Some Antenna Photos and Tests

[tripelo]

Time: Summer/Fall 2010
Location: KY

Yagis (channel-cut for 10, Wade 10y10s) were used as raw material and a starting point in the design. Channel 10 was selected because it seemed a good compromise for the driven element between Ch. 8 and Ch 11. Wade Yagis used a 3-bar folded dipole for the driven element. Such a folded dipole is probably broadband enough to cover channels 8 through channel 11 with good impedance match. Also, since channel 10 antennas generally have longer elements than channel 11 antennas, this allowed flexibility in trimming elements (longer than desired elements can be trimmed).

The front three elements from a YA-1713 were added to the channel 10 antenna. Repeated testing and 4NEC2 analysis were performed. Eventually, an extra length of tubing was added containing two more elements, repeat analysis and testing. Then, all elements were analyzed in 4NEC2 for length and location. Gradually the element positions and lengths were changed toward the 4NEC2 specified parameters, with testing in between small steps to confirm the direction of the change.

When it was thought the antenna was near optimized for channel 11, then searches were performed in 4NEC2 to determine if channel 8 could be improved without adversely affecting channel 11 performance. Few changes were available. A change was made, lengthening the reflector improved gain by more than one dB on channel 8 and reduced gain for channel 11 by maybe 0.1 dB. Measurements confirmed these differences.

Final element progression illustrated in images below:

- 13 elements
- 14 elements
- 15 elements







The antennas were made from parts of:

1. Wade Channel-cut Yagi (trimmed directors, less reflector)
2. Winegard YA-1713 (reflector)
3. Winegard YA-1713 (front three elements trimmed)
4. Portion of Channel Master boom (portion extending out front)
5. Channel Master elements (front directors, trimmed)
6. Portion of Antennacraft boom (support boom)
7. Homemade support boom brackets

The final 15 element Yagi had total length ~170 inches.

Two identical antennas were constructed.

Later, a pair of these long-Yagis were stacked and performance compared to stack of Winegard YA-1713 Yagis.

.

[Pete Higgins]

Later, a pair of these long-Yagis were stacked and performance compared to stack of Winegard YA-1713 Yagis.

tripelo,

When do we get the rest of the story? Anxiously awaiting comparison results. How close were you able to come to 7 dB gain above that obtained with a pair of YA-1713 antennas with coaxial balun's and the additional 5 elements?

[tripelo]

Originally Posted by Pete Higgins

tripelo,

When do we get the rest of the story? ...

Thanks Pete.

Glad you find the story interesting.

For the most part, kept handwritten logs, Excel files, and photos. Takes a bit of time to go back and put a 'hopefully' coherent story together.

Once upon a time...

[tripelo]

Earlier, it had become apparent that sufficient gain (predicted 7dB above pair of YA-1713) to enable reception of WHAS-11 would be difficult to achieve.

To make most use of the available gain, a pair of Research Communications Preamplifiers (Models RC-9267) were ordered. The primary rationale for this approach was noise figure related.

The advertised VHF noise figure of major manufacturers of preamps ranged from a low of 2.6 dB to around 3.0 dB. The RC preamps were advertised to have a noise figure of 0.4 dB. Signal-to-noise ratio is closely related to noise figure. As an example, at UHF frequencies, a real improvement in noise figure of say; one dB, can result in a signal-to-noise improvement of about one dB. It is more complicated than this example, and also less likely at VHF, but this can be a close approximation for TV reception purposes.

This begs the question of whether advertised noise figures are accurate. With measurements, using normally expensive equipment, the question can be directly answered. Indirect measurements at the lab bench have indicated that the major manufacturer noise figures are overly optimistic. In any event, field tests here demonstrate the RC preamp provides improved S/N ratios.

Testing with the RC preamp encountered a hitch. One day a thunderstorm arose quickly. The test station was abandoned with antennas in place and the RC preamp powered on. As the scene was observed from the back porch of the house about 100 feet away, lightning struck the earth or a tree, about 600-1,000 feet distant. After the storm passed, the RC preamplifier was no longer functioning. On the main tower, two Channel Master preamplifiers survived.

There are several complication factors in the above scenario, but it is clear that:

- The RC preamplifier can be susceptible* to atmospheric discharge.
- The coaxial half-wave loop balun contributed to the demise of the RC preamp.

*More susceptible than the Channel Master preamplifiers.

The outcome was that testing with the coaxial loop baluns combined with RC preamps was terminated for a time.



Some different brands of commercial baluns were compared to half-wave coaxial baluns on both the test bench and also installed at the long Yagis. A selected pair of Philmore MT-74 baluns provided the lowest loss at Upper VHF. The signal loss of the MT-74 baluns was ~ 0.4 dB worse on channel 11, than the half wave loop baluns. The Philmore MT-74 baluns were used with an RC-9267 preamplifier throughout the remaining tests.

* Recent purchases and testing of Philmore MT-74 baluns indicate quality of construction, materials, and performance at VHF, have significantly changed (negatively) compared to the units tested in 2010.

.

[tripelo]

Analysis with 4NEC2 software can often help provide insight into antenna modifications and resulting performance.

A computer simulation of an antenna is an approximation that is no better than the models represented in the software. The software model is based on numerical parameters that represent an antennas physical dimensions and electrical specifications. Often the dimensions are measured on a physical model. Electrical specifications may have to be determined by experiment. It is rare that measurements and specifications allow software to fairly completely describe an antennas performance, but is often close enough for practical purposes.

The performance of Winegard YA-1713 without balun, was simulated, both as a single antenna and stacked at 43 inches. The plot below shows gain (dBi) and Standing Wave Ratio* (SWR) over the upper VHF band. For a single antenna, the maximum gain is about 12 dBi, which would be close to Winegard specifications of ~10 dBd.



Interesting that stacking at 43 inches shows higher gain (~2.9 dB) at the low end of the band. Recall that a stack distance of 43 inches was empirically determined as best (within +/- 18 inches) for reception of channel 13. The real world includes ground effects (mainly reflections) that were not included in the simulations. Variations in height above ground can dramatically affect antenna patterns, thus affecting perceived gain. The change in SWR from single unit to 2 units stacked, could be a result of mutual impedance or coupling between the two antennas. A stack distance of 43 inches is relatively close, and some mutual coupling could be expected.

The performance of the long Yagi without balun was simulated, both as a single antenna and stacked at 89 inches. The plot below shows gain (dBi) and SWR (relative to 300 Ohms) over the upper VHF band. For a single antenna, the maximum gain is ~>14 dBi, about 2 dB greater than the YA-1713. Stacking at 89 inches shows additional gain improvement approaching 3 dB over most of the band.



The following plot shows the mismatched gain for both the stacked pair of YA-1713 and the stacked pair of long Yagis.



Mismatched gain is the normal gain (shown earlier) that has been reduced by the effects of SWR (normalized to 75 Ohms). The numerical value of SWR represents the mismatch to the characteristic impedance of the system. In this case, the characteristic impedance is 75 Ohms. Any impedance presented by the antenna other than 75 Ohms reduces the amount of power that can be transferred from the antenna to the transmission line or to a system (preamplifier, receiver etc). The amount of power rejected due to mismatched impedance is the mismatch loss. Mismatched gain is the full gain minus the mismatch loss. Mismatch loss is present to some degree in all systems with SWR greater than 1. It appears this final gain figure is the equivalent of what has been called ‘Net Gain’ by Ken Nist, at HDTVprimer.

The mismatched gain shown in the graph is probably realistic for a system with good SWR with respect to the transmission line and tuner, or preamp (if one is used). If the transmission line and tuner, or preamp, have poor SWR with respect to each other, then the overall mismatch is statistically likely to be worse with greater mismatch losses causing gain for the system to be less than that shown.

* Standing Wave Ratio (SWR): A measured or calculated number that mainly represents an impedance mismatch. SWR=1 represents a perfect impedance match, numbers greater than 1 represent progressively worse impedance match. SWR is an indirect way to describe ‘Return Loss’. To have a ‘Return Loss’, there does not have to be a transmission line physically long enough to support an actual standing wave length.

.

[Pete Higgins]

tripelo,

Clear, concise & very interesting. Your installments keep me anticipating and coming back for more. At my age, I hope you make some miraculous breakthroughs on aging but I’m not sure your talents wouldn’t be better served back in Engineering!

[tripelo]

Originally Posted by Pete Higgins

...Clear, concise & very interesting. Your installments keep me anticipating and coming back for more.

At my age, I hope you make some miraculous breakthroughs on aging...

Soon, I will be your age. Relatively soon, all now alive will be your age.

Human aging is a scientific and engineering problem, it will be solved.

... but I’m not sure your talents wouldn’t be better served back in Engineering!

Thank you Pete for your kind remarks.

[tripelo]

Previously shown were computer simulations of antennas; Winegard YA-1713 and homebrew long Yagi. For the channels of interest (Channel 8 and 11), the analysis data showed gain for a pair of long Yagis compared to the stacked pair of YA-1713 to be as listed in table below:



Having finalized construction of the long-Yagi, several comparison tests at the test range were conducted.

A test involved an antenna-under-test (AUT) mounted on the test mast, receiving signals from the transmit antenna. The transmit antenna was a portion of a YA-1713 mounted on the main tower (aimed towards the test mast) fed with a crystal oscillator as a transmitter. The frequencies available were the 9th and 10th harmonic of a 20 MHz crystal (180 MHz and 200 MHz). The AUT signal was amplified by RC-9267* preamplifier then received on a Blonder Tongue FSM-11 (signal level meter). Signal levels at 180 MHz and 200 MHz were recorded.

*RC-9267 – Particular interest due to specified low noise figure. This model has band pass filtering to attenuate frequencies outside upper VHF (174-216 MHz).

As with previous testing over several weeks, the test started with the homebrew long Yagis stacked at 89 inches. The two antennas with MT-74 baluns fed an Antronix CMC2002U splitter (used as a combiner). The combiner output was coupled to the preamp/FSM-11 receiver system as listed above. FSM-11 readings were recorded.

Then, the long yagis were changed to the YA-1713s stacked at 43 inches, using the exact cables and Winegard CC-7870 combiner that was used on the main tower. The preamplifier was a RC-9267 instead of the original CM-7777. The FSM-11 meter readings were recorded.

Measured Results:

Freq. Delta
(MHz) (dB)
180 . 3.3
200 . 4.8

That these results generally agree with the computer simulations was bit of luck.

In this test, there were some confounding variables, examples (no particular order):

- Antronix combiner on long Yagis had less loss (maybe ~0.2 dB) than Winegard 7870
- MT-74 balun on long Yagis may have had more loss than YA-1713 balun.
- Preamp SWR; briefly explored near the end of these tests.
- There are others

Suppose the gain improvement was ~4.8 dB, this falls short of the predicted needed value of 7 dB gain above that of the pair of YA-1713s, as previously discussed.

A somewhat compensating factor was that the noise figure of the RC-9267 was considerably better than of a CM-7777.

Later in 2012, lab comparisons in thermal noise background, using a Sencore 1456CM, indicated ~3dB better post-detection S/N with the RC-9267 compared to the older version of CM-7777.

There are probably few locations (certainly not this location) in the USA where the background noise at Upper VHF is as low as thermal noise. So, one would not expect an approximately 1 for 1 improvement in S/N with improvements in noise figure. But, there was some reason to think that the combinations of antenna gain and noise figure improvement might suffice for adequate reception of WHAS-11.

Over-The-Air reception at the test site with long Yagis, RC preamp, and a Channel Master CM-7000 DTV converter was observed for a few weeks. WBNA-8 reception was very solid with an occasional pixelated image, and WHAS-11 was watchable. But, signal dropouts were present on WHAS-11, especially during hours between about 10 AM and 5 PM or so. This viewing experience seemed to indicate a lack of sufficient antenna gain, or possibly some other deficiency in reception of WHAS-11.



Previously collected antennas and materials were available, so building two more long Yagis for a quad stack was feasible.

Decided to observe performance of these antennas on the main tower for a period of time.

.

[Pete Higgins]

tripelo,

This is starting to remind me of a time when we listened to the radio every Saturday night for the next installment of Boston Blackie or the Shadow –great stuff.

Yesterday, Time Warner dropped KCBS {RF 43}(channel 2.1 which carries a lot of the shows my wife & I watch, especially the evening news @5,6 & 11) KCAL RF 9 channel 9.1, plus all of our Showtime & Movie channels.

A somewhat compensating factor was that the noise figure of the RC-9267 was considerably better than of a CM-7777.

OTA, I get 9 pretty reliably but CBS {RF 43, 2- Edge with a Noise Margin of -22.1 dB @ -112.90 dBm} comes & goes especially during late afternoon early evening hours. Do you think I would see any improvement swapping out my 91XG’s PCT MA2-M 2.7 dB NF amplifier with one of the new Winegard LNA-200’s that they claim has a 1 dB NF on UHF? Or would the improvement be “in-the-noise” –so to speak?

[tripelo]

Originally Posted by Pete Higgins

...This is starting to remind me of a time when we listened to the radio every Saturday night for the next installment of Boston Blackie or the Shadow –great stuff.

Yep. “…return with us now to those thrilling days of yesteryear!…” .

Listened to radio shows such as: Dragnet, Suspense Theater, Gunsmoke, and the Lone Ranger.

Yesterday, Time Warner dropped KCBS {RF 43}...

They dropped KTVT (CBS) here in Dallas. Such as this, may in long run be good for all broadcasters.

OTA, I get 9 pretty reliably but CBS {RF 43, 2- Edge with a Noise Margin of -22.1 dB @ -112.90 dBm} comes & goes especially during late afternoon early evening hours. Do you think I would see any improvement swapping out my 91XG’s PCT MA2-M 2.7 dB NF amplifier with one of the new Winegard LNA-200’s that they claim has a 1 dB NF on UHF? Or would the improvement be “in-the-noise” –so to speak?

Hard to say.

Seems as any improvement could help with such a low signal level.

Based on my measurements of industrial/commercial drop amplifiers, of which your PCT MA2-M is one, the specified noise figure of the PCT MA2-M is probably realistic. So, if Winegard LNA-200 specifications are accurate, you could see approximately 1.7 dB improvement in S/N ratio. That is worthwhile, especially in a low signal situation as you describe.

Small improvements, even if they do not entirely eliminate dropouts, can reduce dropout time duration. On the other hand, if the dropouts are a result of severe multipath, signal level increases often have less noticeable effect.

-----------------------------

On Winegard LNA-100 and LNA-200 specifications: It is plausible that all the specifications are accurate. The information provided by Winegard suggest a design using pHEMT transistor or IC. With such GaAs FET devices available these days, all Winegard’s specifications are within reason, and preamplifiers using such devices could be manufactured at low enough cost to enable mass marketing.

The thing to watch for in pHEMT based preamplifiers is susceptibility to atmospheric discharge. To a large degree this problem is solvable, hopefully Winegard has done so.

Of course if Winegard's specifications are not accurate, as seems sometimes, then the above speculations are meaningless.

Couple of ways to help answer such questions:

- Test the preamplifier to see if it meets specifications.
- View the layout of the circuit to identify the class of active devices.

------------------------------

Pete, in your situation, I would probably give it a try, and be prepared to add it to my collection of interesting but maybe not very useful preamplifiers.

Note: Based on ADtech's comment in another forum, the LNA-200 may be more like Winegard's previous preamplifiers. In such a case, it is unlikely the preamplifier would help your reception.

[tripelo]

In October 2010, a single long-Yagi was installed at the top of the main tower mast along with an RC-9267 preamplifier. A Blonder Tongue FSM-11 signal level meter was used at the tower top to observe signal strength of both WHAS-11 and WBNA-8.

The sensitivity of the Yagi height above ground was investigated. The Yagi, with RC-9267 and FSM-11 connected, was raised from near the tower top to the upper end of the mast. The FSM-11 was observed as the Yagi was raised, and about every foot or so, the Yagi was fastened in a fixed position for longer observation of the signal level. This process was repeated until the Yagi was at the top position. While raising the Yagi (~11 feet variation in height) there were no obvious signal strength differences from one position to the next. But, signal fluctuation could easily have masked differences in average signal strength at some particular location.

Often, at any position, the signal would fluctuate rapidly, changing as much as 15-20 dB in a time period of 20 –60 seconds. In the depth of the fades, the signal would be at or below the required threshold of detection for 8VSB DTV (at around 15 dB above the noise floor). The fades were frequency selective. For example, the signal level in lower portion of the 6 MHz wide channel might be fairly strong while in the upper portion of the channel the signal could be critically weak, or vice versa.

The image below shows the long-Yagi at the main tower mast.



For several days the single Yagi was left alone while reception quality was observed at the house (using a CM-7000 DTV converter). Reception was fine for WBNA-8 (with occasional image pixelation). Reception for WHAS-11 was good for several hours of the day, but signal loss and breakup was common through midday hours until early evening.

Comment: Rapid Fluctuations and Frequency Selective Fading

The signal observations at the top of the tower with the Blonder Tongue FSM-11 were similar to those seen at the test mast location. Those rapid fluctuations and frequency selective fading were indicative of multipath. The terrain in central KY is hilly with trees, no mountains. This is the path that a 2-Edge signal (according to TVfool) from Louisville must traverse. At first thought, such a path with abundant vegetation as RF absorbers might not seem to support severe multipath. It could be that for a significant part of the time the attenuation of the path was large enough that the propagation mode was tropospheric scatter (Tropo) or a combination of Tropo and 2-Edge. The propagation mode of Tropospheric scatter is always present, but ordinarily the path is so attenuated that such signals are not seen. If this is the case, 2-Edge propagation combined with Tropo scatter may provide multiple paths that combine to cause rapid frequency-selective signal fading. Tropospheric scatter alone often supports multiple varying paths, and even without 2-Edge propagation could be sufficient to cause such fading.

The image below is an excerpt from a recent TVfool report for this location with stations of interest WBNA-8 and WHAS-11 marked.



In 2013, an opportunity arose to replicate the above antenna configuration and observe some of the signal spectrum characteristics with a Sencore SLM1456CM. The graph below illustrates the signal fluctuation.

Notes for graph, below:

- Mid-morning, sunny day, May 9, 2013
- Measurements taken at tower top
- Long-Yagi mounted at top of main tower mast
- Homebrew ferrite balun
- Homebrew pHEMT preamp with power inserter.



The frequency selective signal fading is evident at the upper portion of the 6 MHz spectrum, at about 203 MHz. Relative to Sample 1, the signal level of sample 2 shows a fade of approximately 25 dB.

.

[Pete Higgins]

tripelo,

Another interesting read.

Listened to radio shows such as: Dragnet, Suspense Theater, Gunsmoke, and the Lone Ranger.

You may have listened to the Lone Ranger but I bet you don’t still have a Hopalong Cassidy wristwatch!

I was discouraged by ADTech’s findings with his LNA-200. He’s very up-front about his test equipment and the constraints of his test environment, so even if they are contributors to his absolute measurements, his relative findings are invaluable. He had previously mentioned testing an RCA TVPRAMP1R amplifier and posted his findings for me. Like most others, it didn’t quite meet advertised but was many dB’s closer than some designs that I already own. I ordered one last week (for $22.80 delivered!) and in my environment it didn’t show any signs of overload. Subjectively, my SNR’s seem better with the RCA than with either of my Winegards. I posted his information and my results in a new thread titled “RCA TVPRAMP1R Amplifier” @ {http://forum.tvfool.com/showthread.php?t=13530} if you want to take a look. Last night, I ordered a second one to try with my tower array.

[tripelo]

The signals from a single long-Yagi were observed for a few days. Then, in October 2010, a second long-Yagi was installed on the main tower mast.

The second Yagi was installed 89 inches below the top Yagi. Before the antennas were combined, signal levels were observed on a Blonder Tongue FSM-11 signal level meter. Snap-on connectors were temporarily installed to allow quick switching of the preamplifier (RC-9267) between antennas. There was some indication that the signal level from the top Yagi was stronger than the Yagi at the lowest position. Again, as with the single Yagi, signal variation could have masked any real differences in the average signal levels.

Satisfied the antennas were performing normally, both antennas were connected to an Antronix CMC-2002U splitter (reversed, as a combiner). Signal indications on the FSM-11 showed an increase above that obtained separately from either antenna.

The antennas and connections were secured in their final positions. The configuration:

- Two long-Yagis stacked 89 inches apart.
- Two selected Philmore MT-74 baluns
- Antronix CMC-2002U combiner &
- Research Communications Preamp (RC-9267) mounted between antennas
- Suitable lengths RG-6 cable with custom connectors.

Below is an image of the antennas at the main tower.



Reception was monitored inside the house using a Channel Master CM-7000 converter. Reception of WBNA-8 was good with occasional pixellation. WHAS-11 reception seemed to be improved with the antenna stack compared to the single antenna. WHAS-11 reception was good except there were some dropouts in the troublesome daytime hours. The dropouts were less frequent than seen either at the test location with both antennas stacked, or with the single Yagi at the tower top.

Interesting: The depths of frequency selective fading,seen with the single Yagi, imply that gains in signal strength alone would not be sufficient to overcome the effects of the deep fades. Clearly, a 25 dB selective fade could not be mitigated by a 3 dB increase in signal strength (in reality, stack gain is likely less than 3 dB). Those fades shown in the spectrum (earlier) were not difficult to capture, meaning they were fairly frequent during daylight hours. With the stack of two antennas, the fades were less noticeable. The fades continued to exist but were more difficult to capture via the FSM-11. The equalizer in the tuner demodulator can mitigate the effects of some fades, but equalizers have a limited range. One limit being that the signal level at the depth of the fade must remain above the minimum required for DTV detection (~15 dB S/N) --there are other limits. It seems likely the stack of antennas provided something more than merely an increase in signal strength (theoretical ~ 3 dB).

In 2013, an opportunity arose to nearly replicate the above antenna configuration and observe some of the signal spectrum characteristics with a Sencore SLM1456CM. The graph below illustrates the signal fluctuation.

Notes for graph, below:

- Mid-morning; sunny day; May 10, 2013
- Two Long-Yagi, stacked @ 89 inches, mounted at top of main tower mast
- Homebrew ferrite balun
- Homebrew combiner (transmission line type)
- Homebrew pHEMT preamp with power inserter.
- Measurements taken in house, at converter input
- Signal conditioned: attenuator, filters, UVSJ, industrial drop amplifier, attenuator



Frequency selective signal fading is evident at the mid- portion of the spectrum, at about 200 MHz. Relative to Sample 1, the signal level of Sample 2 shows a fade of approximately 10 dB. The fading depth of ~25 dB (as shown earlier) has not yet been observed with the two antenna stack.

Reception Summary (At end of year 2010):

WBNA-8: Acceptable at all hours of the day.

WHAS-11: Near flawless during evening hours through early morning hours. During daylight hours; could be considered acceptable, but occasional loss-of-lock remains.

.

[Pete Higgins]

tripelo,

That’s a lot of separation between your long Yagi’s. Do you think that ~1.5 wavelengths of spatial diversity is mitigating your frequency selective fading? It would certainly help explain improvements in excess of simple combining gain.

I received my second RCA preamplifier on Monday. I dropped my tower and removed the PCT MA2-M’s from my 91XG and from my Antennacraft Y10 7-13. I made up new antenna cables w/boots and installed the RCA TVPRAMP1R amplifier on the mast midway between the two TV antennas. I configured it for separate inputs with the FM trap selected.

I decided to use the supplied power supply/inserter. The run from the garage to the amplifier is ~125’ through copper coated steel RG-6 so I was expecting I might encounter too much voltage drop. Turns out, it works just fine. The RCA draws <80 ma and my 3 PCT MA2-M’s were drawing almost ten times as much.

My Winegard AP-2870 overloaded with my 8-Bays on my push-up mast, but it overloaded a lot worse on the tower with the 91XG /Y10 7-13, so I was half way expecting the RCA might overload on the tower as well. No such luck. Since I installed it, it’s been rock stable with no signs of overload what so ever. When I was first setting up my system, because of the big price difference, I never would have considered buying the much cheaper RCA amplifiers. Just goes to show ya!

Subjectively, I think the RCA is outperforming the PCT drop amps. especially on VHF. Rescanning I’m picking up 1 low VHF channel, 1 high VHF channel and several UHF channels that I haven’t received in the past. Of course I’ll have to learn new languages to understand some of them. The RCA is specified to have 1 dB more gain on VHF & 7-8 dB more on UHF so that could be the reason with my long cable runs. The PCT’s are specified to have a Noise Figure of 2.7 dB (avg.) & 4.0 dB (max) with no distinction between VHF & UHF so it could be that I’m seeing a NF advantage. My main concern was with overload from my two strong local stations. So far, neither RCA amp. has exhibited any signs of overload. In the same situation, both of my Winegard’s (AP-2870 dual input & HDP-269 single input) displayed debilitating overload. I would sure like to know where the 1 dB compression point is reached with this design.

Looking forward to your next post.

Edit:

I calibrated my NTE/ECG U-106 rotor mid July. When I turned it to 090 deg. to drop the tower I noticed the array was pointing ~110 deg. I parked it so the antennas would be pointing straight up, lowered the tower and made my changes. After I raised the tower I had to recalibrate the rotor again. I’m not impressed with the accuracy of these 3-wire rotors. My 40-50 year old 5-wire rotors always point where they say they are pointing.

[tripelo]

Originally Posted by Pete Higgins

...The RCA draws <80 ma and my 3 PCT MA2-M’s were drawing almost ten times as much...

Yes, the drop amplifiers use more current. The regulators are more conservatively designed. And, all things being equal (which they seldom are), high dynamic range amplifiers consume more power.

My Winegard AP-2870 overloaded ...expecting the RCA might overload on the tower as well. No such luck. ... no signs of overload

That is a bit puzzling. The Winegard preamplifiers that I have scrutinized are, in principle, not much different than the RCA design appears to be. Both use BJT, bipolar junction transistors. Might expect a few dB difference in overload characteristics but not many dB.

Subjectively, I think the RCA is outperforming the PCT drop amps. ... I would sure like to know where the 1 dB compression point is reached with this design.

It appears to be similar to the original CM-7778, and by extension it would be similar to the CM-7777. Both CM's are about as good as one could expect for single-ended BJT designs. But, neither would be considered outstanding in terms of high dynamic range.

...I calibrated my NTE/ECG U-106 ... I’m not impressed with the accuracy of these 3-wire rotors. My 40-50 year old 5-wire rotors always point where they say they are pointing.

Yes. The more mass the three wire rotors have to rotate, the more frequently they become uncalibrated. As you know, the principle involves two synchronous motors (one in the control and one on the mast) that rotate at about the same speed, no feedback loop for correction. Problem is that the mast motor runs slower.

Do you think that ~1.5 wavelengths of spatial diversity is mitigating your frequency selective fading? It would certainly help explain improvements in excess of simple combining gain.

That is good thinking.

The antenna spacing of 1.5 wavelengths is sufficient to provide a signal path that is at least partially independent (not completely correlated with the other antenna). Independent, or at least partially independent paths are necessary for diversity gain.

It seems, with this configuration, there are some difficulties with realizing true space diversity, because of the method of combining the two antennas.

Maybe more thoughts on this later.

[Pete Higgins]

tripelo,

I’m not sure that the programmable “digital display” 3-wire rotors employ two synchronous motors (one in the control and one on the mast).

From what I’ve been able to gather, the control unit has a counting or “timing ckt.” that calibrates itself to the time of travel from 360 deg. to 0 deg. My calibration procedure involves running the antenna motor full CW until it is at the end of its travel and the display reads 36 (360 deg.) and then pressing the “INITIAL” button.

When it gets really badly out of sync, I have to reset the control box several times to just get the array to point north. I’ve had the digital display read over 40 before reaching the CW end of the rotors travel. The control unit then returns the rotor CCW to 0 deg. to complete its calibration. If I run it back around CW the display shows 36 and the rotor is back to the end of its CW travel.

I do hear a slight hum whether the rotor is running or not and a relay click when the rotor starts or stops turning. I haven’t tried to take the control box apart to look for a separate motor but that sounds like a good indoor project on these 100 deg. days.

[tripelo]

Originally Posted by Pete Higgins

I’m not sure that the programmable “digital display” 3-wire rotors employ two synchronous motors (one in the control and one on the mast).

Yes, I'm not sure either.

From what I’ve been able to gather, the control unit has a counting or “timing ckt.” that calibrates itself to the time of travel from 360 deg. to 0 deg...

The well-known Channel Master rotator and clones essentially use the synchronous motor in the control as a timer. So, it could be in newer units that a timing function is implemented in a smaller or less costly circuit.

When it gets really badly out of sync,...I’ve had the digital display read over 40 before reaching the CW end of the rotors travel. ...

I do hear a slight hum ...that sounds like a good indoor project on these 100 deg. days.

Yes, for the curious mind there is always a project.

Seems unlikely that a timing motor would be used with a digital indicator, when a timing circuit could be smaller and probably less expensive.

I haven't used, or looked at circuits of, any of the newer digital indicator rotators, so my comments are speculation.

[tripelo]

Following up on a partial response to Pete’s thought provoking question as to whether space diversity gain could be a mechanism for alleviating frequency selective fading, presumably originating as a result of multipath.

Originally Posted by Pete Higgins

That’s a lot of separation between your long Yagi’s. Do you think that ~1.5 wavelengths of spatial diversity is mitigating your frequency selective fading?

The writing below is an opinion, subject to revision.

There are several methods of combining two or more independent signals to achieve diversity gain. The list below represents some techniques, but is not a complete list.

Wikipedia was consulted for some of the diversity combining names.

1. Maximal Ratio or Ratio Squared
2. Equal Gain
3. Switched
4. Selection or Scanning
5. Other

Maximal Ratio or Ratio Squared: The gain applied to each signal is determined by the S/N ratio of each signal. Higher S/N signals are increased in gain relative to low S/N signals, then combined.

Equal Gain: The signals are phase shifted into alignment and added without altering gain of either signal

Switched: The receiver uses only one signal until the signal becomes essentially unusable or drops below a fixed threshold, then switches to the other signal.

Selection: Signal strength of each separate path is evaluated; only the strongest signal is used.

Other: Other methods have been devised, depending on field of application. An example for TV might include post-detection combination of the video from each separate antenna or path. This method requires separate tuners with timing and delay hardware suitable to align two independent video streams; then combining based on some of the principles suggested by the above techniques.

Combining in a summation/additive device such as a splitter (reversed), hybrid combiner, or transmission line combiner, does not fit into any of the above categories. These directly additive combinations make no allowances for either phase or amplitude adjustment before combination. Having no means of modifying the signals before addition permits potential destructive interference between signals to continue. The main reason multipath is a problem in reception is destructive interference.

Destructive interference arises when signals add in such a way as to reduce signal strength. Electromagnetic (RF) signals are characterized by magnitude (amplitude) and phase; they add according to principles of vector addition.

Examples:

1. Two signals of near equal amplitude and nearly 180 degrees phase shifted relative each other, add to nearly zero amplitude.

2. Equal strength signals with less than 180 degrees relative phase shift can interfere with each other (depending on amount of phase shift) , but not complete cancellation.

3. Signals that are 180 degrees out of phase with unequal amplitudes interfere with each other, but do not completely null to zero.

There are many possibilities in between those of the preceding examples where two or more signals add to a sum (amplitude) that is lesser than the amplitude of either signal alone.

The summation of two antennas (in a stack arrangement) essentially increases the antenna aperture. Antenna aperture is the effective area in space over which signals are intercepted. In special cases such aperture increases could alleviate mutlipath effects by intercepting better quality signals. If so, this is a matter of aperture placement rather than diversity gain. If the multipath effects are random and equally distributed in space, then a larger aperture will not produce diversity gain.

An increase in aperture usually means a decrease in antenna beam width. Decreases in beam width may eliminate or reduce some potential multipath signals. (More on this later)

It appears, the additive combination of two antennas does not eliminate (through the mechanisms of diversity) the potential for the destructive signal combinations resulting from multipath signals.

.

[Pete Higgins]

tripelo,

Interesting: The depths of frequency selective fading, seen with the single Yagi, imply that gains in signal strength alone would not be sufficient to overcome the effects of the deep fades. Clearly, a 25 dB selective fade could not be mitigated by a 3 dB increase in signal strength (in reality, stack gain is likely less than 3 dB).

From the little bit of reading I’ve done on the subject, VHF is more prone to rapidly-changing multipath conditions than UHF although the new demodulators with long equalization spans have dramatically reduced multipath effect, both static and dynamic for 8-VSB reception.

None the less, your early sample 2 data showed ~25 dB of what appears to be frequency selective fading evident at the upper portion of the 6 MHz channel 11 spectrum. Assuming your test setup was undisturbed between samples, I think the apparent loss of signal has to be attributed to some mechanism external to your test setup. In my experience frequency selective fading (when neither the transmitter nor the receiver is moving) is generally due to destructive interference from multi-path causing cancellation of certain frequencies at the antenna. For high VHF & UHF, which are usually received via a LOS mechanism, this is typically from reflections off the ground and nearby objects.

For WHAS-11, using 984’ (300m) Tx antenna height and guessing your tower @ ~65’ I get a line-of-sight path of roughly <50 miles. I reasoned that when there isn't a line-of-sight path between transmitter and receiver the signal has to be both diffracted and reflected along multiple paths before finally reaching your array. I expect some signal arrives via tropospheric scattering due to changes in temperature, humidity and barometric pressure causing slight changes in the refractive index although I’m not sure that would in and of itself explain short term selective fading.

In any case, each of these arrival mechanisms introduces signal anomalies that can destructively interfere with one another at the aperture of a receiving antenna. With a single antenna or two closely spaced antennas, I would expect little difference in coherence of the arriving signal(s) yielding a theoretical overall 3dB stronger signal, but, exhibiting the same overall characteristics (frequency response). By coherently combining two separate apertures spaced ~1.5 wavelengths apart, however, I would expect slightly lower overall gain reflecting a summation of all the different paths coherently combining at the antennas. The phase differences causing frequency selective fading at one antenna would in effect be mitigated by the arrival phase at the other antenna.

From WikipediA:
Antenna diversity, also known as space diversity, is any one of several wireless diversity schemes that uses two or more antennas to improve the quality and reliability of a wireless link. Often, especially in urban and indoor environments, there is no clear line-of-sight (LOS) between transmitter and receiver. Instead the signal is reflected along multiple paths before finally being received. Each of these bounces can introduce phase shifts, time delays, attenuations, and distortions that can destructively interfere with one another at the aperture of the receiving antenna.

Spatial diversity employs multiple antennas, usually with the same characteristics, that are physically separated from one another. Depending upon the expected incidence of the incoming signal, sometimes a space on the order of a wavelength is sufficient. Other times much larger distances are needed…

I guess my thinking is that if your frequency selective fading data is the result of multipath signals arriving out of phase at a single antennas aperture, combining two antennas narrows the aperture (beam width) eliminating some portion of the fade contributor. Additionally, with 1.5 wavelengths spacing, you have signal arrival from different paths at each antenna. As one antenna receives frequency selective out of phase signals (a fade) there is a good probability that the other antenna will receive in-phase signals, mitigating the depth of the fade. If that mitigation is sufficient to keep you above the detection threshold the equalizer can do its job, further minimizing the fade effect on reception.

It will be interesting to see what you postulate in your next installment.

[tripelo]

Thanks Pete, for your thoughts and comments.

Probably soon, more can be written about space diversity.

Would like to explore antenna elevation pattern discrimination as it applies to long distance multipath.