On May 9, 2009 08:34:31 am Andy Walls wrote: > On Sat, 2009-05-09 at 01:19 -0400, Nick Nobody wrote: > > On Friday 08 May 2009 22:47:15 Andy Walls wrote: > > > On Fri, 2009-05-08 at 21:58 -0400, Nick Nobody wrote: > > > > On May 8, 2009 08:04:11 pm Andy Walls wrote: > > > > > On Fri, 2009-05-08 at 16:56 -0400, Nick Nobody wrote: > > > > Everything is grounded properly and I'm using the same coax everywhere > > (75 Ohm, RG-6, 95% shielded, rated for 2200MHz). > > OK. > > > I'm using a splice at the moment > > to reduce any potential loss that I might incur from using a splitter > > (only one device is connected at a time). > > That's bad, if by splice you mean two cables wired to the incoming feed > at the same time. In such a case, without a splitter, you'll get signal > (voltage wave) reflections due to impedance mismatches (which will look > like noise) and the power is splitting anyway. Also at the end of an > unterminated cable you'll get reflections from impedance mismatch that > will travel back up and split down to the other end - more apparent > noise. > > Reflections essentially show up as increased noise floor in the tuner, > degraing your signal to noise ratio.
Sorry, I wasn't very clear in my sleepy state. Right now I have one length of wire that comes down from the antenna (which is on a mast above my roof) and gets terminated inside the house to something like this: http://imgur.com/JG0L.jpg >From there I either connect the television or the capture card, never both at the same time as that would be impossible :P > With digital modulations that have lots of Forward Error Correction > (FEC), the Bit Error Rate (BER) vs Signal to Noise Ratio (SNR) curves > are close to a "brick-wall". There is a very narrow band (~ 2 dB) where > the signal will go from a very low BER (good picture) to a very high BER > (no picture) as your S/N ratio drops. > > This is the joy of digital TV reception that the FCC did not convey top > the public. :( The FCC didn't fully assess the impacts to OTA > broadcast to outlying areas either, I suspect. What does > > http://www.antennaweb.org > > say you can expect for digital reception of WFFF at your location? What > is the bearing that antennaweb.org say you should point you antenna vs > the bearing at which you have it? (antennaweb uses magnetic north) > Antennaweb should be running a Longley-Rice propagation model to give > you some decent expectation of availability. Sadly I can't use that because I don't live in the US. I tried entering my coordinates and it only lists two stations from Plattsburgh. One at heading 201 (WCFE) and the other at heading 188 (WWBI). The WFFF transmitter is southeast of me, I'd estimate the heading to be about 150 degrees and the distance (using google maps) is about 142km (~88mi) "as the crow flies". This may seem far, but there are 3 other transmitters in that area that I'm able to get great signal from (WCAX, WPTZ and WETK). > > This will eventually be replaced with a > > splitter once I get everything sorted. > > Bring an unspliced, unsplit line straight to the HVR-1600's digital TV > input. If that doesn't work well, you'll need a higher directivity > antenna and/or a low noise preamp. As mentioned above, that's what I'm already doing, I'm just bad at explaining. > > Right now the HVR-1600 is working properly, I can get FOX but I suspect > > that it's because it's night time. I'll report back in the morning once > > the sun comes back out. > > Weather fronts and temperature varaition betwenn you and the > transmitting station also make a difference at times. Sure enough, as soon as the sun comes out, "scan" no longer finds the channel nor will gnutv get a lock on the signal. > > One thing that I don't understand is how come the television reports that > > it's getting such a high signal strength (about 75%, and I still get a > > picture all the way down to about 30% while turning the antenna) > > What does you TV report as SNR? At what SNR does your TV start getting > errors on a station? How close is WFFF's ditial station to that > threshold SNR on your TV? > > My Sony TV will start getting errors at an SNR of 13 dB for ATSC OTA > 8VSB modulation. The channel will have no picture if the SNR drops to > 11 dB. IIRC... My TV only reports "signal strength" in percent and no SNR, not very handy,,, > > while the capture card > > can't get a lock > > Managing SNR by obtaining a good receive system noise figure makes a lot > of difference. Remeber that your TV has lots of power and space for > components to build a much more complex receiver with better, but > larger, components and good shielding than a card in your PC's PCI slot. > > For example, I doubt your TV uses a tiny silicon tuner chip like the > MXL5005s. It probably uses larger components with better noise figures > and less loss. > > Also the tuner AGC take over point (TOP) between the MXL5005s and > CX24777 (aka S5H1409) could be set suboptimally in the linux drivers, > resulting in a suboptimal receiver system nosie figure. I have not > tried to experiment with this as I could only adjust the TOP on the > S5H1409 properly. That makes perfect sense, thanks for that clarification. > > (I can't verify what the signal level is on the capture card > > because femon -H always reports "signal 0% | snr 0%"). Are all ATSC > > capture cards this picky? > > The 0% signal is due to a s5h1409 driver deficiency, ignore it. The SNR > should be around 2.3 or 2.4 dB as reported by the s5h1409 driver for > ATSC OTA reception to work. > > Receiver implementation makes the difference in the face of low incoming > S/N ratio. Receivers that have more space and power for components will > typically have better performance. The amount of HVR-1600 real-estate > dedicated to the MXl5005s (in the small tuner can) and the CX24777, > compared to the real-estate afforded to the digital tuner components in > your TV, is probably quite different. > > In the face of low incoming S/N, you best bet is to reduce the noise to > as low a level as possible (eliminate EMI and reflections due to > impedance mismatches) and manage the total *system* noise figure. > > The noise figures (NF) of components before the first gain stage > (cables, connectors, etc.) will dominate the noise figure of your > receive system: > > http://en.wikipedia.org/wiki/Friis_formulas_for_noise > > The NF in dB for a cable, connector, or other passive component is the > same as it's loss in dB. So looking at Friis' equation, to get the best > possible receive system noise figure, an LNA (low noise amplifier) as > close to the antenna as possible is what one wants to manage your > receive system NF. > > > Regards, > Andy > So it looks like my only option is to try out a pre-amp, I'll see if I can find one around here at a reasonable price. Thanks, nick _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
