rogerheflin at gmail
Feb 23, 2008, 11:57 AM
Post #34 of 36
(688 views)
Permalink
|
backuppc [at] sundquist wrote:
> On Feb 22, 2008, 12:37 PM, gull wrote:
>
>> On Feb 22, 2008, at 11:40 AM, Roger Heflin wrote:
>>
>>> One thing that you can do, is test each antenna separately and note
>>> what channels each gets and how strong it is, if both together are
>>> worse than one single you know some canceling is going on.
>> Yes, and to amplify (heh heh) on what Brian said, phase-canceling is
>> very frequency dependent. If the canceled frequencies don't land on a
>> channel you care about, it's not a problem. If it does, you can shift
>> the canceled frequency just by changing the distance between the two
>> antennas by a few inches, or by turning them enough to put the
>> affected channel in the null of one of the antennas.
>>
>> All antenna work contains some measure of "cut and try" because there
>> are so many real-world effects that are difficult to model. If you can
>> make it work well enough for your purposes, feel free to ignore any
>> math that says it shouldn't. ;)
>
> Yes, this is a great plan. I will try each antenna individually and
> tweak the aim to get the best reception and then compare it to when both
> are hooked up. If the signal is much weaker when combined, I will try
> small adjustments to see of steer out of a null for the channels I need.
> Maybe even set up a second attic mast and separate then laterally.
>
> But first I am going to go out and enjoy some of that "bad weather" and
> do some XC skiing today! (there goes my nerd credentials...)
>
> On Feb 22, 2008, 12:39 PM Craig wrote:
>
>> Any chance you can still return the Radio Shack antennas?
>
> Too late now.
>
>> If not, I'd try connecting one and see if it will receive enough out
>> the back to get the channels you want.
>
> After I get the best aim on the north/norhwest one, I'll see if I can
> get the big ABC/CBS/NBC towers located not too far to the south out
> the back of that same antenna.
>
>> SiliconDust provides a utility to work with the HDHomeRun that can
>> help for testing reception. Do a full channel scan on one of the
>> tuners, which will take several minutes, with:
>>
>> hdhomerun_config FFFFFFFF scan /tuner0 scan0.txt
>>
>> Focus on the bits of interest with:
>>
>> grep -B1 -A1 '8vsb' scan0.txt
>>
>> Pick a channel that you want to optimize for and check your reception:
>>
>> hdhomerun_config FFFFFFFF set /tuner0/channel 8vsb:<channel/Hz>
>>
>> hdhomerun_config FFFFFFFF get /tuner0/status
>>
>> Keep checking the status as you aim the antenna...
>
> The pcHDTV came with two great utilties that I'll probably use. First
> is dtvsignal. Here's the current output using my pcHDTV 5500:
>
> user [at] mytht:~/dvb-atsc-tools-1.0.7$ ./dtvsignal 34
> using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0'
> tuning to 593000000 Hz
> video pid 0x0021, audio pid 0x0024
> dtvsignal ver 1.0.7 - by Jack Kelliher (c) 2002-2007
> channel = 34 freq = 593000000Hz
> 30db 0% 25% 50% 75% 100%
> Signal: | . : . | ._____:_____._____|
> Signal: 048 ------------------------
>
> (the signal bar on the last line is "live". Not doing too well on
> channel 34 [Grand Island] today)
>
> For comparison, here's CBC from the CN tower:
>
> channel = 20 freq = 509000000Hz
> 30db 0% 25% 50% 75% 100%
> Signal: | . : . | ._____:_____._____|
> Signal: 011 -----
>
> The other tool is dtvsnr:
>
> Grand Island:
>
> channel = 34 freq = 593000000Hz
> dB -10 0 10 20 30 40
> SNR: | | | | | |
> SNR: 017 ============================
>
> CN Tower:
>
> channel = 20 freq = 509000000Hz
> dB -10 0 10 20 30 40
> SNR: | | | | | |
> SNR: 003 --------------
>
> The hills south of me:
>
> channel = 38 freq = 617000000Hz
> dB -10 0 10 20 30 40
> SNR: | | | | | |
> SNR: 026 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
>
> I'll see what I can get tonight on these using just a single well-aimed
> antenna and with each of the tuners.
>
> Thanks again for all the help.
>
> J.S.
>
J.S.,
A suggest mod for dtvsignal.c this will collect signal for 10 seconds and
give you an average over the 10 seconds, you may want to increase it to more
than 10 seconds, I found the initial long lock time on some channels seemed
to affect it more, maybe just have it ignore the first few seconds, and then
average the rest out.
336,337c336
< unsigned long count=0;
< unsigned long sig_total=0;
---
>
368c367,368
< while(count < 1000) {
---
>
> while(1) {
401d400
< sig_total += signal;
404d402
< count ++;
406d403
< fprintf(stderr,"\nChannel = %3d Signal Average %3ld\n",chan,sig_total/count);
And one might use a script like this:
channel=2
#while [ ${channel} -lt 100 ] ; do
for channel in 7 18 21 23 24 28 31 34 36 42 44 47 48 51 ; do
dtvsignal -dvb 0 ${channel} 2>&1 | grep "Signal Average"
# channel=`expr ${channel} + 1`
done
Run first with the "for channel" commented out and with the while/channel=
uncommented and then run the for channel with the list of actual channels
that you find.
Roger
_______________________________________________
mythtv-users mailing list
mythtv-users [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
|
