[linux-dvb] Re: qt1010 / m9206 / gl861 / au6610 testers wanted

Jan Jan.Nijs at scarlet.be
Thu Jan 25 08:40:58 CET 2007


Jan wrote:
> I will just need more time to perform a decent analysis of the 
> settings that differ on my side and in mercurial.
> I will first test the "zl10353_single_write(fe, 0x5a, 0x0c);" line I 
> needed to add in the past.
> Adding this line worked for me, but this caused problems for comabug.
>
Hello,

today I did a quick test. I added the line above to the latest "megasky" 
checkout:

diff -uNr /tmp/megasky/linux/drivers/media/dvb/frontends/zl10353.c 
/tmp/megasky.jan/linux/drivers/media/dvb/frontends/zl10353.c
--- /tmp/megasky/linux/drivers/media/dvb/frontends/zl10353.c    
2007-01-25 07:41:03.000000000 +0100
+++ /tmp/megasky.jan/linux/drivers/media/dvb/frontends/zl10353.c        
2007-01-25 07:39:35.000000000 +0100
@@ -144,6 +144,9 @@
        zl10353_single_write(fe, 0x5E, 0x00);
        zl10353_single_write(fe, 0x6C, 0xCD);
        zl10353_single_write(fe, 0x6D, 0x7E);
+
+    zl10353_single_write(fe, 0x5a, 0x0c);
+
        fe->ops.i2c_gate_ctrl(fe, 0); /* close i2c_gate */

        // if there is no attached secondary tuner, we call set_params 
to program


 and bang I got a lock:

vector frontends # dvbtune -f 482000000
Using DVB card "Zarlink ZL10353 DVB-T"
tuning DVB-T (in United Kingdom) to 482000000 Hz
polling....
Getting frontend event
FE_STATUS:
polling....
Getting frontend event
FE_STATUS: FE_HAS_LOCK FE_HAS_CARRIER FE_HAS_VITERBI FE_HAS_SYNC
Event:  Frequency: 492600000
        SymbolRate: 0
        FEC_inner:  2

Bit error rate: 0
Signal strength: 0
SNR: 43433
FE_STATUS: FE_HAS_LOCK FE_HAS_CARRIER FE_HAS_VITERBI FE_HAS_SYNC

Frequency 492600KHz is a little bit strange.
And I am located in Belgium, not the UK.
The debug output of the QT module (when locking) is:
Jan 25 08:07:13 vector QT1010: compare reg:25 34 b5
Jan 25 08:07:13 vector QT1010: compare reg:25 b5 b5
Jan 25 08:07:13 vector QT1010: compare reg:26 14 14
Jan 25 08:07:13 vector QT1010: compare reg:22 21 21
Jan 25 08:07:13 vector QT1010: compare reg:1f 05 0b
Jan 25 08:07:13 vector QT1010: compare reg:1f 0b 11
Jan 25 08:07:13 vector QT1010: compare reg:1f 11 14
Jan 25 08:07:13 vector QT1010: compare reg:1f 14 1e
Jan 25 08:07:13 vector QT1010: compare reg:1f 1e 24
Jan 25 08:07:13 vector QT1010: compare reg:1f 24 2b
Jan 25 08:07:13 vector QT1010: compare reg:1f 2b 31
Jan 25 08:07:13 vector QT1010: compare reg:1f 31 37
Jan 25 08:07:13 vector QT1010: compare reg:1f 37 3e
Jan 25 08:07:13 vector QT1010: compare reg:1f 3e 5b
Jan 25 08:07:13 vector QT1010: compare reg:1f 5b 5b
Jan 25 08:07:13 vector QT1010: freq:545000000 05:34 07:37 09:1c 0a:0a 
0b:44 1a:b8 11:fd 12:91 22:d7 05:30 1f:13 20:14 25:00 00:92
Jan 25 08:07:13 vector 00: 92 00 00 19 00 30 00 37 08 1c 0a 44 e1 84 b7 2c
Jan 25 08:07:13 vector 10: 10 fd 91 3f 7f 1f ff ff ff 80 b8 00 89 00 00 53
Jan 25 08:07:13 vector 20: 14 53 f7 ee f3 00 15 29 55 39 23 70 dc 20 00 00
Jan 25 08:07:14 vector QT1010: freq:482000000 05:34 07:36 09:1d 0a:09 
0b:45 1a:78 11:fd 12:91 22:d4 05:30 1f:0e 20:0f 25:00 00:92
Jan 25 08:07:14 vector 00: 92 00 00 19 00 30 00 36 08 1d 09 45 e1 84 b7 2c
Jan 25 08:07:14 vector 10: 10 fd 91 3f 7f 1f ff ff ff 80 78 00 89 00 00 4e
Jan 25 08:07:14 vector 20: 0f 53 f4 f5 f4 00 15 29 55 39 23 70 dc 20 00 00

Removing the line I added to zl10353.c and tuning to 492600000 doesn't work.
The only type of signal that I get then is:
FE_STATUS: FE_HAS_VITERBI

tuning to 482000000 then gives the following debug output for the tuner:
Jan 25 08:19:06 vector QT1010: freq:482000000 05:34 07:36 09:1d 0a:09 
0b:45 1a:78 11:fd 12:91 22:d4 05:30 1f:0e 20:0f 25:00 00:92
Jan 25 08:19:06 vector 00: 92 00 00 19 00 30 00 36 08 1d 09 45 e1 84 b7 2c
Jan 25 08:19:06 vector 10: 10 fd 91 3f 7f 1f ff ff ff 80 78 00 89 00 00 4e
Jan 25 08:19:06 vector 20: 0f 53 f4 f5 f4 00 15 29 55 39 23 70 dc 20 00 00

This looks the same for me as without the zl10353 line added.
Has anybody got any ideas what this register 0x5A's function could be?

Thanks,
Jan




More information about the linux-dvb mailing list