tm6000

tm6000

[Dmitri Belimov]

Hi

Who is maintainer of the tm6000 module??

With my best regards, Dmitry.

Re: tm6000

[Mauro Carvalho Chehab]

Dmitri Belimov wrote:
> Hi
> 
> Who is maintainer of the tm6000 module??
> 
> With my best regards, Dmitry.
I wrote it, although I'm currently lacking time to fix the bugs.
Feel free to fix if you want. I have some boards here and I can
help testing on them.

Cheers,
Mauro.

tm6000

[Stefan Ringel]

Hi Mauro,

First, I have a question, in the function "int tm6000_cards_setup(struct
tm6000_core *dev)" have you reset gpio 1 and 4, but why? In my lastest
patch I have add a few sticks and it's works with and without this
part.What for sticks use this part? For a week I have become a email
with usbsnoop log from a tm5600 based stick and I have analysed it.
Resume is that I think it use two reset gpio's.

1. GPIO1  = 0
2. GPIO4 = 0
3. wait a few ms
4. GPIO1 = 1
5. GPIO4 = 1

/*
     * Default initialization. Most of the devices seem to use GPIO1
     * and GPIO4.on the same way, so, this handles the common sequence
     * used by most devices.
     * If a device uses a different sequence or different GPIO pins for
     * reset, just add the code at the board-specific part
     */
    for (i = 0; i < 2; i++) {
        rc = tm6000_set_reg(dev, REQ_03_SET_GET_MCU_PIN,
                    dev->gpio.tuner_reset, 0x00);
        if (rc < 0) {
            printk(KERN_ERR "Error %i doing GPIO1 reset\n", rc);
            return rc;
        }

        msleep(10); /* Just to be conservative */
        rc = tm6000_set_reg(dev, REQ_03_SET_GET_MCU_PIN,
                    dev->gpio.tuner_reset, 0x01);
        if (rc < 0) {
            printk(KERN_ERR "Error %i doing GPIO1 reset\n", rc);
            return rc;
        }

        msleep(10);
        rc = tm6000_set_reg(dev, REQ_03_SET_GET_MCU_PIN, TM6000_GPIO_4, 0);
        if (rc < 0) {
            printk(KERN_ERR "Error %i doing GPIO4 reset\n", rc);
            return rc;
        }

        msleep(10);
        rc = tm6000_set_reg(dev, REQ_03_SET_GET_MCU_PIN, TM6000_GPIO_4, 1);
        if (rc < 0) {
            printk(KERN_ERR "Error %i doing GPIO4 reset\n", rc);
            return rc;
        }

        if (!i) {
            rc = tm6000_get_reg16(dev, 0x40, 0, 0);
            if (rc >= 0)
                printk(KERN_DEBUG "board=%d\n", rc);
        }
    }

For a week I have become a email with usbsnoop log from a tm5600 based
stick and I have analysed it. Resume is that I think it use two reset
gpio's.

1. GPIO1  = 0
2. GPIO4 = 0
3. wait a few ms
4. GPIO1 = 1
5. GPIO4 = 1

So I think, if it a part from function tm6000_cards_setup is, then is
that part wrong and must remove to tuner_callback function.

Second, I will rewrite the gpio's into a struct.

for example:

in tm6000.h

struct tm6000_gpio {
    int        tuner_reset;
    int        tuner_on;
    int        demod_reset;
    int        demod_on;
    int        power_led;
    int        dvb_led;
    int        ir;
};

in tm6000_card.c

[TM6010_BOARD_TERRATEC_CINERGY_HYBRID_XE] = {
        .name         = "Terratec Cinergy Hybrid XE / Cinergy Hybrid-Stick",
        .tuner_type   = TUNER_XC2028, /* has a XC3028 */
        .tuner_addr   = 0xc2 >> 1,
        .demod_addr   = 0x1e >> 1,
        .type         = TM6010,
        .caps = {
            .has_tuner    = 1,
            .has_dvb    = 1,
            .has_zl10353    = 1,
            .has_eeprom    = 1,
            .has_remote    = 1,
        },
        .gpio = {
            .tuner_reset     = TM6010_GPIO_2,
            .tuner_on    = TM6010_GPIO_3,
            .demod_reset    = TM6010_GPIO_1,
            .demod_on    = TM6010_GPIO_4,
            .power_led    = TM6010_GPIO_7,
            .dvb_led    = TM6010_GPIO_5,
            .ir        = TM6010_GPIO_0,
        },
    },

The rest I send in the patch email. So can the gpio's define in the
board struct and using per label in the functions.

Stefan Ringel

-- 
Stefan Ringel <stefan.ringel@arcor.de>

tm6000

[Dmitri Belimov]

Hi

Anyone can watch TV with tm6000 module??

I try mplayer. My PC crashed inside copy_streams function after some time.

With my best regards, Dmitry.

Re: tm6000

[Bee Hock Goh]

its working fine on my tm5600. Run fine for more than 5mins without crashing.

But it does crashed everytime when its playing on the first channel
and I switch channel in mythtv. I am waiting for usb/serial cable to
arrive from ebay to find out why.

I hope to work on the audio part though.

On Wed, Apr 28, 2010 at 1:49 PM, Dmitri Belimov <d.belimov@gmail.com> wrote:
> Hi
>
> Anyone can watch TV with tm6000 module??
>
> I try mplayer. My PC crashed inside copy_streams function after some time.
>
> With my best regards, Dmitry.
>

Re: tm6000

[Mauro Carvalho Chehab]

Dmitri Belimov wrote:
> Hi
> 
> Anyone can watch TV with tm6000 module??
> 
> I try mplayer. My PC crashed inside copy_streams function after some time.
> 
> With my best regards, Dmitry.

It is still causing panic on my tests. Probably, the routine is still writing
something out of the buffer area.

-- 

Cheers,
Mauro

tm6000

[Dmitri Belimov]

Hi 

At this moment I can start mplayer and see green field with some junk without crash.
Info from mplayer - received 497 frames, drop 69 incorrect frames.

Compile without support DVB-T for our cards.

Now try understand init process working drivers and diff with linux.

P.S. Linux kernel is 2.6.33

With my best regards, Dmitry.

Re: tm6000

[Bee Hock Goh]

Did you comment off the code in the get_next_buff that clear the
previous frame data?

On Wed, May 5, 2010 at 12:50 PM, Dmitri Belimov <d.belimov@gmail.com> wrote:
> Hi
>
> At this moment I can start mplayer and see green field with some junk without crash.
> Info from mplayer - received 497 frames, drop 69 incorrect frames.
>
> Compile without support DVB-T for our cards.
>
> Now try understand init process working drivers and diff with linux.
>
> P.S. Linux kernel is 2.6.33
>
> With my best regards, Dmitry.
>

Re: tm6000

[Dmitri Belimov]

On Wed, 5 May 2010 13:44:38 +0800
Bee Hock Goh <beehock@gmail.com> wrote:

> Did you comment off the code in the get_next_buff that clear the
> previous frame data?

No. 

Git tree + my last patch.

With my best regards, Dmitry.

> On Wed, May 5, 2010 at 12:50 PM, Dmitri Belimov <d.belimov@gmail.com>
> wrote:
> > Hi
> >
> > At this moment I can start mplayer and see green field with some
> > junk without crash. Info from mplayer - received 497 frames, drop
> > 69 incorrect frames.
> >
> > Compile without support DVB-T for our cards.
> >
> > Now try understand init process working drivers and diff with linux.
> >
> > P.S. Linux kernel is 2.6.33
> >
> > With my best regards, Dmitry.
> >

Re: tm6000

[Bee Hock Goh]

If you do that you will get some decent looking video.

On Wed, May 5, 2010 at 3:27 PM, Dmitri Belimov <d.belimov@gmail.com> wrote:
> On Wed, 5 May 2010 13:44:38 +0800
> Bee Hock Goh <beehock@gmail.com> wrote:
>
>> Did you comment off the code in the get_next_buff that clear the
>> previous frame data?
>
> No.
>
> Git tree + my last patch.
>
> With my best regards, Dmitry.
>
>> On Wed, May 5, 2010 at 12:50 PM, Dmitri Belimov <d.belimov@gmail.com>
>> wrote:
>> > Hi
>> >
>> > At this moment I can start mplayer and see green field with some
>> > junk without crash. Info from mplayer - received 497 frames, drop
>> > 69 incorrect frames.
>> >
>> > Compile without support DVB-T for our cards.
>> >
>> > Now try understand init process working drivers and diff with linux.
>> >
>> > P.S. Linux kernel is 2.6.33
>> >
>> > With my best regards, Dmitry.
>> >
>

Re: tm6000

[Mauro Carvalho Chehab]

Dmitri Belimov wrote:
> On Wed, 5 May 2010 13:44:38 +0800
> Bee Hock Goh <beehock@gmail.com> wrote:
> 
>> Did you comment off the code in the get_next_buff that clear the
>> previous frame data?
> 
> No. 
> 
> Git tree + my last patch.
> 

A "green" tree can happen due to lots of conditions, like:
	1) it is not receiving data from xc3028 (or xc5000);
	2) wrong gpio setup;
	3) data sent too fast to tm6000;
	4) need to add some new workarounds to another tm6000 firmware/hardware bug;
	5) the device stopped answer and got disconnected from USB buffer;
	6) signal were too weak after changing to some channel (it seems that the tm6000 
	   chip stops reception with weak signals - I remember I had to add a code that 
	   re-enables xc3028 every time a channel is changed due to this bug, since,
	   after disabled, even if the signal become strong, it keeps showing a green
	   screen);
	7) you hit yet another bug on this device ;)

-- 

Cheers,
Mauro

tm6000

[Dmitri Belimov]

Hi 

tm6000 progress bar :-)

At this moment I can watch TV with mplayer without sound.
Our card based on the xc5000 and tm6010
It is possible with my last patch for xc5000 tuner.

Next task is a sound, check init process and stds settings.

P.S. Compile without support DVB-T frontend for our cards. Linux kernel is 2.6.33

With my best regards, Dmitry.