[linux-dvb] NOVA-t and reception problems

[Tomi Ollila <Tomi.Ollila@iki.fi>]

Hi

Whatever I try to do, I can not get any decent reception out of my
WinTV-NOVA-t -card -- I am getting quite confident that the problem
is with the harware, but until I go to the dealer with the card, I try
here as the last resort.

I have decent UHF antenna on my roof, w/ 20m cable to the computer
(straight cable from the antenna). Where I live, there is 3 uhf "channels"
used for digital broadcasting, 32, 44 and 46. There is also 3 analog UHF
channels here, 24, 35 and 53 - and all those channels comes chryscal clear
to my "ordinary" analog television with that antenna.

I started my tries with the "original" dvb driver -- latest downloaded from
CVS some days ago, and I've used VDR and dvbstream  to test the reception:
One channel comes almost right -- some messy macroblocks in about every
frame, but otherwise. All the rest comes with almost every macroblock messy
-- if I can even get any... All the tuning parameters have been set to
the correct one in all programs I've used, and I even asked from digita
(here in finland) what the correct values are and the info I received from
there was the same I've used mostly in my tests...

Latests tests I've been doing with NEWSTRUCT driver. I tried to port
dvbstream to use the changed api -- I even started to get some streams
with it, but the results were worse than with original version using
"oldstruct" driver -- I can't really test if I did any mistakes porting
the code since I haven't been able to get clear signal in any of my tests.

The last test I did with tzap application; I made channels.conf file
from the one I've been using with VDR by just deleting unnecessary
fields and adding tuning parameters. And, in every channel, the tzap
output was quite identical:

status 1f | signal 2d2d | snr bdbd | ber 00000607 | unc 0000005f | FE_HAS_LOCK
status 1f | signal 2d2d | snr abab | ber 000006bf | unc 00000039 | FE_HAS_LOCK
status 1f | signal 2d2d | snr c0c0 | ber 00000842 | unc 00000032 | FE_HAS_LOCK

signal lowest 2c2c, highest 2d2d
snr    lowest a9a9, highest c4c4
ber    lowest 05ad, highest 22d9
unc    lowest 0000, highest 00f0

What does it mean, that signal strength has such a small value, but 
signal to noise about the values that are considered as `good signal'?

I've tested oldstuct with 2.4.9-xx, 2.4.18-3 (<-redhat) and 2.4.19-pre9,
-- with quite similar results, and newstruct with 2.4.19-pre9.

Any thoughts?

Tomi


-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.