Steve,
Copying the ivtv-devel list (as a rule, I like to keep development
discussions in the open...)
On Mon, 2009-06-08 at 17:59 +0100, Steve Firth wrote:
> Hi
> Thanks for the answer.
>
> I'm not an expert on this area, so not sure how relevant this is,
> but the I have had analog TV working (in UK), admittedly not in
> MythTV, but manually using scan etc for channel setting and then
> "cat /dev/video0 | xine/mplayer" etc So something somewhere is
> connected correctly for analog TV. I'm hoping that this may mean that
> we are already at stage "2" of your list?
Well, we can get the firmware loaded once without knowing the right
reset pin. If we don't have the right reset pin subsequent firmware
loads fail.
The xc3028 seems to need different firmware segments loaded when things
change. Adding
options tuner debug=1
to /etc/modprobe.conf (or your distro's equivalent) will log a little
bit more of what's going on with xc3028 firmware loads during normal
operations (I think).
>
> FYI My installation (Ubuntu 9.04) is fairly standard, I've just
>
> * Upped VMALLOC in boot to get the 2nd card working (VERY memory
> hungry)
It's not real memory. It's a few page table entries to get a 64 MB
address window per CX23418 to map in memory and registers over the PCI
bus.
> * Copied xc3028-v27.fw to the firmware directory (Suggesting that
> your guess is correct)
>
> I could not get tuners then working in mythtv, but as mentioned
> above, I did get analog tv working from the command line.
This could be from having the wrong reset line. Check the kernel logs
(dmesg and /var/log/messages) when you try to tune to different stations
and see if there are any xc2028 firmware gripes.
> The MPC718 card only has an simple RF shield soldered above it, so I
> could remove this to take a peek at the chips inside. Not diffficult,
> as there are only 3 fixing points but it may take me a day or so to
> find out my soldering iron and earthed bench equipment as an ESD
> precaution. There are a couple of Xtals on the underside, one looks
> like 32MHz, the second is impossible to read.
Cool. Just take your time - we can always work with guesses until you
have time to get the cover off. High-res picutres of both sides of the
card with the RF shields off would be nice; but a list of chips might
do.
> There is a photo of the card on the web (google images) where there
> appears to be another Xtal under the RF shield: I'll try to check this
> out.
Yuan has some fuzzed out photos. Not very useful.
> Can I work out the I2C address via Windows XP? It must be useful for
> something ;-)
It might be listed in the *inf file. If not it's hardcoded into Windows
driver - which requires diassembly of the Windows driver - not what I'd
want to do. Using the i2c-dev module and i2c-detect would be infinitely
easier.
For reference, here's i2c-detect output from my HVR-1600
# modprobe i2c-dev
# i2cdetect -l
i2c-1 smbus SMBus PIIX4 adapter at 0b00 SMBus adapter
i2c-0 i2c cx18 i2c driver #0-0 I2C adapter
i2c-2 i2c cx18 i2c driver #0-1 I2C adapter
# i2cdetect -y 0
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- 19 -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- UU -- -- --
50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- 63 -- -- -- -- -- -- -- -- -- -- -- --
70: 70 71 72 73 -- -- -- --
19 = cx25447
4c = cs5345
UU = device has been claimed by a driver
50 = ATMEL eeprom
63 = mxl5005s
70-73 = Z8F0811 IR microcontroller
# i2cdetect -y 2
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- UU -- -- -- -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --
61 = simple tuner mixer/oscillator chip
UU = device has been claimed by a driver
> There's quite a few of these boxes out there, if I can get it working
> I'd be happy to script-up my steps above for anyone else who would
> like to use a decent OS/Media Centre . I'm absolutely fed up with
> trying to use XP-MCE!
OK, give me a few days I'll do modifications closer to the weekend.
Regards,
Andy
>
> Regards
> Steve Firth
>
> ----- Original message -----
> Sent: 2009/06/07 21:37:33
> Subject: Re: Re: [ivtv-users] Acer aspire idea 500?
>
> On Sun, 2009-06-07 at 19:39 +0100, Steve Firth wrote:
> > Hi, I was following with interest the email discussion about
> getting
> > the cx18 drivers etc working with the acer aspire idea 510 (I have
> the
> > 500 model)
> > but the chat seemed to fizzle in june 2008. Did anyone ever get
> this
> > working?
>
>
> It fizzled. There's an entry in the cx18 driver for the Yuan MPC718,
> but it is incomplete. Here's what I think I can do:
>
> 1. Fix the analog basedband part (CVBS, S-Video) of the card based on
> the windows *.inf file.
>
> 2. Get analog tuning working based on knowing the card has an XCeive
> XC3028 analog/digital tuner. For this I need someone to figure out
> for
> certain what GPIO pin is used to reset the XC3028. This will require
> someone with patience as the brute force search may require up to 32
> loads/unloads/recompiles of the driver.
>
> 3. I might be able to get DVB-T working, but will need to rely
> heavily
> on someone with the hardware and access to DVB-T. Here's how things
> will go:
>
> a. Someone really needs to figure out what DVB-T demodulator is on
> the
> MPC718. It's housed in the metal can. :( The Yuan PG718 looks like a
> PCI version of the MPC718, so maybe if someone can provide a list of
> the
> chips on that card, it'll be the same. The Yuan MPC788 is reported to
> use an XCeive XC3028 with an Intel/Zarlink CE6353/ZL10353 - so I can
> use
> that as a first guess.
>
> b. I need to know the crystal freq used with the demod. If it's the
> Zarlink 10353 or equivalent, I'm betting it's a 20.48 MHz crystal.
> Visual confirmation would be nice.
>
> c. I need to know the I2C address of the digital demodulator. If it's
> the Zarlink 10353 or equivalent, I'm guessing it's at 0x1e as most
> Zl10353 eval kits used this apparently. If not, there's a way to use
> the i2c-dev module and the i2c-detect tool (lm-sensors package) to
> try
> and figure it out.
>
> d. I need to know if the MPC718 uses a Parallel or Serial TS. This
> will
> require module loads/unloads/recompiles of the driver. Hopefully this
> will be an experiment that has to be run only once or twice.
>
>
> So,
>
> 1. How bad do you want it to work?
> 2. Do you have a technical contact at Yuan?
>
> ;)
>
> Regards,
> Andy
>
> > Cheers
> > spf
>
>
>
> _______________________________________________
> ivtv-users mailing list
> ivtv-users@ivtvdriver.org
> http://ivtvdriver.org/mailman/listinfo/ivtv-users
_______________________________________________
ivtv-devel mailing list
ivtv-devel@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
Copying the ivtv-devel list (as a rule, I like to keep development
discussions in the open...)
On Mon, 2009-06-08 at 17:59 +0100, Steve Firth wrote:
> Hi
> Thanks for the answer.
>
> I'm not an expert on this area, so not sure how relevant this is,
> but the I have had analog TV working (in UK), admittedly not in
> MythTV, but manually using scan etc for channel setting and then
> "cat /dev/video0 | xine/mplayer" etc So something somewhere is
> connected correctly for analog TV. I'm hoping that this may mean that
> we are already at stage "2" of your list?
Well, we can get the firmware loaded once without knowing the right
reset pin. If we don't have the right reset pin subsequent firmware
loads fail.
The xc3028 seems to need different firmware segments loaded when things
change. Adding
options tuner debug=1
to /etc/modprobe.conf (or your distro's equivalent) will log a little
bit more of what's going on with xc3028 firmware loads during normal
operations (I think).
>
> FYI My installation (Ubuntu 9.04) is fairly standard, I've just
>
> * Upped VMALLOC in boot to get the 2nd card working (VERY memory
> hungry)
It's not real memory. It's a few page table entries to get a 64 MB
address window per CX23418 to map in memory and registers over the PCI
bus.
> * Copied xc3028-v27.fw to the firmware directory (Suggesting that
> your guess is correct)
>
> I could not get tuners then working in mythtv, but as mentioned
> above, I did get analog tv working from the command line.
This could be from having the wrong reset line. Check the kernel logs
(dmesg and /var/log/messages) when you try to tune to different stations
and see if there are any xc2028 firmware gripes.
> The MPC718 card only has an simple RF shield soldered above it, so I
> could remove this to take a peek at the chips inside. Not diffficult,
> as there are only 3 fixing points but it may take me a day or so to
> find out my soldering iron and earthed bench equipment as an ESD
> precaution. There are a couple of Xtals on the underside, one looks
> like 32MHz, the second is impossible to read.
Cool. Just take your time - we can always work with guesses until you
have time to get the cover off. High-res picutres of both sides of the
card with the RF shields off would be nice; but a list of chips might
do.
> There is a photo of the card on the web (google images) where there
> appears to be another Xtal under the RF shield: I'll try to check this
> out.
Yuan has some fuzzed out photos. Not very useful.
> Can I work out the I2C address via Windows XP? It must be useful for
> something ;-)
It might be listed in the *inf file. If not it's hardcoded into Windows
driver - which requires diassembly of the Windows driver - not what I'd
want to do. Using the i2c-dev module and i2c-detect would be infinitely
easier.
For reference, here's i2c-detect output from my HVR-1600
# modprobe i2c-dev
# i2cdetect -l
i2c-1 smbus SMBus PIIX4 adapter at 0b00 SMBus adapter
i2c-0 i2c cx18 i2c driver #0-0 I2C adapter
i2c-2 i2c cx18 i2c driver #0-1 I2C adapter
# i2cdetect -y 0
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- 19 -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- UU -- -- --
50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- 63 -- -- -- -- -- -- -- -- -- -- -- --
70: 70 71 72 73 -- -- -- --
19 = cx25447
4c = cs5345
UU = device has been claimed by a driver
50 = ATMEL eeprom
63 = mxl5005s
70-73 = Z8F0811 IR microcontroller
# i2cdetect -y 2
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- UU -- -- -- -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --
61 = simple tuner mixer/oscillator chip
UU = device has been claimed by a driver
> There's quite a few of these boxes out there, if I can get it working
> I'd be happy to script-up my steps above for anyone else who would
> like to use a decent OS/Media Centre . I'm absolutely fed up with
> trying to use XP-MCE!
OK, give me a few days I'll do modifications closer to the weekend.
Regards,
Andy
>
> Regards
> Steve Firth
>
> ----- Original message -----
> Sent: 2009/06/07 21:37:33
> Subject: Re: Re: [ivtv-users] Acer aspire idea 500?
>
> On Sun, 2009-06-07 at 19:39 +0100, Steve Firth wrote:
> > Hi, I was following with interest the email discussion about
> getting
> > the cx18 drivers etc working with the acer aspire idea 510 (I have
> the
> > 500 model)
> > but the chat seemed to fizzle in june 2008. Did anyone ever get
> this
> > working?
>
>
> It fizzled. There's an entry in the cx18 driver for the Yuan MPC718,
> but it is incomplete. Here's what I think I can do:
>
> 1. Fix the analog basedband part (CVBS, S-Video) of the card based on
> the windows *.inf file.
>
> 2. Get analog tuning working based on knowing the card has an XCeive
> XC3028 analog/digital tuner. For this I need someone to figure out
> for
> certain what GPIO pin is used to reset the XC3028. This will require
> someone with patience as the brute force search may require up to 32
> loads/unloads/recompiles of the driver.
>
> 3. I might be able to get DVB-T working, but will need to rely
> heavily
> on someone with the hardware and access to DVB-T. Here's how things
> will go:
>
> a. Someone really needs to figure out what DVB-T demodulator is on
> the
> MPC718. It's housed in the metal can. :( The Yuan PG718 looks like a
> PCI version of the MPC718, so maybe if someone can provide a list of
> the
> chips on that card, it'll be the same. The Yuan MPC788 is reported to
> use an XCeive XC3028 with an Intel/Zarlink CE6353/ZL10353 - so I can
> use
> that as a first guess.
>
> b. I need to know the crystal freq used with the demod. If it's the
> Zarlink 10353 or equivalent, I'm betting it's a 20.48 MHz crystal.
> Visual confirmation would be nice.
>
> c. I need to know the I2C address of the digital demodulator. If it's
> the Zarlink 10353 or equivalent, I'm guessing it's at 0x1e as most
> Zl10353 eval kits used this apparently. If not, there's a way to use
> the i2c-dev module and the i2c-detect tool (lm-sensors package) to
> try
> and figure it out.
>
> d. I need to know if the MPC718 uses a Parallel or Serial TS. This
> will
> require module loads/unloads/recompiles of the driver. Hopefully this
> will be an experiment that has to be run only once or twice.
>
>
> So,
>
> 1. How bad do you want it to work?
> 2. Do you have a technical contact at Yuan?
>
> ;)
>
> Regards,
> Andy
>
> > Cheers
> > spf
>
>
>
> _______________________________________________
> ivtv-users mailing list
> ivtv-users@ivtvdriver.org
> http://ivtvdriver.org/mailman/listinfo/ivtv-users
_______________________________________________
ivtv-devel mailing list
ivtv-devel@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-devel