AW: [mythtv] (Hauppauge DVB-s Nova / Hauppauge WinTV DVB-s) patch

[Kenneth Aafløy]

On Monday 14 June 2004 18:32, Christian Höhle wrote:
> > What is the equipment connected to the card?
> > Sounds like you only have a single LNB
> > connected directly to the card by the comments in the patch?
>
> That's right. I have a two single LNBs connected to my two cards.

So you have an universal LNB which does not support setting the polarization 
via voltage and high/low via tone..strange.

> > Also, have you tried the diseqc_type 1 (ToneSwitch) and 0 
> > (ToneVoltageLnb)? 
>
> I tried both, but the result stays the same.
>
> > Most diseqc switches have compatability mode with ToneSwitch,
> > and if you only have a LNB directly connected to your card,
> > you would be better off with just using ToneVoltageLnb,
> > since that gives much faster zapping.
>
> What do you mean with "faster zapping"? The time between starting a tuning
> and the message "successfully tuned to ..." is much less than a second for
> me. Then, mythtv seems to prebuffer a bit which lasts for approx. 2
> seconds, then I have got the output on tv.

Pretuned channel:
2004-06-14 16:50:36 DVB#0 Successfully tuned to channel 3 (50ms).
Channel on same satellite, but different frequency, same pol:
2004-06-14 18:02:55 DVB#0 Successfully tuned to channel 21 (66ms).
Different satellite, different pol:
2004-06-14 19:07:51 DVB#0 Successfully tuned to channel 5 (113ms).

These are very variable, because I have a CAM which does not always want to 
start decrypting right away (in the same order as above):
2004-06-14 17:21:25 First frame after pause. (224ms)
2004-06-14 18:03:02 First frame after pause. (4663ms)
2004-06-14 19:07:56 First frame after pause. (2622ms)

> I use diseq_type 0 at the moment, that's why my changes are located mainly
> inside ToneVoltageLnb().
> Before I clean it up and adopt the changes to other functions which could
> need it, I wanted to hear from a "professional" wether my patch isn't
> useless or stupid at all ;-)
>
> It seems to me that the szap-code contains a polarization- and
> frequency-dependent diseq command, whereas neither ResetDiseqc() nor any of
> ToneSwitch or ToneVoltageLnb contains something similar (as far as I can
> see).
>
> //cmd.cmd.msg[3] = 0xf0 | (((0 * 4) & 0x0f)
> //                 | ((tuning.tone == SEC_TONE_ON) ? 1 : 0) |
> //                   ((tuning.voltage == SEC_VOLTAGE_13) ? 0 : 2));
>
> On the other hand, szap sets the tone to "SEC_TONE_OFF" before applying the
> other settings. Perhaps my cards are a little bit non-standard in that way.

I'm no diseqc expert or professional, but could you check if a clean cvs 
version with the following patch would work ok for you?

I can't see anything else different if you use diseqc_type == 2, except for 
the DiseqcReset that essentially only powers up the diseqc equipment.

Kenneth

Index: dvbdiseqc.cpp
===================================================================
RCS file: /var/lib/mythcvs/mythtv/libs/libmythtv/dvbdiseqc.cpp,v
retrieving revision 1.4
diff -u -r1.4 dvbdiseqc.cpp
--- dvbdiseqc.cpp       11 Apr 2004 15:14:39 -0000      1.4
+++ dvbdiseqc.cpp       14 Jun 2004 17:19:04 -0000
@@ -223,7 +223,7 @@
         }
     }

-    if (ioctl(fd_frontend, FE_DISEQC_SEND_BURST, SEC_MINI_A ) == -1)
+    if (ioctl(fd_frontend, FE_DISEQC_SEND_BURST, SEC_MINI_B ) == -1)
     {
         ERRNO("FE_DISEQC_SEND_BURST failed");
         return false;