OTA - why so variable?

Discussion in 'DirecTV TiVo Powered PVRs & Receivers' started by rjnerd, Sep 20, 2007.

  1. rjnerd

    rjnerd New Member

    43
    0
    May 27, 2007
    My OTA signal reception for a couple of the channels varies a lot. The thing will be sitting at 80%, then all of a sudden drop to 20, then back to 80. Varies 3-4% every second, every 5-10 seconds it drops to near or actual zero. Not every station, all but two are rock solid all the time (whatever they read, usually in the mid 90's, it dosn't change)

    Its a clear, dry day. I am in a very high signal area (about 5 miles to the antenna's) No trees waving in front of my antenna, etc. Nice clear shot to the towers , drag out bino's and I can see them. (all of the local stations are on the same hill) The stations with problems share a tower with rock solid ones.

    Antenna is indoor directional (one of the phillips un-amplified stamped tin yagi's) Its not moving. Its at a window, about 15' off the ground. Feeds two h10's from a single wideband splitter (3.5db). Less than 10' of cable from the back of the antenna, to the connection on the box. (and I have tried it without the splitter)

    I could understand constant poor reception. I could understand dropouts if I were near the noise floor. I assume my problems are mostly multipath. But since multipath is usually structural, I wouldn't expect it to change, especially second to second. (I suppose it could be passing traffic, but I would expect it to affect more than one channel)

    This happens on both tuners, of both boxes. In one of the boxes, its so lame, I can't use OTA on that channel. The other box was working pretty well, but decided to get worse recently.

    Any suggestions for improving antenna directionality? I don't have too many placement options, as the house is covered in foil insulation, and makes a good Faraday cage execpt by windows...
     
  2. JimSpence

    JimSpence Just hangin'

    30,905
    36
    Sep 19, 2001
    Binghamton, NY
    Multipath isn't just structural. I installed a large antenna at one end of my house on the peak. I couldn't get a solid signal there. I moved the antenna about 25' along the peak and now I get a much more stable signal. Not perfect, but a lot better. At least until the leave fall, when the reflected signals are no longer attenuated as much. It is a matter of trial and error, just as it was in the old analog days to try to eliminate the ghosts.

    Outside as high as you can get is better.

    Another thing to take into consideration is that your locals aren't transmitting at the same power. Try removing the splitter to see if that helps. I also have an antenna mounted amplifier.
     
  3. milominderbinder

    milominderbinder New Member

    434
    0
    Dec 18, 2006
    Chicagoland
    I also think your problem is multipath.

    Imagine someone yelling at you in a cave. Your house is an Radio Frequency echo chamber. For an SD TV, you see ghosting. For an HD signal this is called Multipath. You either see bad pixelation or there are so many bits piled on top of each other, you can't can't lock onto anything. Signal levels may swing wildly as the receiver tries to decipher a mass of signals.

    With all the echoing you can't make out anything. The antenna 5 miles away actually makes the problem much worse. What should be small signal reflections are huge at 5 miles out. People aren't just yelling in the cave, they have a megaphone.

    Though 1
    A good Directional antenna can help reduce multipath a lot. "Highly Directional" is better.

    Thought 2
    EDIT: Some people need to amplify their signal. I think you need the opposite. Consider a variable attenuator. It will let you dial down your signal.

    Thought 3
    Probably the thought you will like least is to move the antenna outside. If you do not have a good directional antenna it may do you no good.

    The good news is that this should be fairly easy to remedy and for not much money.

    One other thought is that splitters by their nature can also induce reflections. If your splitter is really only 3.5 dB it much be a Channel Master, Channel Plus, or Spaun. The others can be up to 3 times worse as the splitter comparison shows.

    - Craig
     
  4. rjnerd

    rjnerd New Member

    43
    0
    May 27, 2007
    Thanks for the tips...

    Turns out the antenna is one of the zenith "silver sensor" models. Unfortunately, I can't find an angle and F/B ratio spec for it. Its a log-periodic design, so it has some directionality. Since the gain is less than 10db, its likely >40 degrees. Any particular favorites for directionality? (did find one that claimed 25 degrees but didn't give a F/B spec... its one of those big grid and bowtie jobs..)

    I have an attenuator that should cover that range but its not set up for TV connectors. (microwave test gear, not down to consumer standards). I should have a TV spec one somewhere, but they are cheap, it will be faster to hit rat shack than search....

    The splitter is branded "monster cable" - no telling who made it for them. It does claim 3.5db and 1.5ghz bandwith... I did try it without the splitter, and still had the problem.

    I did notice that it is sensitive to the post-antenna cable routing. Going from a tangled heap to straight runs above the rest of the gear seems to help. Figured coax wouldn't really care. Perhaps I should do some ground work...
     
  5. milominderbinder

    milominderbinder New Member

    434
    0
    Dec 18, 2006
    Chicagoland
    First of all anyone with microwave gear is a man after my own heart. Are those connectors BNC's by any chance?

    It sounds like your cables may not be new and well sheilded as well?

    The links I gave you were to SolidSignal.com. If you call them, they will help you find the perfect antenna in a matter of minutes.

    I am a big fan of Channel Master and Winegard antennas. They have a great selection.

    - Craig
     
  6. milominderbinder

    milominderbinder New Member

    434
    0
    Dec 18, 2006
    Chicagoland
    I kept digging and find some good antennas for you from Winegard:

    http://www.solidsignal.com/winegard_antenna_chart.asp

    Lots of people just swear by the directionallity of the SquareShooter for multipath but I bet Solidsignal will have cheaper options.

    I hope that you have a blast with this!

    - Craig
     
  7. rjnerd

    rjnerd New Member

    43
    0
    May 27, 2007
    The existing cables are fairly fresh, (hell one of them came with the antenna) but they are pretty flexible, and I am sure were lowest bidder coax, so no telling shield quality...

    The connectors are SMA's I think, very tiny suckers - but I do have some pigtails that adapt them to BNC. Its always interesting what shows up at swap meets. Will check out the antenna suppliers.
     
  8. TyroneShoes

    TyroneShoes HD evangelist

    3,604
    0
    Sep 6, 2004
    "Most people" actually do NOT need amplification. Some do, but usually only when they are in a fringe reception area. And amplification can not improve multipath at all, so it will not improve reception that is plagued by it. The HR10 sometimes needs more attenuation than one would expect. Signal that is too strong can mimic multipath (it manifests very similarly) and an attenuator is an easy fix, assuming it is the right fix. But only better directionality or moving to a less-reflective location can really help with multipath.

    I wouldn't put too much stock in that splitter chart. Just because someone can bang out an Excel file doesn't mean the information is good, and even if its good it still has to be meaningful. There is no information regarding the frequency rating for each splitter or even what frequency is being measured. For instance, if you measure a signal at 1 GHz on a splitter designed for 890, you may see more loss, or you may not. To really do this right, you would have to use a sweep generator and a spectrum analysis sweep receiver, and also do return-loss measurements or VSWR tests, which is how manufacturers and MATV/CATV labs do it.

    Digital signals will also measure with an offset when measured by a conventional analog signal level meter, simply because the modulation characteristics of both are so different (one has spiked carriers, the other has its energy equally spread throughout the channel and has no real carrier spikes). That has to be considered as well (IOW, comparitive tests using real-world reception must use one or the other). It also makes no sense to compare the loss of hybrid splitters with different output port counts in the same list, as they can't really be compared. Those with more output ports will naturally have more loss.

    Their also has to be some understanding of how the information impacts the question at hand. Signal level is far less important for digital signals, which can be received at much lower levels perfectly, than can analog signals. Adding or subtracting a few dB will likely not change reception characteristics of ATSC channels making the differences between equal-port splitters as listed here largely unimportant, but NTSC channels need high levels to over-ride noise, so loss becomes important for that. Whether these numbers are important depends upon the application, and even more on the particular distribution and reception situation.

    If you are buying a splitter, about the only thing to really look for is the frequency rating. If it's rated to 1 GHz or to 2.5 GHz, you can generally expect them to perform similarly for ATSC signals (which are generally below 700 MHz anyway), and even if they vary a couple of dB in output level, that rarely if ever impacts ATSC reception.

    Concern about internal reflections is also not really something to worrry about. It's true that "better" splitters reflect less signal than "worse" splitters, but that really won't make a difference. The time variable in a reflection that short is well within the active eq parameters of any modern ATSC tuner, which means they will not affect reception. The multipath signals that plague ATSC reception fall into a much longer time constant, such as the difference in time it takes direct compared to the time it takes reflected off of a watertower adjacent to the reception location.

    The reason high multipath causes wildly swinging signal levels, is because although the nature of the reflection is largely the same from moment to moment, the amount of interference in how well the signal is received can vary greatly, and does so dynamically. Multipath causes the same info to be received multiple times in the data stream, which can fool the tuner into thinking it is relevant data, confusing it.

    But not in a constant manner. There are moments of light confusion and moments of heavy confusion. Its similar to when you are trying to overhear a conversation that is far away or in a noisy environment. Some words are picked up, while some aren't. One moment, your "reception" is "in the 90s", and the next its in the 20s.

    Decoding MPEG-encoded data uses a similar kind of intelligence, and in a multipath environment, some bit words can be interpreted intelligently by the tuner while some are obliterated and make no sense at all. Some have error correction or other data to supplement the interpretation, while sometimes the corrected data is obliterated as well. Also, when there are multiple paths for reception, the active eq dynamically varies, trying to gamely adjust on the fly to interference as it sees it, and the success of that is dynamic as well. The end result is wild swings in how successfully the signal is decoded from moment to moment, which is reflected by the readings on a digital meter.
     

Share This Page