Reply by glen herrmannsfeldt October 28, 20042004-10-28

Andre wrote:
(I wrote)

>> It is useful for digitally generated color signals, like the >> IBM CGA card for the original IBM PC, to have a sample rate >> a convenient multiple of the color subcarrier, but it is less >> obvious with YCbCr outputs. Note that all ISA bus machines must >> have a 14.31818MHz crystal, as there is an ISA bus pin defined to >> have that frequency, as used by the CGA card.
> ... which is 3 times 4.77 MHz, the clock of the original IBM PC...
The 8088 uses a clock with 33% duty cycle. The 8284 clock generator has a divide by three to generate it. The frequency was selected so that the CGA could use it. Starting with 80286 machines the 14.31818MHz oscillator is needed to supply that pin on the bus, just in case one has a CGA card. -- glen
Reply by Jerry Avins October 28, 20042004-10-28
Steve wrote:
> Very good, you really read that. I appreciate your reply. > > I noticed that part too. But one thought caused me to my original post > is that: The table 2-2, for my understanding, should refer to the > sampling rate for the input signal. Table 2-1 is indicating the output > format. According to you, the data rate for what ever output format > (10bit or 20bit) would have to the same rate relating the sampling > rate on Table 2-2. Are we on the same track now? I think I will stop > here to make sure I am right so that the next question mark is valid. > > glen herrmannsfeldt <gah@ugcs.caltech.edu> wrote in message news:<clp14u$kac$1@gnus01.u.washington.edu>... > >>Steve wrote: >> >> >> >>>I need help in understanding a group of video decoder chip >>>specifications. Let me use TI's TVP5146 as an example. The first thing >>>I want to know is the output digital data rate. On the spec, it says: >> >> >> >>>Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...; >>>Single 14.31818MHz reference crystal for all standards. >> >>The sample rate is on page 2-11, and depends on the TV standard >>being used. 13.5MHz for NTSC, 12.2727MHz or 14.75MHz for others. >>It seems there is a PLL so that it can generate frequencies that >>are multiples of 14.31818MHz*2**-23. (I think that is the way >>the PLL works. It does have a 2**-23 in it, anyway.) >> >>-- glen
To have the input and output rates a bit different -- not 2:1 or such -- a decoder chip would need to be made of rubber, not silicon. Maybe someday .... Jerry -- Engineering is the art of making what you want from things you can get. &#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;
Reply by Allan Herriman October 28, 20042004-10-28
On Thu, 28 Oct 2004 14:33:42 +0200, Andre <no_spam@fischer-zoth.de>
wrote:

> >glen herrmannsfeldt wrote: > >> Jerry Avins wrote: >> >>> Steve wrote: >> >>> 14.31818MHz is exactly 4 times the color subcarrier frequency. That >>> ought to tell us something. Given that the carrier is modulated with two >>> signals in quadrature, it's probably the sample rate, in and out. >> >> It is useful for digitally generated color signals, like the >> IBM CGA card for the original IBM PC, to have a sample rate >> a convenient multiple of the color subcarrier, but it is less >> obvious with YCbCr outputs. Note that all ISA bus machines must >> have a 14.31818MHz crystal, as there is an ISA bus pin defined to >> have that frequency, as used by the CGA card. > >... which is 3 times 4.77 MHz, the clock of the original IBM PC...
It's the other way around. They wanted to save money, so they used a cheap colourburst (uh, that'd be colorburst if it's NTSC) crystal, and the Intel 8284 clock generator divided the crystal frequency by 3 ('cause the nmos 8086 and 8088 needed a 33% duty cycle), to give a 4.77MHz processor clock. Regards, Allan
Reply by Andre October 28, 20042004-10-28
glen herrmannsfeldt wrote:

> Jerry Avins wrote: > >> Steve wrote: > >> 14.31818MHz is exactly 4 times the color subcarrier frequency. That >> ought to tell us something. Given that the carrier is modulated with two >> signals in quadrature, it's probably the sample rate, in and out. > > It is useful for digitally generated color signals, like the > IBM CGA card for the original IBM PC, to have a sample rate > a convenient multiple of the color subcarrier, but it is less > obvious with YCbCr outputs. Note that all ISA bus machines must > have a 14.31818MHz crystal, as there is an ISA bus pin defined to > have that frequency, as used by the CGA card.
... which is 3 times 4.77 MHz, the clock of the original IBM PC... -- Please change no_spam to a.lodwig when replying via email!
Reply by Steve October 28, 20042004-10-28
Very good, you really read that. I appreciate your reply.

I noticed that part too. But one thought caused me to my original post
is that: The table 2-2, for my understanding, should refer to the
sampling rate for the input signal. Table 2-1 is indicating the output
format. According to you, the data rate for what ever output format
(10bit or 20bit) would have to the same rate relating the sampling
rate on Table 2-2. Are we on the same track now? I think I will stop
here to make sure I am right so that the next question mark is valid.

glen herrmannsfeldt <gah@ugcs.caltech.edu> wrote in message news:<clp14u$kac$1@gnus01.u.washington.edu>...
> Steve wrote: > > > > I need help in understanding a group of video decoder chip > > specifications. Let me use TI's TVP5146 as an example. The first thing > > I want to know is the output digital data rate. On the spec, it says: > > > Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...; > > Single 14.31818MHz reference crystal for all standards. > > The sample rate is on page 2-11, and depends on the TV standard > being used. 13.5MHz for NTSC, 12.2727MHz or 14.75MHz for others. > It seems there is a PLL so that it can generate frequencies that > are multiples of 14.31818MHz*2**-23. (I think that is the way > the PLL works. It does have a 2**-23 in it, anyway.) > > -- glen
Reply by glen herrmannsfeldt October 27, 20042004-10-27

Steve wrote:


> I need help in understanding a group of video decoder chip > specifications. Let me use TI's TVP5146 as an example. The first thing > I want to know is the output digital data rate. On the spec, it says:
> Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...; > Single 14.31818MHz reference crystal for all standards.
The sample rate is on page 2-11, and depends on the TV standard being used. 13.5MHz for NTSC, 12.2727MHz or 14.75MHz for others. It seems there is a PLL so that it can generate frequencies that are multiples of 14.31818MHz*2**-23. (I think that is the way the PLL works. It does have a 2**-23 in it, anyway.) -- glen
Reply by glen herrmannsfeldt October 27, 20042004-10-27

Jerry Avins wrote:
> Steve wrote:
>>I need help in understanding a group of video decoder chip >>specifications. Let me use TI's TVP5146 as an example. The first thing >>I want to know is the output digital data rate. On the spec, it says:
>>Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...; >>Single 14.31818MHz reference crystal for all standards.
(snip)
> 14.31818MHz is exactly 4 times the color subcarrier frequency. That > ought to tell us something. Given that the carrier is modulated with two > signals in quadrature, it's probably the sample rate, in and out.
I thought of that, too, but it says for all standards, and 14.31818MHz is only 4 times the NTSC color subcarrier. It is useful for digitally generated color signals, like the IBM CGA card for the original IBM PC, to have a sample rate a convenient multiple of the color subcarrier, but it is less obvious with YCbCr outputs. Note that all ISA bus machines must have a 14.31818MHz crystal, as there is an ISA bus pin defined to have that frequency, as used by the CGA card. It does seem likely, though, that it is the sample rate, but in general it doesn't have to be that rate, especially for non-NTSC singals. -- glen
Reply by Jerry Avins October 27, 20042004-10-27
Steve wrote:

> Hi there, > > I need help in understanding a group of video decoder chip > specifications. Let me use TI's TVP5146 as an example. The first thing > I want to know is the output digital data rate. On the spec, it says: > > Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...; > Single 14.31818MHz reference crystal for all standards. > > But doesn't mention the output data rate. I guess either the > 14.31818MHz implies the output data rate or video output formats > indicate the data rate (it can output YCbCr or VBI data format). But I > need some skilled tech. confirmation. The link is on > http://focus.ti.com/lit/ds/symlink/tvp5146.pdf. > > Since I have to make reservation for the next stage digital process, I > think I need to know what the data output rate is.
14.31818MHz is exactly 4 times the color subcarrier frequency. That ought to tell us something. Given that the carrier is modulated with two signals in quadrature, it's probably the sample rate, in and out. Jerry -- Engineering is the art of making what you want from things you can get. &#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;
Reply by Steve October 27, 20042004-10-27
Hi there,

I need help in understanding a group of video decoder chip
specifications. Let me use TI's TVP5146 as an example. The first thing
I want to know is the output digital data rate. On the spec, it says:

Four 10-bit, 30-MSPS A/D converters with analog preprocessors ...;
Single 14.31818MHz reference crystal for all standards.

But doesn't mention the output data rate. I guess either the
14.31818MHz implies the output data rate or video output formats
indicate the data rate (it can output YCbCr or VBI data format). But I
need some skilled tech. confirmation. The link is on
http://focus.ti.com/lit/ds/symlink/tvp5146.pdf.

Since I have to make reservation for the next stage digital process, I
think I need to know what the data output rate is.