The Official AVS Antenna and Related Hardware Topic! - Page 95

You have a Yagi right now. (Technically, it's a yagi-corner reflector hybrid, but most commercial UHF antennas are.) But the bulk of your antenna is a log-periodic setup for VHF, with the UHF section just pasted on the front.

What makes an antenna more directional is the number of elements it has (or the arrangement thereof, but that's moot for the current question.) Ignore the VHF elements for a minute - they aren't doing anything for your UHF reception. There's probably between 20-30 UHF elements on your CM3678. There's 91 on the 91XG. It is going to be far more directional when the yagi half of the hybrid is being used. (On lower channels, the corner reflector becomes the primary means of gathering the signal, and these are not, by rule, very directional.) When the corner reflector portion comes into play, the 91XG has a better corner reflector than the CM3678, but most of that is going to be gain, not directionality.

It is a fact that a UHF-only antenna will outperform a similar UHF/VHF combo antenna with similar-sized UHF sections. The UHF/VHF combo is a compromise, one that works well for many, but still a compromise.

Is there a DIY for antennas to build your own UHF antenna?

I have seen web pages for HAM stuff but I haven't located a resource for a kit or make your own.

Wow, I'm getting overloaded(pun intended) with information here.

The info I have for the 91XG was from the AntennasDirect site which apparently is lacking in useful data. I recall that it was a clone of a Blake model sold in the UK? Not sure which one though.

holl_ands, with your data from above on pre-amp overloads, you first indicate that the Winegard model would be better or did you refine that to say the CM 7775 was the same as the Winegard?

AntAltMike, you mentioned that the 8-bay CM4228 would be a better choice than a Yagi style? I was under the (probably misinformed) impression that Yagis were more directional and that would be better for my situation. I have a 4228 still in the damaged box it came in, would require some straightening but I could probably install in place of the 91XG this weekend. Before I go up there though I'd like to know if a pre-amp swap-out is in the works. I also still have the CM4248 which can be used.

Too bad I can get you all here at my location, my treat! It sounds like my situation could be a challenging one that appears to intrigue or interest you. This stuff definitely interests me but I don't have the knowledge to understand a lot of it.

AntAltMike, you mentioned the RS ATSC tuner, I had one but returned it as it was problematic fitting into the scheme of my whole system. But, I am going to pick it back up as I recall it had a SNR readout meter...which I'm assuming would be of some usefulness..?

All I meant was that I was visualizing the 4228's channel 22/23 sidelobe to be just a little more favorable in terms of rejecting the undesired signal from that angle, because the undesired signal would arrive at each of the 4 bay sub-stacks at a different time, but that is just one selection consideration. I can't say that the 4228 antenna is better for your situation than a 91XG, because yours has more gain, and if you "misaim" it by a few degrees, you probably can achieve the same rejection of the undesired signal.

Right now, you may understand less about your siituation than you did before you entered this thread. Originally, when you had concluded that there must be obstructions between you and Sutro because you got better performance misaiming your antenna, the real reason you got better performance was probably that you were mitigating channel 68/47 signal at the same time. Holl_ands' preference for your situation would be to go with a lower gain preamp. I'd like to get that spreadsheet of his to open before I concurred, because I still suspect that it would develop marginal signal levels of the distant transmitters, and even if a topo program says that solid planet earth is not in your way, its vegetation still might be, and you might desire some additional margin to sustain reception under wet weather conditions. My preference is to have you drop in a 750Mz low pass filter and a UT-2700 notch to weaken 47, but either would likely improve your system performance.

See my Post #93 above for the corrected RevD spread sheet and listed errata.
I also added calculations for the W-G AP-4700 Preamp.

Calculations show that both the AP-4700 and the HDP-269 should keep the strong signals coming
from KRCB (both analog and digital) well below the spec overload point when pointed toward Sutro Tower.

The high gain CM-7777 and CM-7775 (UHF only) is shown to be unsuitable for use in a
urban/suburban environment with broadcast towers in close proximity
(but then most of us knew that to begin with).

A couple of additional notes and questions for you all.

Solid planet earth is definitely in my path. There are a couple of 1000' plus peaks on my line of sight to Sutro.

Using the RS Accurian tuner I get the following readings from their "meter", this is with the antenna orientated for maximum signal strength for KBHK.

KBHK.......SNR-18-19dbs with signal strength at around 66-67

KCRB.......SNR-30-32dbs with signal strength at around the low 90's

KTLN........SNR-15-16dbs with signal strength at around 60

Additional Sutro Tower station,

KRON(ch 4-digital 57) constant picture breakups unwatchable ...SNR about 15-16 and signal strength about 40-60

Mt St Helena,

KFTY(ch 50-digital 54).....SNR-2-15db with signal strength of 9...this station would be behind me so I guess it indicates the antenna is doing a fair job of rejecting signals from the backside.

Whether this info is of any value, I don't know, you guys are the experts. Obviously, the best course of action is to try and see what works better, if anything. I guess I'm looking for answers on whether things like horizontal stacking and such are fundamentally sound. I'm also unsure from the discussion whether a different pre-amp is warranted, again, testing it is probably the best solution. Again, I do appreciate the input and I have learned a whole lot more than I knew before starting this journey, and for that I want to thank everyone.

Now just get me better signal strength and I go away...

http://www.wfu.edu/~matthews/misc/dipole.html

http://www.clarc.org/Articles/uhf.htm

I bought one of those RS tuners because the clearence price of $90 was too low to resist. I am not sure what the SNR number represents but it is not based on the AGC value or channel power for an analog station. The channel power does seem to affect the signal strength level. When I tune to an analog station number (after clearing all channels by doing an auto re-scan with the antenna unconnected) The SNR number is usually an erratic 17 dB regardless of how strong the station is but the bar graph and signal strength readout is a consistant 49% for strong stations. Both those values seem to have more meaning when tuned to a digital station but I don't think you can read RF signal strength using them.
My recomendation for reducing channel overload would be to try a WG-4700 or 8700 pre-amp plus a filter for channel 68 if necessary. If KRCB is a problem even with the WG pre-amp the best solution to eliminate it would be to get another XG-91 and horizontally stack them to place a notch at 30 degrees off the mainlobe peak.

What would be your eventual goal in a DIY project? I don't think "kits" exist.

Which channels do you want (a few or entire band) and which need the most gain?
Do you want to build a really big super high gain antenna (but smaller than Aricebo)?
Or are you simply looking for a moderate gain antenna on the cheap?

Would it help if you only knew the detailed measurements for commercial antennas,
such as are used for input to NEC simulation runs?

Would you want to use an NEC simulation program to try out your design and tweak various parameters?

The HAM stuff you mention has lots of guidance on building very narrowband high gain Yagi type antennas (one or two channels)
as well as very wideband, moderate gain LPDAs (Log Periodic Dipole Array).

What's your preference?

Well actually both. I have one channel (35) I would like to build for and then see if I can improve upon the commercial grade antennas. i have built a lot of stuff both electronic and not so the above references are a good start.

When you say narrow band are you talking only a single channel with good rejection from the adjacent channels?

I saw that channel master made a parabolic dish at one time but you can't buy that unit anymore and besides it is the higher UHF band I would like to receive better. Ideas?

That's the direction I'll probably take initially, a horizontal stacking and see what happens. I'm having a little trouble wrapping my head around the idea of using a lower gain pre-amp, but that is probably my lack a understanding more than anything so I'll probably get one of the WG pre-amps and try it as well.

You mentioned center to center boom width should be around 22-24"?

What's the best way to join the leads? The antennas have a balun contained in a box attached to the boom so it's a 75ohm coax feed from the antenna.

Another DIY webpage:

http://www.mtmscientific.com/yagi.HTM

Ron

Use 22.5" unless there is a problem with the antenna elements getting too close.
holl_ands had this post about the best way to combine antennas:
http://www.avsforum.com/avs-vb/showt...da#post5768104
Cheap way- same null but insertion loss may cancel much of the gain of using two antennas: CM model 0538 available from:
http://www.warrenelectronics.com/Antennas/joiners.htm
The gain of the pre-amp needs to overcome the losses in the lead-in cable and reduce the effect of the noise factor of the receiver input. The usual rule is 10dB more gain then lead-in loss is sufficient; after that additional gain won't help much and overload from too much gain can make things much worse.

Okay so the Lindsay 2WCU1469 would be the one to use. With a cable run of 100-125' would RG11 be a better choice than RG6?

Would it be a better idea to stack a CM4228 on top of the 91XG and join them?

You're picking the most expensive choice first. Why would you want to try a horizontal stack first? It will cost you several hundred dollars and will not mitigate any overload coming from 47/68 if that turns out to be the problem. You can rule that problem out much more readily and economically either with the filters that I recommend or the preamp that Holl_ands recommends.

If your stack gets you more net gain, stacking will increase the undesired signals from 47 and 68 as well. You can rule out channel 22/23 by putting a $30 Jointenna temporarily on the line. It will crush ABC 24, but it will at least tell you if mitigating 23/24 helps you. If it does, you might then try forming a horizontal null which could slightly weaken 22/23 without weakening 24, but until you establish that is contributing to your reliability problem on 45 and otherwise, I think you would be wasting a lot of time and money by stacking first.

While horizontal stacking can mitigate an off-axis 22/23 Mz undesired signal component, which I doubt it your problem, there is no optimal horizontal spacing to get the two antennas to resonate favorably at all frequencies. If you want to do something really cool, rig up something motorized that will let you vary the horizontal spacing to optimize on each distant channel. That's what I'd do if I didn't have to work for a living.

I used to have this math teacher named Mary Tighe. One day, she saw that I was doing nothing (that could have been any of a number of days) and said, "What are you doing?" I replied, "I'm thinking" She said. "You're not supposed to think. I do the thinking here!"

She was a useful math teacher for most of the other students but not for me.

Keenan is doing to much thinking. He keeps thinking of the most expensive and most difficult to implement solution but he doesn't yet know the problem. He needs to know for sure if he has a preamplifier problem caused by overload, which I doubt. He can definitively resolve that for under a hundred dollars with filters. If he chooses to go with the low gain amplifier, he might as well use RG-11 because, without a real signal meter, he doesn't know if his distant signals will be near the threshold or not. RG-11 will cost him maybe twenty bucks.

LOL, you're right, because I don't understand most of the math and theory regarding antenna usage my first inclination is to throw money at it to fix a problem that I may not even have..

I'll order a JoinTenna model 0585-1 and go from there. I think this was already answered, but can I insert this in the line post pre-amp so I don't have to go up the tree, at least to see if it has an effect?

No. Post- preamp notch attenuation would only "help" your receiver correctly process channel 22.1 if the receiver was previously not able to process it. You're going to have to tough this one out, but again, you are overly relying on your signal strength meter's numerical quality value of channel 22.1. According to the FCC website's transmitter power figures, it is not your strongest station. 47 and 68 are.

Actually, just moving the preamp down will reduce all of the inputs by about three dB. If you are primarily afflicted by an overload problem, you might get somewhat improved performance from doing that before any filters even arrive.

You can accomplish something by temporarily moving the preamp down, but don't move it down any further than you have to. If you make a splice insertion point even twenty feet above the ground and put the preamp there, you can then mix and match with the THREE jointennas (23, 47 and 68) to see which channel, if any is overloading your preamp.

If the problem is 68, leave in the Jointenna. If it is 47, then buy a tuned UT-2700 or even an MWT-U to weaken that without disturbing your 45. If it is 22/23, then you might FIRST try a UT-2700 tuned to the analog visual carrier of 22, because that can be notched without damaging DT-24, but if the problem is DT-23 (which you see as 22.1), there is no notching possible, so then and only then do you try to pull off a minor miracle with the horizontal phase cancellation.

I've seen receivers display 90% to 100% on lots of channels. I never even bother to measure those signal's actual signal strength with a field strength meter because they almost never are causing a problem. With some products, you can have a signal strength of 100%, increase it by a hundred fold (which would make it 10,000%, if the 100% were a measure of actual signal strength referenced to some upper limit), and still not have a problem. And I have bad news for you if you think that increasing the nominal signal quality from, say, 40% to 50% or from 50% to 60% is going to make much difference in the reliability of your problem signals if you do not have line of sight reception from the San Francisco towers...

Of course you want it all....don't we all.
Narrow band would not have good rejection for any of the near adjacent channels.
For that you would need an input filter.

Various versions of NEC Sim can determine how well a design will work (gain, VSWR, etc.)
You might find the hdtvprimer NEC files of interest. Start with readme and xls files.

Suggest you look in the fol. section on "Practical Antennas: VHF/UHF"
http://www.cebik.com/
Especially note that he presents lots of NEC simulation run results and tips.

There are several narrow band Yagi and wideband LPDA designs for the
UHF band (up to 40 elements!!!).

=======================================================
Some comments re the so-called Yagi. The original Yagi-Uda design uses
a dipole as the only fed element (and usually a folded dipole for wider
bandwidth and better impedance) .
A slightly longer simple rod is used as the reflector element and any number
of decreasing length directors may be used.
The resultant designs are fairly simple to calculate using a spread sheet.

An array of rods (W-G PR4400 and PR8800 use a Triad Reflector) or a complete
Reflector Screen may be used to supplement the reflector element
(Corner Reflector) to increase the gain and F/B ratio.
Some antennas, such as the XG-91 type, use wider bandwidth bow-tie elements.

Beyond that, modern Yagi design becomes more black magic.....
A classic Yagi-Uda would have very narrow bandwidth (only a few UHF channels),
as you'll see in the various NEC Sim runs.
In order to expand the bandwidth to cover nearly the entire UHF band, the element spacing
and lengths are moved around to "de-tune" the antenna.
With a computer, thousands of possible detunings can be analyzed.
Some have said that a monkey randomly picking alignments can do about
as well as a trained antenna engineer.

And the DAT-75 type is an entirely different animal, using three stacked Yagis
pointed towards the reflector screen.

========================================================
You might find the fol. LPDA spread sheet calculator of interest.
I punched in the max sized LPDA that the spread sheet would permit:
result was a fairly big 15 element array, with 10 dBi = 7.85 dBd gain.

An LPDA is fairly difficult to fabricate, esp one built to withstand the elements.
But once you understand the spacing rules, it shouldn't be too difficult to model and
build either a Log Periodic Toothed Array (such as the overly small Silver Sensor)
or even a simpler Zig-Zag Log Periodic Array, like the ones I described in the fol:
http://hdtv.forsandiego.com/messages/1/2944.html
I have an old roll of RG-58 out in the garage that would be ideal to simply string up
on a frame in the garage whenever they fire up DTV in the VHF band.

I also found the Corner Reflectors fed by a simple Dipole element to be interesting.
But these are SUPER-SIZED reflectors with a lot more gain.
I just wonder what happens when a super sized reflector is used with a big Yagi...

Unfortunately both the Yagi and the LPDA are fairly large, require significant rigidity
to survive outdoors and need to be truly monsterous to out perform the CM-4228.

=======================================================
The CM2251 parabolic may not be available, but Wade/Taco and Lindsay in Canada still are:
http://www.wade-antenna.com/upa.htm
http://www.lindsayelec.com/antenna/c...6-pbu-uhf.html

Note that the Lindsay has a short LPDA pointed towards the reflector.
Hard to see, looks like maybe a 4-element LPDA (too keep it close to the focus point).
I can't make out what the Wade/Taco uses as a feed antenna.
If you blow up the below jpg, you'll see the CM4251 7-foot Parabolic is fed by a
2-Bay Bow-Tie with Reflector Screen. A Balun module attaches to the screen.

You could do a DIY project, either constructing your own reflector or an old "BUD"
(Big Ugly Dish for C-Band SAT).
A 3-D parabolic would be best, but a 2-D parabolic is also possible by simply folding over some wire mesh.
Or use a Super-Sized 2-D or 3-D Corner Reflector.

Once you have a reflector, you'll need some sort of small UHF antenna to feed it.
You might want to try a 2-Bay Reflector (like the DB-2) or a 2 or 3-element YAGI or a Silver Shadow LPDA.
Or one of the short broadband zig-zag structures like the radio telescope feeds shown in the zig-zag link above.

Let us know what looks good in NEC and of course, what works.

 

UHF LOGPERIOD.zip 10.298828125k . file

While the horizontal stack with a low insertion loss combiner is an expensive solution it will improve on-axis performance across the UHF band. You would only need to change the spacing if you needed to move the null. I have modelled a stack of two CM4248s at 22.7 inches. I have enclosed a comparison horizontal beampattern with a single 4248 plotted at 526 megahertz. I agree that Keenan should first try something simple to see if overloading is his problem. I would suggest he just try switching to a WG-4700 or 8700 pre-amp which is much more overload resistant than a CM 7777. I doubt he would overload his receiver with that pre-amp given a long RG-6 lead-in from a tall tree. That seems to me even simpler than messing with filters.

I finally had some time to do a couple quick RADIO MOBILE propagation prediction runs for Keenan location.
My goal is to encourage others to use this FREE and relatively easy to use program, along
with some handy-dandy spread sheets to help understand what is possible with OTA DTV reception.
I'm still collecting data on how much Fade Margin various situations "need".

=====================================================
It confirms that KRCB-DT and A-KRCB are operating uncomfortably close to the spec
overload point for the CM-7777, but depending on how much +/- 28 degree sidelobe
notch depth is truly available against CH22/CH23.
Unfortunately, most antennas don't develop a strong sidelobe notches at the lower frequencies.
[Fortunately, the W-G UT2700 tunable notch can be used, one notch for each channel.]
The high gain of the CM-7777 also may be causing intermods in the HDTV Tuner.
Assuming a 10 dB sidelobe notch, the input to the HDTV would be close to -6 dBm (43 dBmV),
which would generate low level intermod noise products preventing reception of weak stations.

The higher gain W-G AP-4700 (or equal) Preamp is a slightly better choice than the HDP-269,
providing a slightly higher Fade Margin for the desired KBHK-DT from Sutro Towers,
and has a spec overload point that is much higher than the maximum input signal
so that the SFDR (Spurious Free Dynamic Range) can be maximized.

There is a delicate balance between using a higher gain Preamp to overcome the
coax downlead and HDTV Noise Figure Losses vs providing ample "head room" below the spec overload point.
The HDTV Tuner is also likely to be overloaded, so a Variable RF Attenuator should
be inserted prior to the HDTV and tweaked for "best" reception.

==========================================================
For KBHK-DT, RADIO MOBILE shows two hills in the way, the first where the KTLN tower is located
and second on the side of the hill S. of Santa Rosa.
The two knife edge diffraction path causes 52 dB loss over and above the Line-Of-Sight path (that's a lot!!!)
The received signal level was calculated to be about -77.1 dBm (or -28.4 dBmV) for the XG-91 feeding a Preamp.
RADIO MOBILE calculated a Fade Margin of 15.5 dB, but there are a bunch of parameters
not included in the results, resulting in a very minimal 8 dB after making adjustments.

I use one and only one parameter calculated by RADIO MOBILE: Total "Path Loss".
It is calculated using the unit locations, the terrain database, the transmitted frequency
and not much else. The "Path Loss" is independent of any assumption re TX ERP, TX/RX
Antenna Gain and various other parameters that can be entered into the program.
Path Loss is then plugged into the DTV FADE MARGIN CALC spread sheet, which incorporates
all of the site specific parameters (as best as I understand them, so tweak away).

KTLN (CH47 and 68) signals originate on the back side of a directional tower antenna
and then diffract off the side of the "hill" South of Santa Rosa, resulting in a Diffraction Loss
of about 40 dB in excess of Line-Of-Sight.
The receive signal levels were calculated to be within a few dB of KBHK-DT,
which isn't strong enough to be interfering with KBHK-DT reception.
Hence the CH47 Join-Tenna isn't attacking the real problem (CH22/CH23).

More antenna gain could help, such as Horizontal Stacking, but only yields 2-3 dB gain.
Of more importance would be the "space diversity" effect of having different parts of
the antenna
separated so that a multipath null in one part of the array doesn't affect a different part of the array.
This is the primary reason the CM-4228 has been shown to be such a high performer.

A Horizontal Array also can develop a very deep notch that can be directed towards CH22/CH23,
but if the Preamp is chosen to operate in the linear region to begin with,
this may not be needed.

See attachment for path profile and calculation files.

Please let me know if you find any problems, there are a lot of meticulous details.
No, I'm not volunteering to do entire SF area. I've already done LA and San Diego.

 

Kbhk_map.zip 498.216796875k . file

 

Santa Rosa DTV Fade Margin Calc RevA.zip 23.30859375k . file

Unfortunately, I can't get Holl_ands's two most recent links to open either, and when I try the "right click" workaround that works for him, it gets me a different, insurmountable, "no can do" message.

Has a link been furnished to this free, Radio Mobile program? And why is it free? 100,000 Watts is no longer free, and the free, NIA calculator is gone, too.

Since Keenan's nearby hill takes the possibility of overload by 47/68 out of the picture, he is basically left with two problems: too much local 22/23 signal, and insufficient or qualitatively inadequate, distant 45, 24, etc.

I'd be interested in seeing some of the assumptions Holl_ands is using. Is 23 up to its licensed power level or still running at its temporary authority? For the thermal noise floor, I have always assumed that since the figure I used for analog estimates was between -59 and -60dBmV, and since I thought that was based on the entire 6Mz bandwidth, and since the DTV tuner industry would like us to use dBm instead, I have been assuming that the noise floor is about -108dBm. Is that about what the industry is using for a reference level?

It has been posted elsewhere that there is some kind of mandated or engineered need for a 15.6 or 15.9 dB S/N ratio to be seen at the "tuner". Two thoughts: first, must one add the tuner's noise figure to that, meaning that the minimum input level would be about -82 to -83dBm if the tuner had about a ten dB noise figure?

Second: are the tuners engineered to process signals at anemic input levels? Consider DBS I.F. signal processing parameters. The DBS I.F. frequencies are just above UHF (950Mz to 1,450Mz) frequencies. Since the bandwidth is about 30Mz, then the measured noise floor would be about five times greater than a 6Mz noise floor, or -103 dBm.

Supposedly, the DBS, QPSK signals can be processed if their S/N ratio is as low as 8dB. But the DBS tuner manufacturers say that their tuners need a minimum input of -60dBm, and when Sonora Design bench tested a bunch of early Sony receivers, it saw that performance started to deteriorate when the input signals dropped below -66dBm. Yet, assuming a -103dBm noise floor and tuner noise of 10dB, the tuner noise should have been inconsequential. But when you take a DBS signal that has dropped to, say, -75dBm to -80dBm and boost it with a "dime store" inline 20dB amp, the signal gets recovered to a processable level. That has led me to conclude that determining tuner input sufficiency requires more than maintaining the minimal S/N ratio.

I've only seen one published figure for minimum input level by one manufacturer, and it was either -80dMb or -85dBm (I'm getting senile, too!). If anyone here has seen any published threshold input specs, I'd like to see them. In fact, published, manufacturer's info for DTV tuners is so sparse that it might be worth creating a sticky post where we could store it, if there is even enough info available to warrant such a post (I think Drake and PDI are the only two manufacturers in whose literature I have ever seen processor noise figures: 9dB and 10dB).

Preserving "headroom" is going to become a much more important consideration in establishing the integrity of off-air amplification systems than it ever was with with cable or even broadcast analog for a number of reasons.

With cable TV, all signals are at the same level, so any "wideband" noise affects all channels similarly, whereas in an off-air system, noise and intermodulation products that would have no effect on equal strength signals would decimate the weaker ones. Thus, the notion of arbitrary levels at which amplifiers "overload" will be much less useful than it was previously.

The byproducts themselves will be wider, since before, all we cared about was the simple interactions between two or three visual carriers, and whether they fell on or very near to another visual carrier, but now, we have these 6Mz wide signals interacting, and they may adversely affect the weak, desired DTV signal regardless of where, within the channel band, they land on it.

Given that the broadcast bandwidth has been reduced from 68 channels to 50, and in most markets, it has been effectively reduced to 45 since very few VHF lowband stations will be reverting to lowband, we are going to be dealing with a lot more adjacent channel problems.

I surmise that Keenan may have ABC DTV available through other means, like cable TV or satellite, so that 24-D from San Francisco is not essential, but other people will not be so lucky. Most of my customers are in Washington, DC and are within a few miles of ABC-D on 39 and WB-D on 51. Baltimore is presently using 38 for CBS-D, and 40 for WB-D and 52 for ABC-D. Most of my customers are within five miles of the Washington towers but those towers are about thirty to thirty five miles from the Baltimore towers. Because my customers basically encircle the Washington transmitters, I can have two headends just a few miles apart where, from one building, adjacent channel digital transmitters are directly inline, with one being five miles away and the other thirty-five miles away, but, five miles to the east, from another headend, the transmitters form a ninety degree angle to one another.

FWIW, strong DTV channel 39 adversely affects weaker, lower adjacent DTV channel 38 more than it does weaker, upper, adjacent channel 40. I don't know if that is because the DTV signal has a pilot carrier that is just .9Mz above the lower band edge that makes the differential between the it and the adjacent lower channel greater or if I have just been unlucky for some other reason.

Anyway, I can usually get the weak, adjacent upper channel to be processed if the differential between it and the stronger, adjacent channel below it is less than about fifteen dB, but don't take those numbers to the bank, because they are based on an inadequate sample and imprecise observations and recollections.

Assuming 10dB off axis rejection, what does Holl_ands estimate the differential between local 23-D and distant 24-D to be? The 22 analog carrier can be squashed with a tunable notch without affecting 24, but nothing that convenient can be done to 23-D without injuring 24-D.

If someone in Keenan's locale needed 24-D, like I might if I was servicing a multiple dwelling unit there, they could point a second antenna directly at 24-D, bandpass filter it, and couple it with their main antenna coax to effectively phase cancel the undesired signals. Funny thing though. If one aimed a horizontal stack at San Francisco to take advantage of the "space diversity" effect that Holl_ands alluded to in the previous post (an underappreciated characteristic of multiple bay antennas that explains why they consistently outperform Yagis in attics), then I don't think the phase canceling would work well because the undesired signal components from the two stacks could not both be kept in phase with one another. It could be developed, but not at the top of a 70 foot mast or tower. On a flat roof, I could put one antenna on a 3' x 3' non-penetrating frame with no ballast, and slide it "side to side" until I saw the most favorable spectrum form on my analyzer, as I actually had to do once for a 550 unit condo, but again, such extreme remedies are not for the do-it-yourselfer.

The "space diversity" effect that Holl_ands alluded to in the previous post is an underappreciated characteristic of multiple bay antennas that explains why they consistently outperform Yagis in attics.

When I googled radiomobile this is the first link I got:
http://www.cplus.org/rmw/english1.html
It is free for the same reason that Linux and much other quality software is free- the author chose to release it that way. It is also based on path loss work and elevation data that have already been paid for through US tax dollars. I have used radiomobile and it is a great tool for calculating path loss. The only remaining path loss question is the effect of items very close to the receive antenna like trees and buildings etc. The program can also calculate signal strength margins but they were designed for mobile radio applications and it can be easier to just use a spreadsheet to finish the calculations. Anyone that is interested will find a good deal of relevant info at the sites that holl_ands has posted.

Here's the link for Radio Mobile,

http://www.cplus.org/rmw/english1.html
Radio Mobile WEB Site

I see that deconvolver has already posted the link. It involves DL'ing about 5-6 files but it works, not that I know how to use it...

If you like, PM me with your email address and I can send you the unzipped files that you can't download.



holl_ands, there is some wonderful work here, thank you. There's references to many things I don't understand but your note sections gives me the gist of what I believe I need to know. I have always felt that the hill that is in close proximity to me has been the biggest impediment to quality SF reception. Just west of me about 4-6 miles there's a forum member here in Santa Rosa that has been getting most of the stations from SF and even the KNTV signal from Loma Prieta, clear down in the south bay past San Jose, with a CM4228 in his attic!! Of course, he doesn't have the hill to deal with.

As I mentioned before, holl_ands has posted a good deal of info over at the hdtv.forsandiego.com site. You will find most of the answers to what he is doing there. Here is a link to a spreadsheet he created for calculating fade margin (I didn't look to se if he has posted a more recent version):
http://hdtv.forsandiego.com/messages...ample-3108.xls

I posted the Santa Rosa files on the fol thread, which includes info on RADIO MOBILE:
http://hdtv.forsandiego.com/messages...tml?1126976722

=======================================================
PS: Thanks AntAltMike for pointing out the footnote in the Winegard Catalog.
[I KNEW I had seen something somewhere on W-G overload....]

It's also in the 2002 edition. But not on the W-G or retailer websites.
Maybe because the wording is so quirky:

"Outputs stated for 7 VHF and 5 UHF channels at -46 dB cross modulation"
Where what is "stated" in their table is the "Maximum Total Input# (Microvolts)".

Since the table shows input overloads vs output, perhaps the fol is more accurate:
"Maximum input levels are for 7 VHF and 5 UHF equal strength channels,
resulting in cross modulation products 46 dB below the Output signal levels."

======================================================
So the high overload (AP-4700 or equal) Preamps have a 93,000 uV maximum UHF input.
That is 20 log (93,000) = 99.4 dBuV or 39.4 dBmV or -9.4 dBm (for 75 ohm load).
With a gain of 19 dB, the max output level would be +9.6 dBm (58.4 dBmV).
The output intermod noise floor would be -36.4 dBm (12.4 dBmV ).
To provide the ATSC minimum SNR of 15.5 dB, the weakest output signal would need to
be at least -20.9 dBm (27.9 dBmV), corresponding to an input of -39.9 dBm (8.9 dBmV).

The thermal noise floor is about -106 dBm, hence the minimum detectable ATSC signal
is (-106 +15.5 + Cascade Noise Figure). With 100 feet of RG-6, that's about -85.8 dBm
if a pesky intermod product isn't sitting on the desired channel.
So the intermod products can limit the sensitivity to only -39.9 dBm.
If we had a Preamp with a 10 dB higher overload point, the intermod products could be
expected to be reduced by perhaps as much as 30 dB, for a sensitivity of -70 dBm.

========================================================
References explaining this 3:1 intermod reduction were cited in my Post #60 on 9/9/05.
You want to operate well below the spec overload point in order to maximize
the SFDR (Spurious Free Dynamic Range).
If the maximum signal levels can be reduced by 3 dB, the intermod products are reduced
by 9 dB.
Similiarly, if a Preamp with a 10 dB higher overload point is used, the intermod products
can be expected to be reduced perhaps 30 dB for the same input signal levels.

=====================================================
The CM-7777 has a specified maximum OUTPUT level of 51 dBmV or +2.3 dBm
for 2 input channels, with cross-modulation levels 46 dB lower.
Derating is 5 dB for 4 and 8 dB for 8 (equal strength) channels.
So compare the CM-7777's 4 channel derated max input level of -23.8 dBm
to the W-G AP-4700 5 channel max input level of -9.4 dBm, where both state
a 46 dB (0.5%) cross-modulation output level.

Of course, all of this presumes that the manufacturer specs mean what they say.

==================================================
To convert between microvolts and various dB units:
dBuV = 20 log (uV)
dBmV = dBuV - 60 = 20 log (uV) - 60
dBm = dBuV - 48.75 - 60
dBm = dBmV - 48.75

Yes, I posted an updated REVC version.
The Broadcast Antenna Polar Field Plot needed to be converted from relative volts/m to a dB level.
The correction uses 20 log () vice 10 log (), cuz it's volts not power.

========================================================
AntAltMike: In the cited hdtv.forsandiego.com thread above, you'll also find spread sheets
summarizing important specs/measurements for Antennas, Baluns, Preamps, Cable and ATSC Tuners.

And I woud appreciate any additions....

Thanks so much for all your input, sregener! What I need is an antenna that will have much higher gain for the Twin Cities towers, and I'd like to pick up the Eau Claire channels on the back side. (I'd like to stay away from a rotor) I need the following channels: 15-26-32-35-39-50(and someday 49). Nothing is above 50. Given all my desires, which type would you try first?
I'm actually wondering if I NEED the biggest of the big because the dB specs quoted for the 43XG and the DB4 are both significantly greater than the unit I currently have, and I usually get servicable reception for all but 26 (FOX9) and 50 (ABC5).

Some thoughts for the future (improvements are "just over the horizon"...so to speak...)

Performance of so-called 5th Generation ATSC Receivers has been discussed
in fol thread. Look for post dates cited below:
http://www.avsforum.com/avs-vb/showt...24#post6085224

8/16, 8/17/05 et. al. posts have links to lab and on-air tests conducted by CRC (Canadian Research Center)
on the Zenith/LG 5th Gen Prototype Receiver and the Linx (now Micronas) ATSC Receiver.
It actually measures the sensitivity and overload levels, the minimum required SNR and the
additional SNR required for the worst multipath test cases thus far encountered.

7/5/05 post has discussion and links re. European DVB-T Diversity Receivers,
showing how they nearly eliminate signal dropouts.
They have been available for NTSC analog TV reception in automobiles for years.
My '98 Volvo Station Wagon came with a killer Dual Diversity Antenna for FM that
runs circles around my friends earlier Nissan 300Z Diversity FM system.
Maximum Ratio Diversity Combiners are standard for many mobile microphone packs.
And it's in most any Wireless Router that has two or more antenna stubs on it.
SO WHERE ARE THE ATSC DIVERSITY RECEIVERS??????
I've seen various reports re development and field test, but still nothing you can buy.

6/30, 7/15 and 8/3/05 has discussion and my evaluation of the one and only Smart Antenna
system that can automatically maximize desired signal direction and (if they do it right)
has the potential to steer nulls towards undesired directions.
It also has automatic control of the Preamp's AGC level to avoid overload.
A worthy DIY project would be to adapt the Preamp/Combiner Module for 2 or 4 large
external antennas, cuz the DTA-5000 only provides 0-3 dB antenna gain.
It should be capable of outperforming the Lindsay antenna coupler for about the same price, plus of course a CEA/EIA-909 capable STB (e.g Sylvania 6900DTE).

========================================================
Fol. thread has a discussion re advantages of COFDM vs ATSC waveform:
http://hdtv.forsandiego.com/messages...html#POST19348
COFDM is used elsewhere in the world for DTT, in the U.S. for Cellphones and
is expected to be used for DVB-H mobile video in the U.S. and will be used to
provide a variety of services on CH52 and above when they are cleared for Home Security
and other uses. Such as Qualcomm's nation-wide CH55 MediaFLO system.

Big advantages are 1) support for SFN (Single Frequency Networks) wherein multiple transmitters carry the same signal
and 2) relatively simple to build Diversity Receivers which provide very significant protection against signal dropouts due to multipath fading.

The ATSC equivalent of SFN may cause multipath problems for all but the newest ATSC Receivers, depending of course on topography.
But if they finally get it to work, networks of repeaters could be used to fill in the terrain gaps that severely limit ATSC reception today.

I have seen several reports re Diversity Receivers for ATSC, but so far nothing is available for purchase.
With the market for stand-alone STB's being surplanted by integrated HDTVs (and expanding HD on SAT/CATV),
and the cost expected to plumet to $50-75, I have to wonder why the STB and chip manufacturers
don't provide a hi-end OTA STB that has Diversity Receivers and Smart Antenna control.

7/19/05 post in the above thread provided links to laboratory and on-air comparison tests for ATSC vs DVB-T vs ISDB-T.
Performance parameters were measured for several different (unindentified) ATSC Receivers, although they are a bit outdated now.

Bon Appetite!!!

re. BACKSCATTER signal path......and ATSC Diversity Receiver references.

A few dozen posts ago, I mentioned BACKSCATTER as being one way that a signal can bounce off nearby objects,
back into the beam of your antenna, even though the source of the transmissions might be on the back side of the antenna gain pattern.
For the same reason a mirror reflects the most light back towards the light source,
the worst Backscatter sources originate from the back side of the antenna pattern,
and will be scattered back towards the highest gain portion of the antenna pattern.
Depending on geometry, sidescatter can also be a problem, esp for those directions where the antenna has significant sidelobe response.

A measurement of the backscatter and sidescatter levels can be found by looking at the relative strengths
of the received multipath signals for a given DTV channel.
[Unfortunately, STBs only display this information to the Design Engineers.]

======================================================
Refs. [a] and [b] below give some insight on this process.
Our buddies at the Canadian Research Center, the Canadian government's technical experts on ATSC and COFDM,
conducted some signal level surveys in various locations, including the described survey around Ottawa.
The ATSC (and COFDM) waveforms transmit a known pseudorandom sequence of bits (PN-511) for 47.48 usec every 24.2 msec.
The ATSC Receiver uses this known data sequence to initialize the equalizer by minimizing
the mean square error between the received sequence and the known PN-511 sequence.
The equalizer tap settings can be observed to log the multipath strength vs delay.

And once the multipath delay is known, it can be used to find the path difference between the direct and reflected path,
using the time it takes light to travel one mile: 5.376 usec/mile.

They found that for a large number of suburban locations, the strongest multipath signals arrived within about 0.2 usec,
which is a path difference of only a couple hundred feet.
That means that your neighbors (and perhaps your interior/attic) are the strongest contributors to multipath.
(Especially your neighbor's two story house with the wire mesh stucco, Al. foil back insulation, Al. gutters, electrical wiring, etc.)

How strong? Roughly equal strength down to 10 dB below the direct path.

On the average, the multipath reflections were insignificant (20 dB below the direct path) for multipath delays
in excess of about 5 to 10 usec, which is a path differential of 1 to 2 miles.

Urban multipath delays were about twice those in surburbia.
So how much multipath should an STB accomodate?
About +/- 20 usec ought to do the job unless there is a tall skyscraper more than a few miles away from you.
So most people shouldn't expect the Zenith/LG 5th Gen chip to work miracles.
Indeed, the Linx/Micronas chip had much better performance for the 0 dB Echo test cases,
including Brazil ensembles, such as might occur for indoor and attic locations.

Of course, there were some infrequent excursions well beyond the "average".
ATSC A/74, "ATSC Recommended Practice: Receiver Performance Guidelines", 18 June 2004,
contains several dozen of these "exceptions" for test purposes.

(So why is it that this information is never collected and released from U.S firms???)

========================================================
References re ATSC Diversity Receivers:

a. "UHF Wideband Mobile Channel Measurements and Characterization using ATSC Signals with Diversity Antennas", Semmar et. al., May 2004:
http://lrts.gel.ulaval.ca/publicatio...ication_11.pdf

b. "UHF DTV Band Channel Characterization for Mobile Reception", Semmar et. al.,
Canadian Conference on Electrical and Computer Engineering, Vol.3, pp 1339-1342, 2-5 May 2004.

c. "ATSC Digital TV Receiver using Spatial Diversity Technique", Ju-Yeun Kim, et. al.,
IEEE Vehicle Technology Conference, Sep 2004.

d. "Antenna Diversity for Improved Indoor Reception of US Digital Terrestrial Television Receivers",
Meehan, IEEE Trans on Consumer Electronics, Vol. 48, No 4, Nov 2002.

e. "Joint VSB Terrestrial Digital TV Indoor Reception Project", Wetmore, Schnelle, Meehan, et. al.,
IEEE Trans on Broadcasting, Vol 48, No. 4, Dec 2000.