discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] HD-[23]000 RAW?


From: Cory Papenfuss
Subject: Re: [Discuss-gnuradio] HD-[23]000 RAW?
Date: Thu, 18 Nov 2004 07:52:20 -0500 (EST)

Very interesting. I posted awhile back about trying to revitalize the bttv->gnuradio attempt, but didn't hear too much about it. Of course if the pcHDTV card is a more general purpose mixer/ADC board it'll be easier to use. It would be nice to get the cheaper bt878 cards to work in some sense as well.

I'm curious about the videosyncronization being turned off... what does it actually do and how did you find out about it? I tried contacting the chip guys about it, but of course received no response.

Too bad about the holes in the data. It certainly seems like there should be a way around this (some test setup or something). Maybe it would be worthwhile to try to contact the chip guys again about it?

-Cory


*************************************************************************
* Cory Papenfuss                                                        *
* Electrical Engineering candidate Ph.D. graduate student               *
* Virginia Polytechnic Institute and State University                   *
*************************************************************************


On Thu, 18 Nov 2004, Martin wrote:

Hi,
I don't know if this will make it to the mailinglist because I am not subscribed.
I am working on code that makes bttv cards usable for gnuradio.
A lot of the restrictions the bttv card has to be usable as a general purpose ADC can to some extend be solved in software. The restrictions most tuner modules on bttv cards have need hardware modifications ore are just not usable. It would be cool if there are indeed cards (like HDTV cards or sattelite receiver cards) which are more generic.

My progress with bttv until now.

video ADC usable in gnuradio
samplerate settable from 28 Mhz to about 80 MHz (maybe higher I didn't dare set it higher then 80 Mhz because it is supposed to run at max 35.468950 MHz)
resolution 8 Bits

videosynchronisation can be turned off (by setting the undocumented test register to 239)

I use the RAW capture mode (meant for teletext VBI capturing) with a modified bttv driver to get more datalines.
This way I get around the horizontal videotimings.
I can get a continuous stream of sampledata for about 19.4 msec and then a gap of 0.6 msec then 19.4 msec data 0.6 msec gap and so on. This has to do with the fact that the bt878 invalidates samples during the vertical sync pulse (which it even generates internally even if it is not trying to synchronise with the videostream)

If I just leave out the gaps in the datastream I can use gnuradio to demodulate FM stations and listen to them. Because of the gap there is some distortion (garbling) but usable for listening to speech.

I have not found any way around this gap.

The tuner hase more problems.
It tries to lock to the TV station using a PLL in the synchronous demodulator. I made a small hardware modification which breaks this lock by connecting the RC-circuit which generates the AFC (automatic frequency control) to Vcc. Now I can actually freely tune and use the sampled 'videosignal' to feed gnuradio.

There is however one more problem. Because my tuner uses a switching synchronous demodulator it now has many harmonics of the received signal. If I receive a 2Mhz signal I also get 4 Mhz 6 Mhz 8 Mhz and so on.
Possible solutions:
1 Use the 38.9Mhz IF output directly in stead of the videosignal.
2 Trick the videodemodulator so it stops the PLL oscillator completely (You would then also get the 38.9 Mhz IF)
3 Only use the lowest output frequency of the videosignal.

I currently use option 3 .
The lowest frequency it outputs now is about 1.5 Mhz
So I use 2Mhz which is still below the first harmonic of the lowest frequency (2*1.5=3Mhz)

This way only narrowband (<1Mhz) signals can be received.

I am going to try option 1 and 2 also.
Best would be option 1 but I have to find an easy way of turning the differential IF signal into a single ended one. It is also a bit difficult to get to the If signal because it is not on any of the external leads of the tuner and inside it is all SMD and very hard to mount additional components.

I also made a small modification to the linux btaudio driver so I can get 850 kHz samplerate. This stream has no gaps so it is possible to get a full 425 kHz bandwidth signal.

The good thing:
I have not had to buy any additional components yet and can experiment with gnuradio. I also made the AFC trick non-permanent, I can just switch back and forth between gnuradio and normal TV reception.






reply via email to

[Prev in Thread] Current Thread [Next in Thread]