Re: Extracting the video signal from a 38.9MHz TV IF

On Fri, 31 Jul 2009 16:31:02 +0100, Terry Casey
<k.type@xxxxxxxxxxxxxxx> wrote:

In article <00b81b96$0$4795$c3e8da3@xxxxxxxxxxxxxxxxx>, usenet09
@philpem.me.uk says...

Hi guys,
I'm hacking together a VBI inserter/extractor. Basically, it's a little
black box with an FPGA and some analogue support parts inside it, which
receives a TV RF signal, strips the Teletext data off of it, and spits
said data out of a parallel port. On top of that, it reads Teletext page
data out of a RAM, and can either inject it into the output video
(composite, CVBS, FBAS, whatever you want to call it) or leave the
original data intact.

"It slices, it dices, it juliennes"

The problem -- and, as always, there's a problem -- is with the RF input
side of things. While I can live with CVBS in/out, I'd rather like to be
able to feed it a signal from my TV aerial, and tune in (say) BBC1 and
browse Ceefax from the comfort of my computer desk. Or just leave the
thing idling on BBC1 and dump the entire of Ceefax to my PC's hard drive.

Now obviously I need a tuner -- that's not a problem. I have a box full
of Philips UV916MD tuners (think: UV916M with a booster amplifier tied
to the front, and a LO/DX input). I give the tuner a frequency, and half
a second or so later either the LOCK output goes active, or it stays low
and I notice (after lots of debugging) that the aerial cable is
unplugged... :)

The tuner spits out a 38.9MHz IF signal. That is to say, the signal I
want is centred at 38.9MHz, with the sound carrier at (38.9+6)=44.6MHz.
Frankly I couldn't care less about the sound, as long as the video is in
good enough condition that I can extract it and pull the data out of the
VBI.


UK vision IF is 39.5MHz, sound IF is 6MHz LOWER at 33.5MHz.

Some UK receivers do use the european 38.9MHz but the sound will be @
32.9MHz. (Not really any difference, as far as the tuner is concerned.)

What I haven't been able to find out for definite is what modulation is
used for UK-standard (which if memory serves is PAL I) RF video. About
half of the references I've looked at say it's ordinary double-sideband
AM, the other half say it's vestigial sideband modulation (aka VSB).
Does anyone know for definite?


Lower vestigial sideband - only the upper sideband is tranmitted in full
but be carefull - the IF is inverted in the tuner (which is wht you got
the sound IF wrong above.)

Now as I understand it, if it's AM I can just filter off the sound
carrier, then use a synchrodyne-type receiver to get the video (which
would probably involve designing a PLL to recover the carrier, and
finding a suitable <100MHz mixer in my junkbox). Alternatively I could
take the cheapskate's way out, and use a germanium or Schottky diode and
a capacitor to demodulate it. I'd be a little worried about HF response
with this option, though...

If it's SSB, then my only real option is to lock a PLL off the carrier
(from experience, it's hard enough to get the things to lock to a pure
50% duty 5V/0V signal) then use a mixer to pull the signal from IF to
baseband.

And now for the problems:
* I don't have any 38.9MHz TV IF (SAW) filters, nor can I find them
in any of the catalogues I've looked in. I know Maplin used to sell them
*years* ago. Does anyone know of a UK supplier of small quantities of
these?


Can't help but bear in mind that the SAW replaces a complex filter which
ensures the correct vestigial sideband response, including co-channel
and adjacent channel sound traps and adjacent channel sound and vision
traps.

Most important part of the design of this filter, as far as teletext is
concerned, is the group delay response. SAWs were first used in early
teletext receivers because of the difficulty of eliminating group delay
using discrete LC filters. You really need to keep it withing 50nS
across the band and it probably rather late in the day to start your own
learning curve!

* There's no way a 4046 PLL is going to run at nearly 40MHz; the
data*** for the HC4046 says 12MHz max. So if I go down the synchrodyne
route, I need something that can handle a higher input frequency. Does
anyone have any suggestions?



There are lots of ICs specifically designed for the purpose. The Philips
TDA3853T is one example picked purely at random

http://www.datasheetcatalog.org/datasheet/philips/TDA3853T.pdf

Note that the (test) application circuit specifically omits IF filters
and rejectors (see SAW above)

Your best bet will be to get a SAW IF strip from a scap TV.

Hurry up though, otherwise you won't get your project finished before
the analogue transmissions are switched off!

AIUI, the signal that's modulated onto the video carrier is the
composite video signal, so all I'd have to do at the end is a bit of
amplification and offset-correction (probably including AGC) to bring
the sync, black and white levels back to where they should be...

Lastly, does anyone know of a reliable reference (or indeed more than
one reference) for this type of technology? As in, TV, video and so
forth (maybe radio as well, though I'm already aware of the existence of
the ARRL Handbook). I've seen "Basic Television and Video Systems" by
Grob which looked like it might be worth a punt (or the £5 Amazon wanted
at least), but some other references would be useful.


None I can recall although I'm sure there are lots (I never read one
because I learn't my TV and electronics stuff the 'drip-feed' way from
the age of about 12 - nearly 20 years before teletext and SAWs
appeared!)

Try you local library. You never know what you might find on the shelves
- if it's a decent library, of course! Try both lending and reference
sections. Dewey references are 621.384 (Radio) & 621.388 (TV), IIRC.

Geez, i may have started both earlier and younger, under ten and about
50 years ago. Damn, half a century in electronics already.
.

Loading