Re: [linux-dvb] Pinnacle PCTV HD pro USB stick 801e

[Matthias Schwarzott <zzam@gentoo.org>]

On Samstag, 3. Mai 2008, Jernej Tonejc wrote:
> On Fri, 11 Apr 2008, Steven Toth wrote:
> > Hi Jernej,
> >
> > The s5h1409 is a different beast to the s5h1411, so you're wasting your
> > time trying to make that work.
> >
> > That being said, I'm kinda surprised you're having i2c scan issues. I
> > don't work with the dibcom src so maybe that's a true limitation of the
> > part, or maybe something else is just plain broken on your design.
> >
> > Googling/searching the mailing list, or reading the wiki's at linuxtv.org
> > might show a reason why I2C scanning isn't supported.
> >
> > In terms og the GPIO's, you'll need to understand which GPIO the xc5000
> > tuner is attached to (because the xc5000 needs to toggle this). You might
> > also need to drive other gpio's to bring the tuner, demod or any other
> > parts out of reset - and able to respond to i2c commands. I tend to add
> > /* comments */ around the GPIO code for each product detailing any gpio's
> > I know (or suspect), which helps other devs maintain the code later in
> > it's life.
> >
> > Maybe you could make some progress with understanding why I2C scanning
> > doesn't work, and perhaps dig deeping and try to establish which gpio's
> > are connected to what.
> >
> > With these two things, and a s5h1411 driver we should be able to get
> > support for this product pretty easily.
>
> Hi Steve,
>
> I managed to get the s5h1411 frontend attached for the Pinnacle HDTV PRO
> 801e USB stick. I scanned through all i2c ports (in the code, command-line
> utils like i2cdump, i2cdetect don't work for some reason) and I got
> response on addresses 0x19, 0x1a, 0x44, 0x50. The first two are almost
> certainly related to the s5h1411 frontend. I suspect that xc5000 is
> sitting on 0x50. The problem however is that the reply to the
> XREG_PRODUCT_ID command is not one of the two replies that you have in
> your xc5000.c code. Originally it returned 0x423. After adding that as one
> of the replies (both for loading the fw and for not loading it), I tried
> to tune to some frequency but the tuning failed. Also, the USB device ID
> changed the next time I plugged it in, it's 10b8:0066 now (it was
> 2304:0231 before). I have no idea what caused this. The device is still

First: I having absolutely no clue about this device.
But I do on i2c: I guess the device you found at 0x50 is an eeprom. So you may 
have overwritten the subvendor/subdevice id there.

> recognized (after I updated the IDs in the code) and the frontend gets
> attached and seems to work - the device accepts all commands for the
> s5h1411 part. But strangely, everytime I try to tune, the reply on 0x50
> i2c port changes. It only changes between unplugging and plugging in the
> card, not between repeated attempts to tune while the stick stays in. The
> few different replies I got so far are: 0x423 (original, with the
> original USB IDs, it never reappeared), 0x34d1, 0x14d1, 0x0f51, 0x0dd1.

As you wrote: Originally it returned 0x423 - that is just endian swapped the 
original vendor id of 0x2304.

Matthias

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb