On Friday 23 Sep 2005 05:04, Oliver Endriss wrote: > Johannes Stezenbach wrote: > > On Wed, Sep 21, 2005 Oliver Endriss wrote: > > > Imho the 'enhanced' code looks broken: > > > > cvs annotate blames this to Andrew, maybe he can enlighten us? > > Otherwise, if it is obviously broken we should just remove it. > > Yep, the fine-tuning code was introduced in Rev 1.31: > | Revision 1.31 - (view) (download) (as text) (annotate) - [select for > | diffs] Mon Feb 23 23:20:29 2004 UTC (18 months, 4 weeks ago) by quincy > | > | Fixes to sort out tuning at low symbol rates > > Hm, I turned-on the fine-tuning code for my BSRU6 tuner and did some > tests. > > Imho this code can be fixed by moving > state->tuner_frequency = p->frequency; > into the else clauses. tuner_frequency must not be set for fine-tuning. > > Anyway, recovery seems to be faster without fine-tuning, at least with > my Nexus, higher symbol rates and a good signal... Hmm, weird, I added the enhanced code because thats what hauppauge appeared to be doing, and I thought it seemed to work better at the time.... however looks like that isn't the case. I've no problem if you want to rip it out. BTW: I finally tracked down why I was having extreme lock problems on one site - satellite installer problem - they'd mounted the arm with the LNB on it out of position!! (and they claimed they'd tested it!)
