, and which ranges from around 10**16 to 10**19; and f is the radio frequency (in GHz).

Since the amount of Faraday rotation is inversely proportional to the square of the radio frequency of the signal being rotated, the effect is much larger at 3.8 MHz than it is at 3.8 GHz (about a million times larger)

Thus, for NVIS propagation, it is plausible that the "if" part of the Law of Reciprocity is not satisfied (because of Faraday rotation in the path), and therefore that the Law of Reciprocity does not apply.

Stu

P.S. I know that some readers of this posting have other theories of what causes "the diode effect". For example, higher levels of absorption at one end of the link vs. the other end of the link. However, that, in and of itself, would not make the link non-reciprocal... i.e., the link would still satisfy the Law of Reciprocity if the absorption were not uniform along the path through the link. 

[Bacon, WA3WDR]

Imagine that the polarization twist on some signal path holds steady for a while at 45 degrees clockwise, and that the antennas of two stations also differ in polarization by 45 degrees. Station one will hear station two just fine (perfect polarization alignment), yet station two will hardly hear station one at all (crossed polarization).

Terman stated that reciprocity always held, except in a plasma medium in the presence of a magnetic field. I read that, and I thought, "But... But... that describes the ionosphere!"

[AB2EZ]

Mark's question and Bacon's amusing quote* are both "right on target". At this point, it is surprising to me that reciprocity appears (by experience) to hold so well for long distance, multi-hop, shallow-angle-reflection-based propagation. It would be nice to see a complete analysis, comparing the physics of NVIS and shallow angle reflections, that would explain what's really happening in each case.
Stu

* I look forward to having an opportunity to use that quote in some context.

[KA1ZGC]

So what does this have to do with "the diode effect" on AM?

Well, for typical paths between NJ and New England (for example) the mode of propagation is Near Vertical Incidence Skywave propagation... in which the transmitted signal goes (nearly) straight up and, as a result of various physcial mechanisms, is reflected (nearly) straight down.

When a signal passes upward (or downward) through the ionosphere, it experiences Faraday rotation... an effect that is well known in satellite communication.

(...)

Since the amount of Faraday rotation is inversely proportional to the square of the radio frequency of the signal being rotated, the effect is much larger at 3.8 MHz than it is at 3.8 GHz (about a million times larger)

Thus, for NVIS propagation, it is plausible that the "if" part of the Law of Reciprocity is not satisfied (because of Faraday rotation in the path), and therefore that the Law of Reciprocity does not apply.

That's funny, we've always blamed it on higher noise levels in your neighborhood, Stu! 

--Thom
Killer Agony One Zipper Got Caught
p.s. Sorry, couldn't resist.

[steve_qix]

What is interesting about the so-called "diode" effect, is that I really only notice it during the day.  Maybe this is due to the weak nature of the signals during the day (on 75 meters).... Anyway..

Ok, so we have the D layer at 40 to 55 miles, and the E layer at about 60 to 100 miles above the earth.

Any reflection is probably taking place in the lower E layer during the day - say 80 miles for round numbers.   For a signal to propagate from Northern Massachusetts to Northern New Jersey - what is that distance - maybe 200 miles - the signal has to have a traverse some 256 miles through the atmosphere.  Figured (roughly): 2 triangles with a height of 80 miles (the E layer), and each base of 100 miles (1/2 the horizontal distance from Northern MA to NJ)...  Assuming the numbers are close, the takeoff angle will be less than 45 degrees (38.6 degrees to be exact).

More to the question about the "diode" is this:  Is the reflection equal - or does the ionosphere "bend" the waves more than one would get during a straight reflection.  In this way, a lower angle signal may be bent to earth at a much steeper angle than would occur during a reflection.

Just some musings.

This is an interesting topic.

Regards,

Steve

[Bill, KD0HG]

Imagine that the polarization twist on some signal path holds steady for a while at 45 degrees clockwise, and that the antennas of two stations also differ in polarization by 45 degrees. Station one will hear station two just fine (perfect polarization alignment), yet station two will hardly hear station one at all (crossed polarization).

Terman stated that reciprocity always held, except in a plasma medium in the presence of a magnetic field. I read that, and I thought, "But... But... that describes the ionosphere!"

I have never heard of or observed a cross polarization effect on HF. 50 MHz and up, yes, but nothing seems to remain linear after ionospheric propagation on HF. It's all scrambled. I look at the ionosphere layers as 'benders' rather than mirror-type 'reflectors'. Spray a garden hose at a sheet of plywood and what bounces off the sheet in no way resembles what hit it. That's probably a better analogy of ionospheric propagation than optics and a mirror.

Also, on that hypothetical 200-mile path, the stations might be close enough to experience destructive or constructive interference between the space wave and sky wave. I do believe the atmosphere can bend, enhance or extend MF and HF signals, not to the extent of what happens on VHF, but to some degree. No one has ever established a lower frequency limit for atmospheric effects on radio propagation (AFAIK)...

[KA1ZGC]

No one has ever established a lower frequency limit for atmospheric effects on radio propagation (AFAIK)...

Well, sort of. If by "atmospheric" you mean "ionospheric" (and I'm assuming you do), then there is such a thing as Lowest Usable Frequency (LUF) between two points, which rises over the course of the morning and falls over the course of the afternoon. Unlike the MUF, it hits its lowest value some time after sunset and pretty well stays there until just before sunrise.

More to your point would be the question of how low the LUF can go. Of course, like the MUF, it all depends on where points "A" and "B" are.

Once you start getting below AM broadcast, the ionosphere plays a diminishing role. WWVB, for example, runs many wristwatches and wall clocks all across the country, and I've only ever seen very slight variations in signal strength into Albion over the years. WWVB is propagated almost entirely by groundwave.

--Thom
Keep Away One Zorched Ground Conductor

[Bill, KD0HG]

No one has ever established a lower frequency limit for atmospheric effects on radio propagation (AFAIK)...

Well, sort of. If by "atmospheric" you mean "ionospheric" (and I'm assuming you do), then there is such a thing as Lowest Usable Frequency (LUF) between two points, which rises over the course of the morning and falls over the course of the afternoon. Unlike the MUF, it hits its lowest value some time after sunset and pretty well stays there until just before sunrise.

More to your point would be the question of how low the LUF can go. Of course, like the MUF, it all depends on where points "A" and "B" are.

Once you start getting below AM broadcast, the ionosphere plays a diminishing role. WWVB, for example, runs many wristwatches and wall clocks all across the country, and I've only ever seen very slight variations in signal strength into Albion over the years. WWVB is propagated almost entirely by groundwave.

--Thom
Keep Away One Zorched Ground Conductor

Nope, I mean atmospheric effects on shortwave radio propagation...Like ducting, inversions, etc.
We're familiar with such on TV, FM, 6 and 2 meters, but I believe there are effects that go down to the broadcast band, which includes the HF spectrum.

I can't prove it, but I do believe you can get atmospheric ducting and other similar weather system effects affecting 75-10 meter propagation..

An old buddy of mine who has moved away from the area once ran a strip chart recording of the carrier signal level of 50 KW 670, WMAQ out of Chicago here in the Denver area. On a big loop or beverage, there is always a carrier from them, even during the day over the 1,000 mile path. Scott claimed there was an observable correlation between their signal level and mesoscale weather systems. The test had to end after a small local daytimer was licensed on 670. Before that, there wasn't a single other occupant of that 670 clear channel frequency between Denver and Chicago.

[AF9J]

Bill,

The only ducting I know of that occurs at HF, is ionospheric ducting.  For tropospheric ducting to work at HF, the duct would have to be of world record proportions.  With regards to radio - tropo phenomena (ducting, inversions, etc.) start higher in freq. (because the shorter wavelengths are more easily effected by physically smaller sized tropo refracting phenomena), and work their way down frequencywise.  I can remember plenty of times where I had an easier time working tropo on 432, than I did on 2m.  Tropo on 6m is VERY hard to work.  Tropo ducts seldom happen on 6, and mainly on coastal paths.  By the time they do occur, the tropo's been going hot and heavy for some time, on the higher bands.  I've never heard tropo on 10m. 

I just find it hard to believe that mesocyclonic conditions could affect HF propogation, unless they have very steep air pressure gradients, that are conducive to refracting HF sized wavelength RF.  This is the only way I could see it working.   Even then, the propogation would be more troposcatter in nature, and like tropo scatter, the power levels used for decent effectiveness, would have to be quite high.  As it is, the only mesocyclonic systems I know of that offer any enhancement to tropo propogation are stalled out high pressure centers, and RF behaves tropowise with regards to these, in the manner I mentioned above.

73,
Ellen - AF9J

[K1JJ]

Any reflection is probably taking place in the lower E layer during the day - say 80 miles for round numbers.   For a signal to propagate from Northern Massachusetts to Northern New Jersey - what is that distance - maybe 200 miles - the signal has to have a traverse some 256 miles through the atmosphere.  Figured (roughly): 2 triangles with a height of 80 miles (the E layer), and each base of 100 miles (1/2 the horizontal distance from Northern MA to NJ)...  Assuming the numbers are close, the takeoff angle will be less than 45 degrees (38.6 degrees to be exact).

Yes, I'd agree.  I think the "optimum" angle on 75M during the day is very low. When I hear Tron on from Maine during the day, (300 miles) he is usually weak on the high angle dipole at 60'.  But when he gets on his low angle wire array and I use my quads at 190', we often see a 15-20db difference.  This indicates angles of around 25-30 degrees.

However, during the evening, the angles are higher cuz of the higher reflection layer. Many times we see little difference on the arrays vs: the low dipoles for 300 miles.


About long range propagation on 75M...  it astounds me how at our sunset during December, the Ja's will come through on 75M. Sometimes they are S9 +20 over. This must take many hops to pull off and indicates to me there is some kind of HF ducting going on. When it's ducting, the signals are big... when not, they are gone. If it were standard hops, you would think the attenuation on 75M would be greater over this 11,000 mile path. (The path is all ocean - South Atlantic over the S. Pole which does help.) There is a 50KW JA beacon on 3925 - a Tokyo BC station. This is a great way to check prop to JA.


Here's my guess on one area of the diode effect on 75M DX....  I see it happen quite often while still in daylight while Europe is in  darkness. Sometimes we can hear them strongly, but they cannot hear us at all. It's the opposite before our sunrise - they hear us but we have trouble with them in their daylight.  I think this may have to do with the first hop reflection. In darkness, you have a good first reflection. But the other station, if in daylight, may have a poor daylight layer to work with on the first reflection. What would happen if the path was just ONE hop and you had a good first reflecton and the other station did not?   ie, Both stations may not be using the same area of the sky both ways...  This is exaggerated when the two stations are using different antennas, with lower or higher angle takeoffs.  


T

[Bacon, WA3WDR]

The polarization of downcoming signals on HF is indeed jumbled, but any given signal goes through varying jumble states over time.  See "The Enhancement of HF Signals by Polarization Control" by B. Sykes, G2HCG, in Communications Quarterly, November, 1990, or find the original from Practical Wireless.

Sykes showed how incoming signals from 10 and 15 meter beacons went through various angles of linear polarization, and various circular and elliptical polarizations.  It was his conclusion that most of the short-term fading we see while the bands are 'open' is caused by downcoming signals varying between optimum polarization match and cross-polarization mismatch, with respect to the receiving antenna.  My own limited experience tells me that he was correct.

Something that I consider to be remarkable is the coincidence of the angle of incidence and refraction, and the time delay associated with the signal path.  At frequencies close to the MUF, the signal path may appear to involve a remarkable height, but it seems that this apparent height is caused by a slow propagation mode in a much lower refracting layer - and yet, the time delay associated with the signal path is substantially equal to the time delay that would be expected for the virtual layer height, based on the vertical angle of the signal path at the surface of the earth.

[AB2EZ]

Of note

Non-intuitive as it may be... a multi-hop channel with varying amounts of attenuation along the path, still satisfies the "if" part of the Law of Reciprocity. To get non-reciprocity, you need some effect that doesn't satisfy the "if" part of the Law of Reciprocity. You need a truly non-reciprocal effect like Faraday rotation.

Radio waves do not (when you look at the details of the physics) bounce off of the ionosphere like reflections from a mirror. However, if you think of the ionosphere as a black box, you have a wave traveling upward toward the ionosphere and (under the right conditions) a wave traveling downward from the ionosphere... that is equivalent of having a mirror at some point within the black box. A key issue is whether or not there are any non-reciprocal effects, like Faraday rotation, that actually make the end-to-end path loss non-reciprocal.

Stu

[AF9J]

Understood.   BUt doesn't Faraday Rotation typically exhibit its effects, the higher into the ionosphere (or longer duration in it) a signal goes?   Hence the reason why you really don't hear about pronounced Farady Rotation, unless you're doing EME?  If I remember right, Farady rotation is more of a lens type refraction effect (akin to the inversion of a light image passing through a lens).

Ellen - AF9J

[w4bfs]

hmmmm ... I wonder if rf can tunnel or some other quantum effect?

[Steve - WB3HUZ]

Take a look at the magnetic/auroral oval. This has a huge effect on seemingly one way propagation to Europe on 75 meters.

I think most propagation on 75 meters other than at times of high D-layer absorbtion or over very high angle paths is non-reciprocal. The very nature of the ionosphere dictates.