13 Sep
2002
13 Sep
'02
7:28 p.m.
Tony Duell wrote:
[snip]
I don't want it to generate a TTX data stream. I want it to generate a video
signal for the TTX to lock onto and the Cub to sync to. I don't care if it
has no TTX data to lock onto - I can just load its RAM and send it a
"DISPLAY ON" command.
Might be worth a shot. Thanks. Now, I wonder if the central library is open
on a Saturday...
Ooh - another company to track down :-)
Ah, right. So it
doesn't matter if I use, say, a Velleman "professional" 10x
probe and leave the outer contact unconnected? Or I could just shove a piece
of tinfoil in there to short it out...
What it _won't_ do is decode PAL (or any other)
colour video. There's
a separate chip for that in the TV.
I don't WANT it to decode PAL. I'm
using it for a display system in a
selfbuilt computer. I'm going to add a video tuner later (i.e. when I've
reverse-engineered one of the VCR tuners in my junk^M^M^M^M quality used
parts box).
The RGB outputs from the teletext
board are combined with that chip. There's often another output from
the teletext board which is used to blank the 'live' video. So you
can have live video only (teletext decoder produces no RGB ouptuts),
teletext only (teletext decoder produces RGB _and_ blanks live video)
That's
the mode I'm planning to use it in.
vidgen ->
TTX board.
Anyway, I'll see if my PICDREAM video generator can generate a good
enough signal for the TTX board to generate some form of video
signal. The 5281
I doubt the PICDREAM will produce a valid teletext data stream...
Oh, how I wish
I had a TX91G service manual so I could find out just
how this pesky thing works and what Thomson originally designed it
to use... The
Have you tried the local public library? Main ones often have the
Radio and Television Servicing books, and the TX91 is likely to be in
one of those. Alas my only collection of such books only goes up to
1981. Be warned that when they changed the teletext decoder
board they may
also have changed the control microcontroller. So it's not definite
that your board repsonds to the same commands as the SAA52xx series.
I think
I've got the MCU somewhere. Don't know if it still works, though.
Know anyone
that stocks the SAA5243? Can't find it on Farnell's
site, nor does my CPC catalogue list it.
Grandata used to do it (I bought one from
there to use in a very
simple I2C colour video display (4 chips IIRC -- the SAA5243, a
74LS04 (clock oscillator, mostly), a 6264 RAM and a 74LS244 (output
buffer)). See if you can track down the Teletext PCB from a
Thomson ICC5
chassis. It would be ideal for you. There are several versions, but
every one I've seen contains an SAA5243, SAA5231 and a 6264 RAM. Some
have a custom on-screen-display chip that you can easily desolder (to
keep it off the bus). The support components are there too. From what
I remmeber you need to feed it +12V (and/or +5V, I forget just what
PSU components are on-board) and composite video. It's got a standard
I2C bus to control it. The RGB outputs are transistor-buffered to
strange levels, but you can either add a few more transistors to make
them sane again or tap the signals off the SAA5243 pins.
Time to track down a
junked ICC5 chassis. Anyone on here got one they don't
mind pulling the Teletext board out of? :-)
I'll see
if I can persuade the TTX board to generate any video
tonight. I've got a Tek 466 storage scope and a Gould OS1100A in
case I need an extra two channels. Anyone know of a source for Tek
1x/10x probes? The 466 seems to want the ring around the BNC
connector shorting to the shield (GND) to push it into 10x mode.
Guess how many probes I've found that do that... That's
That only sets the
on-screen display and/or the lamps on the range
switch. It doesn't actually affect the operation of the 'scope (the
probe is a passive /10 divider, the 'scope will always display the
signal it receives). You can use any probe with the 466 (provided you
can match it to the input capacitance of the 'scope), the extra
contact is just to make it easier to read the range switch, etc. IIRC, that pin is shorted to ground for the
'identify' mode (moves the
trace vertically slightly and changes the on-screen-display to
'INDENTIFY' so you can tell which trace corresponds to the probe).
It's connected to ground via a rrsistor to tell the 'scope what
division ratio the probe is. Older 'scopes just support *1
(open-circuit) and *10. Later ones support *100 (and maybe others)
too.
The 466 seems to support 1x/10x only. No OSD either.
The trace does look a bit "grainy", but I guess that's because it's a
storage scope.
BTW, anyone interested in a Gould OS1100A? 30MHz, delay sweep, single
timebase, dual channel, no manual, a bit heavy. Make me an offer - buyer
collects or sends a courier round for it.
Later.
--
Phil.
philpem(a)dsl.pipex.com
http://www.philpem.dsl.pipex.com/