Originally Posted by dr1394
However, that long of a delay suggests a geometry that should be easy to overcome with a directional antenna. That is, the multipath direction is either off to the side or behind the antenna.
Your comment got me to thinking about my multipath situation and just exactly where it's all coming from. I think when you're blocked to the transmitter but you receive strong reflections, that it's very difficult to overcome the problem.
Peter mentioned he was using a 91XG and I looked up that antenna on the HDTV Primer page to see it's pattern and compare it other UHF antennas.http://www.hdtvprimer.com/ANTENNAS/comparing.html
The 91XG has the best pattern of any of the antennas reviewed. The response over the back 180 degrees or more of the antenna is running around -30dB in both the horizontal and vertical planes. No other antenna comes close.
In an earlier post Peter showed us the spectrum analyzer displays for KQED on the direct path and the reflected path. The latter was 10 dB stronger! I don't think it's surprising that even a 91XG isn't good enough to reject the reflection enough to allow reception. And who knows how many other peaks in reflections there might be.
I decided to take look at the Walnut Grove stations for signal strength on the direct path and rotate my antennas to look for other peaks. Walnut Grove is almost due west of here and I know from past experience that I have a lot of reflections from the east. There's a range of 3500' - 4500' mountains that span NE through SE of here. I've been up on one of the peaks with a battery operated TV and all the analog stations were very strong up there. I remember pointing my antennas in that direction and although the analog stations were strong, the pictures were completely scrambled. It was impossible to make out anything.
I'm using a pair of 91XGs as anything less simply doesn't work here. Here's what I got looking at the AGC value on my Sony plus the SNR value when on the direct path. (Lower AGC = Stronger signal)
Station - SNR - AGC Direct - AGC Back
KCRA - 19 - 24 - 24
KOVR - 25 - 24 - 24
KSPX - 29 - 23 - 25
KMAX - 19 - 26 - 28
KTXL - 17 - 37 - 33 (weaker than average)
KQCA - 26 - 27 - 34
KVIE - 24 - 26 - 25
KXTV - 24 - 25 - 25
KMMW - 31 - 24 - 38
I was stunned to see such strong signals coming from the east on most stations. Even worse though was that the lowest numbers stayed fixed with antenna headings from 60 - 120 degrees for all the Walnut Grove stations. That's 60 degrees of strong reflections while on the direct path the strongest signal was only over 10 degrees.
KVIE and KXTV are both VHF and for those I use a pair of ch 7 - 13 log periodics. I know the F/B on those channels is >20 dB.
I included data for KMMW since it's low power (250 w in my direction) and LOS. It acts like you expect for an LOS station, high SNR, strong signal, and comparatively weak in every other direction. The reading of 38 was only at one specific heading, much higher numbers in all other directions.
Peter could try a pair of 91XGs. That made a big difference for me but I wouldn't want to put money on that fixing his problem.
What people with lot of multipath need is an antenna that can reject all far off main lobe signals by around 40 dB. I don't know of any antennas that can do that.