[linux-dvb] Yakumo QuickStick Basic DVB-T question

Tobias Stoeber tobi at to-st.de
Sun Sep 10 17:08:26 CEST 2006


Hi,

Alan Chandler wrote:
> On Sunday 10 September 2006 01:56, Tobias Stoeber wrote:
>> Hi anyone,
>>
>> I have got a recent Yakumo Quicktick Basic DVB-T here. Linux distribution
>> is Debian sarge, kernel is 2.6.17 from kernel.org with recent drivers from
>> mercurial. DVB-T reception is in Germany in region 'Braunschweig'.
>>
> This seems to be the same as the freecom sticks I purchased here in the uk.  
> The vendor/product ids are 14aa:0225 (in cold state, warm state changes them 
> to 14aa:0226).

Yes, thats correct, I have the same usb ids here. Maybe somebody should 
update the hardware page on

http://www.linuxtv.org/wiki/index.php/DVB_USB

to reflect this change of a new revision of Freecom / Yakumo sticks.

>> After that the usual /dev/dvb/adapter0 devices are present.
>>
>> Using 'scan' problems begin, as not all channels are found and tuning
>> errors occur. Using a channel that is recognized gives me the following
>> output, e.g.:
> 
> See the post just above, where I had to input more channels into the scan 
> file.  I am new to this game, so I didn't know what to expect.  Does scan use 
> the channel tuning information it is given as a starting point, because as I 
> said, once it had tried all the channels I gave it, it went off trying some 
> more?

Well I am using the channels definitions from the debian package 
dvb-utils. So far this contains all channel (freqs) used here. As some 
regions here in Germany use both UHF and VHF band (incl. Braunschweig) 
with the mentioned antenna I can not receive the mux an channel 8 (ARD 
bouquet). Normally there are 4 tv channels within a mux on a 
channel/freq. in Berlin there are additional radio channels. (just for 
your info).

The scan starts with the first entry (ch 8, 198 MHz) and fails tuning 
(which is correct with this antenna, see above). then continues with the 
next channel ch 23 / 490 MHz, where it sometime finds the ZDF mux or 
sometimes fails.

But as I can receive all UHF channels/muxes with my STB receiver and 
also using the stick under Windows I am really wondering what's the 
problem with linux.

> I am only a few miles/kilometers from the Crystal Palace transmitter in the 
> UK.  I can look out of my house and see it near the top of a hill across the 
> valley with easy line of sight.  I have a loft aerial, which is split (just 
> by wiring the split) into two coax cables, one of which comes down the back 
> of the house and into a sony tv set.  Picture there is perfect.  The second 
> side comes down the front of the house and is split (again via a simple y 
> splitter) to go to an unused socket in our dinning room, and the other side 
> into my study where all my computers are.
> 
> In the study, it is split three ways, one into a Panasonic PVR, the other two 
> into two of these freecom sticks.
> 
> Nevertheless, despite all the splitting, I would have expected a strong signal 
> (on the other side of our house, my wife has a digitial television in our 
> kitchen driven from an internal aerial).
> 
> However, without any amplification, I get problems with pixellation and sound 
> breakup on the Panasonic PVR, but placing a small aerial amplifier just 
> before the two y splitters that give the three feeds in my study, these go 
> away.  
> 
> 
> Strangely the signal value reported by tzap when I do this changes from about  
> 4c4c when there is no amplifier to 3838 when there is.  IN fact, I can make 
> the signal strength numbers rise, by adding coax segments (ie the little 
> aerial extension leads with a plug/socket on each end) into the run from the 
> aerial.  So despite the documentation saying that signal strength should be 
> above 8000, it almost seems to me that it needs to be below 4000 to work.
> 
> (is this a dvb driver issue inverting a value somewhere?)

Well, first of all the stick seems to report only 8 bit of data for both 
signal strength and snr (because it repeats the output of the first 2 
digits).

If the stick really does report -dBm for signal strength, this would 
mean, that lower values would be better. So that in fact 38(38) is 
better than 4c(4c) ...

But how is snr scaled? Does ffff mean 100% snr? Can anyone say how to 
interpret the snr value on a freecomm/yakumo stick?

In general it is not a goo idea to split the antenna signal (too often) 
without additional amplification. Usually you should also observe, that 
a amplifier adds additional noise to the signal (so snr value should 
change too).

Just fyi, the distance between the tx in Braunschweig and my location is 
about 50 km.

> 
>> No teletext/videotext possible using alevt.
> 
> Not tried this

Maybe you can try? Have a look at alevt (or look at this webpage:
http://www.linuxtv.org/wiki/index.php/Alevt

>> EPG works using klear in a recent version. Using klear 0.4 gives me for
>> instance the following output:
>>
>> AudioPid: 386
>> VideoPid: 385
>> Not multiplexed? 0xbd
>> Not multiplexed? 0xbd
>> Not multiplexed? 0xbd
>> Not multiplexed? 0xbd
>> Not multiplexed? 0xc0
>> Not multiplexed? 0xc0
>> Not multiplexed? 0xbd
>> Not multiplexed? 0xc0
> 
> Not tried this - and I don't understand what you are saying (but maybe I would 
> if I did try it).

I have tries different programs, basically ending with using kaffeine 
and klear. You can find klear on http://www.klear.org

The output above shown that klear tunes to some station and video and 
audio pid is shown. After that some errors occur.So my question is what 
does "Not multiplexed" and the hex value mean??

I am using both klear 0.4.? and 0.5.4. Last version is also able to 
create EPG information (electronic program guide). This works!!!

> I have another problem, and that is, as soon as I plug my second stick in it 
> appears to interfere with the first one.  This seems to be the usb layer 
> reporting protocol errors, I was about to try and sprinkle some printk code 
> in the modules to understand what was happening a bit better, as I wondered 
> if the firmware in the sticks are responding to messages that are not meant 
> for it or something like that.

I don't have a second stick to test, so I can't be helpful in testing it.

Thanks so far. Maybe, Alan, you can try to test klear (EPG) and teletext?

Regards, Tobias



More information about the linux-dvb mailing list