Impedance matching, directionality, and gain

[retired17]

Originally Posted by No static at all

I agree as well. Just a few inches can make a world of difference, even with a strong signal. UHF can be very fickle.

Just registered. I was using a CM4228 8-bay antenna to get the UHF channels because I believed it would help my reception plagued by multipath. Some channels were strong and some were weak. I would move the antenna a few inches and the weak ones became strong but the strong ones became weak. I could never find a spot where all stations were strong. I read on HDTVPrimer that a large antenna (which has a narrower beamwidth) could actually be more lossy than a smaller antenna. Thus I tried the CM-4221 4-bay antenna and to my surprise all the UHF channels were strong (high S/N), some even stronger than before. Didn't have to move the antenna at all.

My theory of why I got better reception is that the radiation pattern across the smaller antenna was more uniform than across the larger antenna. According to the reciprocity theorem, the multipath signals received creates a very complex signal that is not matched to the transmission line thereby causing reflection which is loss. The losses must have been greater than the 3 dB increase in gain. The received signal from the smaller antenna better matched the transmission line resulting in less loss.

Sorry for being long winded. Your comments or thoughts are welcomed.

[Dave Loudin]

Originally Posted by retired17

My theory of why I got better reception is that the radiation pattern across the smaller antenna was more uniform than across the larger antenna. According to the reciprocity theorem, the multipath signals received creates a very complex signal that is not matched to the transmission line thereby causing reflection which is loss. The losses must have been greater than the 3 dB increase in gain. The received signal from the smaller antenna better matched the transmission line resulting in less loss.

You are onto something, but don't think of it as a match to the transmission line. Think about how the signals received by the different elements add up in strength and in phase (the underlying principle of antenna design.) This page at Ken Nist's site has excellent pictorial descriptions of what you're after.

John, I believe I understand what you're thinking. However, that's a lot more metal in the air than necessary. Better to have additional elements cut to the appropriate, i.e. smaller, wavelengths.

[retired17]

Originally Posted by Dave Loudin

You are onto something, but don't think of it as a match to the transmission line. Think about how the signals received by the different elements add up in strength and in phase (the underlying principle of antenna design.) This page at Ken Nist's site has excellent pictorial descriptions of what you're after.

Just to clarify what I said. I looked at Ken Nist's site and he does discuss why larger antenna sometimes don't work any better. It's due to the non-uniform field. Imagine that you are transmitting from your antenna. What signal do I have to feed it to produce all those multiple multipath signals coming from different directions? A very complex signal with significant imaginary components along with the real components. This implies the antenna impedances vary significantly. Based on reciprocity, the reception of the multipath signals also has the same complex impedances. If there is a poor match to the 300 ohms or 75 ohms transmission line then there is reflection (known as high VSWR). This was what I was trying to say in my original post.

[GroundUrMast]

Originally Posted by retired17

... A very complex signal with significant imaginary components along with the real components...

What components of a radio wave are imaginary?

[Dave Loudin]

You've challenged me to rethink things through carefully. Let me begin with what I believe are the facts involved and go from there.

1) even though we can use complex numbers to represent them, what we're ultimately dealing with are modulated sine waves.

2) the impedance of the antenna does vary by frequency, but for a single frequency it remains the same no matter the direction the signal comes from.

3) therefore, the ratio of power delivered into the transmission line remains the same no matter the arrival angle.

4) signals that take different paths to the receive antenna are no longer in time, or phase, sync. A phase delay has been added. Additionally, indirect paths have increased variability in phase and polarity.

My view of what is happening is the multiplicity of signals out of phase with each other march down the coax until they hit a detector. It is the mixing of these signals in the detector that causes the problem. As an example, when the skywave and ground wave signals of an AM station are both present at a receiver, the effect of the delay between the two is a slow selective fade. A notch filter runs across the passband, and when it hits the carrier, distortion like over modulation results. I'll have to brush up on the modulation scheme used in ATSC to comment on why a spectrum that has non-uniform amplitude is a bad thing.

Reciprocity holds here only in the sense that if I turn around and transmit a single signal, it will arrive back at the transmitter in a hodge-podge from scattering and multiple-path. You can't steer a frequency by changing its properties. You have to add delay across the antenna.

[GroundUrMast]

Per Wikipedia, an imaginary number is: "... a number that gives a negative result when squared..."

I'm quite comfortable using the term "imaginary" in the context of the math... but I'm not aware of current, voltage or magnetic flux being described as imaginary.


I'm also comfortable with the idea that many wave fronts can arrive at an antenna though there is only one source. There can be many paths of various lengths and may have fixed or variable attenuation in addition to still other signal effecting factors.

My understanding is that all of the magnetic flux, from all sources, impinging on a conductor at one moment, may be summed algebraically to then determine the change vs time that will induce current in the conductor. Even though many magnetic fields may be influencing the conductor, only one current is induced. This would be reasonable given that energy has changed state, from a magnetic field to an electrical current.

I see multipath beginning at the transmitter and ending at the antenna, not the junction of the antenna and transmission line or later in the tuner. But yes, the tuner will have to deal with the effects of multipath because it will have possibly modified every measurable attribute of the signal. Multipath can be viewed as an interfering source of modulation. It will obviously alter the amplitude. Doppler shift can result from reflection off of moving objects, and phase shift would result as well. Perhaps CW would be the least likely modulation.

Yes the gears are slowly squeaking in this old head...

the impedance of the antenna does vary by frequency, but for a single frequency it remains the same no matter the direction the signal comes from.

Agreed. And the impedance will not change if the amplitude of a given signal varies.

My view of what is happening is the multiplicity of signals out of phase with each other march down the coax until they hit a detector.

I believe fourier analysis gives you a mathematical tool set to analyze from that view... though in a real series circuit, the current is equal at all points, (phase not withstanding).

[retired17]

Originally Posted by GroundUrMast

What components of a radio wave are imaginary?

My bad, I should have used the word quadrature instead of imaginary. I was not referring to the radio waves in the air, but how a RF signal travels along a transmission line and antenna.

I need to think about what Mr. Loudin said. I have been challenged, too.

[andy.s.lee]

A Fourier analysis will probably not help too much in analyzing multipath since you'd be looking at the frequency domain instead of the time domain. Multipath is primarily a time domain problem.

You can actually get the same effects as multipath within the coax itself if you have dangling ends (unterminated) or impedance mismatches that create reflections inside the coax network.

In real-world multipath, the line-of-sight signal (even if severely attenuated by trees and buildings) is always the first to arrive at the antenna (shortest distance between two points). All multipath reflections arrive later than that since they travel a longer distance.

From the point of view of the receiver, it all looks like delayed copies of the same signal interfering with itself.

There is one aspect of multipath that is frequency dependent, and that is when it comes to frequency selective fading. If you happen to have two strong signal paths of nearly equal strength reaching your antenna and they happen to be 1/2 lambda out of phase, then the signals will cancel each other out to some degree. This kind of fading is very sensitive to the relative path lengths, so moving the antenna a few feet is often enough to get out of a fading situation.



To some degree, digital tuners are actually able to "undo" some of the damage caused by multipath. As it turns out, multipath is related to one of the problems faced by distributed transmitter (DTx) systems that some broadcasters have been experimenting with. Sorry for the long read, but a more in-depth discussion of this is available in this thread: http://forum.tvfool.com/showthread.php?t=101



Also, just to get even more thought gears churning, multipath can even be useful in some situations. In the world of WiFi, in particular 802.11n and MIMO technologies, multipath is actually a good thing because they can take advantage of that spatial diversity. When you have multiple antenna elements and multiple digital receivers (a mini "array" of antennas), you can use signal processing to separate each of the multipath components and treat each "spatial channel" as an independent data stream.

[retired17]

Thank you all for your thoughts and comments on my original post on why my CM4228 (8 bay antenna) did not give me better results than the CM4221 (4 bay antenna).

I was directed to HDTVPrimer hosted by Ken Nist and I think I found the answers to my question. There are two reasons for degradation. Ken said that if two colocated antennas are not in a uniform field than you would not get the expected 3 dB gain when combined. Some of the stronger signal from one antenna will be reradiated out of the other. The problem is in the combiner and an explanation of why this happens is provided by Ken Nist.

The second source of loss is that a larger antenna with its narrower beamwidth will only see one multipath signal thereby losing the hot spot advantage. He said hot spots are good for reception since the hot spot signal could be up to 6 dB stronger. You lose some of this 6 dB if the antenna only sees one multipath signal.

My CM4228 antenna is composed of two colocated CM4221s mounted side-by-side with a combiner so it must be behaving in the manner Ken describes.

That is my understanding and I am satisfied with his explanation. Thanks again to Dave for directing me to Ken Nist's website.

[GroundUrMast]

...Sorry for the long read...

No way! It would be nice if you were able to weigh in more often. Thanks for keeping a tidy forum.

[andy.s.lee]

If you are talking about the newer CM4228HD (horizontal rods as reflector), then Ken has a separate page describing how the phasing lines between the two halves of the antenna ruin its performance: http://www.hdtvprimer.com/ANTENNAS/TemporaryPage.html

The old 4228 (wire mesh as reflector) was actually better than the 4221, but that's no longer the case now that the design has been changed.

[mtownsend]

Originally Posted by retired17

if two colocated antennas are not in a uniform field than you would not get the expected 3 dB gain when combined. Some of the stronger signal from one antenna will be reradiated out of the other.

A non-uniform field might be a valid explanation for one of your channels, but it's generally unlikely to be the case for multiple channels at the same time. The "hot" and "cold" spots described in Ken's discussion about non-uniform fields usually only occur when you have two multipath signals that are almost equal in strength. These equal, yet out-of-phase, signals can cancel each other out almost perfectly, creating the "cold" spots spaced several feet apart.

In most multipath situations, the dominant paths are not equal strength. It's even less likely that multiple channels will have this effect at the same time. The reflection/absorption of RF energy varies by frequency, so the behavior of multipath is often different from one channel to the next.

The second source of loss is that a larger antenna with its narrower beamwidth will only see one multipath signal thereby losing the hot spot advantage.

It is generally good for reception to be able to isolate one multipath component out of many.

If there equal-strength multipath components creating the "hot" and "cold" spots, then yes, it's possible for a larger antenna to not-quite hit the "hot" spot, but as stated above, equal-strength multipath components are generally not that common.

Even if you had a hypothetical situation where you had say the line-of-sight path (to many co-located transmitters) plus a giant perfectly reflecting wall off to one side creating near equal-strength reflections across all of your channels. In such a situation, the "hot" and "cold" spots would occur for each channel at a slightly different location (as your hypothesis was suggesting). In that situation, you could use the higher gain antenna (with narrower beam width) to point at just one of the signal paths (i.e., the line-of-sight path), thus reducing the cancellation effect of the other signal path. This would make reception more uniform across all of the channels. Some of the channels will no longer perfectly capture the "hot" spots (possibly 3 dB worse than before), but by the converse analogy, none of the channels will perfectly hit the "cold" spots either.

In the majority of cases, it's better to have a tight antenna pattern to focus on the strongest clean signal path. By doing so, the system becomes less sensitive to the channel-by-channel multipath-induced fading variations that can occur.

This doesn't work if you have transmitter clusters in two different directions and need a wide antenna pattern to capture signals spread in multiple directions.

I was using a CM4228 8-bay antenna to get the UHF channels because I believed it would help my reception plagued by multipath.

Do you know what is causing multipath in your area (tall buildings, trees, hills)?

Were you using the new 4228 or the old one?

Do you have transmitters spread in multiple directions?

[retired17]

@mtownsend,

Do you know what is causing multipath in your area (tall buildings, trees, hills)?
My multipath signals are coming from a ridgeline that is blocking my LOS to Mt. Wilson in Los Angeles.

Were you using the new 4228 or the old one? I am using the old one.

Do you have transmitters spread in multiple directions?

All signals are coming from the same direction.

[mtownsend]

Originally Posted by retired17

My multipath signals are coming from a ridgeline that is blocking my LOS to Mt. Wilson in Los Angeles.

Thanks for the info. Do you have a link to a TV Fool signal analysis you can post?

I'm familiar with the Los Angeles area. If you are blocked by mountains, then you are probably better off with an Antennas Direct 91XG for UHF and a Winegard YA-1713 for high-VHF (KABC, KCAL, KTTV, and KCOP).

The 91XG mounting bracket allows for the antenna to pointed nose-up toward the ridge line. This puts the heart of the antenna pattern up where the signal is really emanating from.

The high-VHF antenna usually does not need to be tilted because 1) VHF signals can diffract over mountains better, and 2) the vertical beam pattern of these Yagis is very tall so you wouldn't notice much of a change anyway.

The two antennas should be combined through a Pico Macom UVSJ to prevent any mutual interference between them.

[retired17]

Originally Posted by mtownsend

Thanks for the info. Do you have a link to a TV Fool signal analysis you can post?

Here is my TV Fool link:

http://www.tvfool.com/?option=com_wr...8d17d4624564ff

Thank you for your suggestions. I am using my CM4221 with a Y5-7-13 VHF antenna. They both have 300 ohms outputs so I am using a preamplifier with separate 300 ohms inputs to combine them.

I have a signal meter on my TV and when I was using the CM4228, the meter would typically show around 2/3 full scale. With the CM4221, it is almost full scale. I would guess the CM4221 sees more multipath signals thus I get the bonus of hot spot boosts.

[retired17]

Originally Posted by Dave Loudin

You are onto something, but don't think of it as a match to the transmission line. Think about how the signals received by the different elements add up in strength and in phase (the underlying principle of antenna design.) This page at Ken Nist's site has excellent pictorial descriptions of what you're after.

I was looking at Ken Nist's hdtvprimer site where he was explaining why the net gain of an antenna may be less than the raw gain of an antenna. He stated that because of the complex impedances, there is a transmission line mismatch and some of the signal is reflected and re-radiated. I don't understand why but the raw and net gain plots of the CM4221 look the same to me but the plots of the CM4228 were different. Could be something in the CM4228 combiner plus the transmission line mismatch.