[linux-dvb] Pinnacle PCTV Sat HDTV Pro USB (PCTV452e)

garlicdevel at ntlworld.com garlicdevel at ntlworld.com
Wed Oct 24 01:42:20 CEST 2007

On Sunday 21 October 2007 18:00, Dominik Kuhlen wrote:
> Hi, 
> On Thursday 18 October 2007 09:54, Daniel Golle wrote:
> > hi all!
> > 
> > i finally got the pinnacle device mentioned above, willing to
> > contribute whatever i can to make it work under linux. As I have to
> > study USB programming as a part of lab class, I hope to combine
> > pratical and educative interests... :)
> > 
> > please send me all known information about the device and let me know
> > about exiting code (Dominik? Roland? Patrick? anybody?)
> Here it is: 
> A new try for a driver for the pctv452e usb device.
> What's working:
>  - Remote control
>  - LNB control (14/18v, 22kHz)
>  - DiSEqC
> What's not working:
>  - tuning (neither DVB-S nor DVB-S2) :(
>  - reading data (which probably results from failing to tune)

I've been working on one of these myself
but you've beat me to it :)
I was in the middle up updating the init registers on the Sunday when I spotted this mail on the list

still I might have some useful info / input
to begin with I started out by writing a perl script to decode the output from usbsnoop for the 452e while using the win driver
(it's the first time I've written something in perl so it's a bit crude)
I've put it up here http://homepage.ntlworld.com/garlicbread/Projects/452e/USBsnoop_parse.tar.bz2
you just need to unpack / copy the usbsnoop.log into the extracted directory and run "./452e.1.pl | less"
this should give a nice colorful output on the console as to what's going on
but be aware it doesn't always name the stb0899 registers properly because it doesn't yet take account of offsets
still the registers are set in the same order they're seen in the data structures so it's easy to figure out what's being set
note I wouldn't use it with any logs over 100Mb otherwise it'll eat your memory :)

for the results I've had with the mentioned patch above
with the dish pointed at Astra-28.2E
using a patched version of szap
or mplayer dvb://BBC1
with kernel 2.6.23

I've got a motor and a meter hooked in line with the dish so I can check for power on the line
without the right tuning info the power to the LNB flicks on for half a second at a time repeatedly until stopped
with the correct tuning info (I'm using "BBC1:10773:v:0:22000:5000:5002:6301" in channels.conf)
the power stays on and is constant with the correct polarity setting

this suggests it is tuning
but there's no visible power output reading from szap, or mpeg output from mplayer
so I'm not sure what needs looking at next

Next I've noticed that you've used the old stb0899 repo hg clone http://linuxtv.org/hg/~manu/stb0899
For info I've spotted a newer repo over here http://jusst.de/hg/multiproto/ (spotted in one of manu's mails about multiproto)
but I think the older driver could be better at the moment as with the newer driver the device doesn't appear to initialize for the stb0899

also I hope you don't mind
but I took some of the changes and combined it with the driver I've been working with

1. lnbp22.patch -  this is from the patch above, it's better than the one I'd setup so I've left this as is
2. pctv452e_initreg.patch - separate header file containing the registers to be sent to the device on init
I think this is more up to date because it's based off the usbsnoop logs with the win driver
but I need to give it one more final sanity check to be sure
3. stb0899_tstrs3.patch - minor change to one of the registers I've spotted while trailing through the usb logs
4. stb0899_misc.patch - changes you've made to the stb0899 driver (originally part of the above patch)

3. pctv452e_main.patch - a combination of the above driver and some code I've been working with
it's basically a slightly changed version of the more recent ttusb driver
with bits of code from the above patch (remote control especially, this does work with evtest)
the newer budget driver in manu's tree (another func added for the stb6100) with some other bits

at this point the results are the same (meter lights up but not much else)
so I can't say if this is better or worse at this point
although I've noticed the LED is staying at orange when it should be going green in both cases, which suggests the GPIO register isn't being set when it should

quick question has anyone managed to get any kind of visible video output from the device at this point?
I'm wondering if the mpeg stream settings (for the iso transfer) need tweaking (I'm not sure if the endpoint or some other setting is set correct)

Email sent from www.virginmedia.com/email
Virus-checked using McAfee(R) Software and scanned for spam

More information about the linux-dvb mailing list