[linux-dvb] TechnoTrend C-1501 - Locking issues on 388Mhz

klaas de waal klaas.de.waal at gmail.com
Sun Oct 12 22:21:46 CEST 2008


On Fri, Oct 10, 2008 at 2:36 AM, hermann pitton <hermann-pitton at arcor.de>wrote:

> Hi,
>
> Am Donnerstag, den 09.10.2008, 22:15 +0200 schrieb klaas de waal:
> > Hi Jeremy,
> >
> > I have the Technotrend C-1501 now locking at 388MHz.
> > The table tda827xa_dvbt contains the settings for each frequency
> > segment.
> > The frequency values (first column) are for  the frequency plus the
> > IF, so for 388MHz
> > this is 388+5 gives 393 MHz. The table starts a new segment at 390MHz,
> > it then
> > starts to use VCO2 instead of VCO1.
> > I have now (hack, hack) changed the segment start from 390 to 395MHz
> > so
> > that the 388MHz is still tuned with VCO1, and this works OK!!
> > Like this:
> >
> > static const struct tda827xa_data tda827xa_dvbt[] = {
> >     { .lomax =  56875000, .svco = 3, .spd = 4, .scr = 0, .sbs =
> > 0, .gc3 = 1},
> >     { .lomax =  67250000, .svco = 0, .spd = 3, .scr = 0, .sbs =
> > 0, .gc3 = 1},
> >     { .lomax =  81250000, .svco = 1, .spd = 3, .scr = 0, .sbs =
> > 0, .gc3 = 1},
> >     { .lomax =  97500000, .svco = 2, .spd = 3, .scr = 0, .sbs =
> > 0, .gc3 = 1},
> >     { .lomax = 113750000, .svco = 3, .spd = 3, .scr = 0, .sbs =
> > 1, .gc3 = 1},
> >     { .lomax = 134500000, .svco = 0, .spd = 2, .scr = 0, .sbs =
> > 1, .gc3 = 1},
> >     { .lomax = 154000000, .svco = 1, .spd = 2, .scr = 0, .sbs =
> > 1, .gc3 = 1},
> >     { .lomax = 162500000, .svco = 1, .spd = 2, .scr = 0, .sbs =
> > 1, .gc3 = 1},
> >     { .lomax = 183000000, .svco = 2, .spd = 2, .scr = 0, .sbs =
> > 1, .gc3 = 1},
> >     { .lomax = 195000000, .svco = 2, .spd = 2, .scr = 0, .sbs =
> > 2, .gc3 = 1},
> >     { .lomax = 227500000, .svco = 3, .spd = 2, .scr = 0, .sbs =
> > 2, .gc3 = 1},
> >     { .lomax = 269000000, .svco = 0, .spd = 1, .scr = 0, .sbs =
> > 2, .gc3 = 1},
> >     { .lomax = 290000000, .svco = 1, .spd = 1, .scr = 0, .sbs =
> > 2, .gc3 = 1},
> >     { .lomax = 325000000, .svco = 1, .spd = 1, .scr = 0, .sbs =
> > 3, .gc3 = 1},
> > #ifdef ORIGINAL // KdW test
> >     { .lomax = 390000000, .svco = 2, .spd = 1, .scr = 0, .sbs =
> > 3, .gc3 = 1},
> > #else
> >     { .lomax = 395000000, .svco = 2, .spd = 1, .scr = 0, .sbs =
> > 3, .gc3 = 1},
> > #endif
> >     { .lomax = 455000000, .svco = 3, .spd = 1, .scr = 0, .sbs =
> > 3, .gc3 = 1},
> > etc etc
> >
> > I plan to do a test on the all frequencies in the near future, at
> > least on all the Dutch Ziggo frequencies.
> > Because I cannot test what will happen if the driver is used for DVB-T
> > (what
> > the name of the table suggests) it might be best to make a separate
> > tda827xa_dvbc table.
> >
> > About the timeout messages, they come from the SAA7134 and they happen
> > fairly random. I have looked at debug traces and everytime it happens
> > it
> > does a retry and then succeeds, so I think this can be ignored for the
> > time being.
> > Maybe you can check if the fix/hack also works for you?
> > If there is an official maintainer of this driver, maybe he can
> > comment?
> >
> > Groetjes,
> > Klaas
> >
>
> just scrolling through mails and did not look it up yet.
>
> But you likely mean tda8274a DVB-C, tda10023 and saa7146.
>
> Are we still here?
> http://www.linuxtv.org/pipermail/linux-dvb/2008-April/025634.html
>
> Please don't top post, you get more readers.
>
> Cheers,
> Hermann
>
>
> >
> > On Tue, Sep 30, 2008 at 11:18 AM, <jerremy at wordtgek.nl> wrote:
> >         Hi Klaas,
> >
> >         Perhaps its an idea to post this on the linux-dvb mailing
> >         list, if anything
> >         it keeps the subject alive.
> >
> >         I've spent an hour or so playing with several of the
> >         parameters of the
> >         demodulator (the tda10023), mostly because this was suggest in
> >         one of the
> >         older posts about this issue. However none of my efforts gave
> >         any desired
> >         result and quickly got tired of unloading / reloading my
> >         drivers (which
> >         every so often required a hard reset as well).
> >
> >         But if you find anything, that would be great ;)
> >
> >         Gr. Jerremy
> >
> >
> >         On Mon, 29 Sep 2008 11:35:03 +0200, "klaas de waal"
> >         <klaas.de.waal at gmail.com> wrote:
> >         > Hallo Jeremy,
> >         > I have exactly the same problem with my C-1501 card: tuning
> >         problems on
> >         > 388
> >         > MHz but OK on most other frequencies.
> >         > It works OK with WIndowsXP on all frequencies including the
> >         388MHz, so
> >         the
> >         > hardware is OK and it must be a software issue.
> >         > I have over the weekend put in a lot of printk for debugging
> >         but have not
> >         > found it yet.
> >         > I will keep you updated.
> >         >
> >         > Groetjes,
> >         > Klaas
> >         >
> >         >
> >         > On Thu, Sep 11, 2008 at 5:50 PM, <jerremy at wordtgek.nl>
> >         wrote:
> >         >
> >         >> Hi,
> >         >>
> >         >> This issue has come up at least once a bit more then a
> >         month ago and is
> >         >> still present in the current release of the V4L-DVB
> >         drivers. The
> >         >> Technotrend C-1501 drivers are unable to get a lock on
> >         388Mhz (and a
> >         > couple
> >         >> of other frequencies, like 682Mhz and 322Mhz, but I can
> >         only test
> >         > 388Mhz).
> >         >>
> >         >> The dmesg will mention an I2C timeout when this occurs, I'm
> >         not sure if
> >         > its
> >         >> related (as it'll randomly give those timeouts when viewing
> >         working
> >         >> channels too).
> >         >>
> >         >> I have two seperate installs of Linux (Ubuntu 8.04 64-Bit
> >         with 1
> >         > received
> >         >> and Ubuntu 8.04 32-Bit with 2 receivers) which both suffer
> >         the same
> >         >> inability to lock onto that frequency. So its unlikely to
> >         be a hardware
> >         >> problem, also the Windows drivers do not seem to have any
> >         issues.
> >         >>
> >         >> Is anyone looking into this issue? If not, what would be
> >         the place to
> >         >> experiment?
> >         >>
> >         >> Gr,
> >         >>
> >         >> Jerremy Koot
> >         >>
> >         >>
>

Hello Herman,


Thanks for your post. I had not seen this message
http://www.linuxtv.org/pipermail/linux-dvb/2008-April/025634.html
before. Probably a lot of work has been done since because the card worked
quite OK, only on the 388MHz it failed on me.

As mentioned in an earlier post, first I got my tuner working by modifying
the frequency map table tda827x_dvbt to use a different setting at 388MHz. I
have now put in a frequency map table tda827x_dvbc for DVB-C tuners only.
This works OK for me and it should not modify the behaviour with other
non-DVB-C demodulators.
With the dvb-apps/util/zap utility I get a lock on all Ziggo (Dutch cable TV
provider) frequencies. With MythTV-0.21 I can receive all the channels that
I am entitled to.

Remaining (non-fatal) issues are:

Error message:
*saa7146 (0) saa7146_i2c_writeout [irq]: timed out waiting for end of xfer*
This message comes once in a while and it happens at random places, it
cannot be correlated to a single place in the tda827x code. It really seems
random. I have read though a lot of traces and when this error happens it
does a retry and the retry succeeds.

Error message:
*tda827x: tda827x_config not defined, cannot set LNA gain!*
This message comes on every tuning action. I think that the TT-C1501 board
does not have a separate LNA outside the tuner so this is not really an
error. It shoudl be possible to fill in a tda827x_config struct so that it
does not give this error message. I could change the code to achieve this
but I do not want to change everything at once.

Minor issue:
*Signal strength not the same across the band.*
On the frequencies I can receive (from 252MHz to 826,75MHz) there is quite a
lot of signal strength variation; on MythTV display ranging from 60+ to 80+
percent, on zap output from 90 to d0 or so. On a KNC-One DVB-C card (cu1216
with tda10023 plus saa7146) the signal strength is practically constant.

I have attached my version of the tda827x.c driver, both as complete source
and as a diff against the hg version of 10 october.

Groetjes,
Klaas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.linuxtv.org/pipermail/linux-dvb/attachments/20081012/78760cc9/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tda827x.c.gz
Type: application/x-gzip
Size: 4946 bytes
Desc: not available
Url : http://www.linuxtv.org/pipermail/linux-dvb/attachments/20081012/78760cc9/attachment.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tda827x.c.patch.gz
Type: application/x-gzip
Size: 1214 bytes
Desc: not available
Url : http://www.linuxtv.org/pipermail/linux-dvb/attachments/20081012/78760cc9/attachment-0001.bin 


More information about the linux-dvb mailing list