[pvrusb2] Hauppauge HVR-1950 Not Locking Signals Less Than 60%?

[Roger]

After much trouble shooting or debugging, I've found the reason was due to the 
signal(s) of the frequencies being to hot or too strong for the Hauppauge tuner 
chips!

On my 2013 Samsung ATSC televisions, I'm seeing a signal from one solid weak 
station or frequency being around 10-20%.  The remaining four or five 
frequencies from the stronger stations are received at around 80-90% on my 
Samsung ATSC televisions.

On the Hauppauge ATSC USB HVR-1950 & PCI-E HVR-2250 tuners, the signals were 
reporting in at extremely weak, bouncing from 0% to 30 or 50% if I were lucky, 
only locking onto a few channels or frequencies.  This was while being 
connected directly to one of the Channel Master television distribution 
amplifier feeds.


Using the following command to detect and see signal levels from the tuner 
chips:

$ dvbv5-scan --adapter 1 -I CHANNEL ~/.mplayer/channels-initial_data_v3.conf

I found dvbv5-scan output invaluable or excellent while trouble shooting, as 
dvbv5-scan output signal strength almost in real time.  The file 
channels-initial_data_v3.conf is basically just a listing of frequencies, so 
users can easily add their own frequencies for better fine tuning or testing.

In summary, I've added a simple two-way splitter onto one of the feed lines 
from the Channel Master television antenna distribution amplifier and 
immediately noticed the Hauppauge tuners finally detect the previously not 
found frequencies, as well as the other frequencies registering more sane 
signal strength values of 85% or more instead of only 40% previously!  A little 
later, I just replaced the two-way splitter with a five-way splitter and the 
remaining weaker stations were finally detected by the Hauppauge tuners.

Amazing.  Results almost the opposite of what one would expect!  But to those 
that understand what is going on here, easily explained but still easily 
over-looked.

I'll follow-up later if I find addition info, but think I can pretty much close 
the book on this incidient.


--
Roger
http://rogerx.freeshell.org/