[linux-dvb] Strage SkyStar HD modprobe
abraham.manu at gmail.com
Fri Oct 19 22:07:59 CEST 2007
Konstantin Dimitrov wrote:
> On 10/18/07, Manu Abraham <abraham.manu at gmail.com> wrote:
>> @ all
>> Reinhard Nissl wrote:
>>> Manu Abraham wrote:
>>>> One question to all the folks how are having I2C transaction errors.
>>>> Do you have the CI board connected ? If so can you please try the
>>>> same _without_ the CI board connected ?
>>> FYI: as you can see from my picture posted earlier in this thread, I do
>>> not have a CI board attached to the card.
>> Can you guys apply these two patches to the current multiproto head
>> and check whether any issues still ?
> this morning applied the patches to changeset 5890 (74d544ae8190),
> modprobe problem is hard to reproduce, so i don't know if it's finally
> fixed or not, but the time will show that.
> locking is still very unreliable, maybe even more unreliable than
> before, because on the tests i run with patched changeset 5890 i got:
> 91 locks out of 100 tries on Astra 19.2E, DVB-S, NRJ Hits
> 90 locks out of 100 tries on Astra 19.2E, DVB-S2, Pro7HD
> actually it's the worst result on DVB-S2 i got till now, i always got
> less than 10 failures till now.
Can someone verify whether the STB6100 has component values as in
this image ? http://abraham.manu.googlepages.com/snapshot9.png
in the problematic case ?
If one is seeing ANALOG CARRIER from what i can see, this is due to a
large Noise Floor, the large noise floor can be caused from phase errors
between the I and Q inputs. This can happen as far as i can guess a
wrong setting in the pump setting. The pump settings are calculated
from default reference values. If the design has been changed from
the default then there can be issues, based on the explained.
(This is a guess, as currently i can't see clearly why there can be an
issue. Definitely i need more proper inputs, from the users)
To verify this: Will require inputs from all STB0899 users what they are
More information about the linux-dvb