On Mon, Jan 2, 2017 at 1:56 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
Hi folks,
Happy new year to all!
Is anyone here familiar with the innards of Viewdata? I'm still deep in the
Yes, and teletext. But not this chip....
bowels of this Executel viewdata phone system and one
thing that it should
be doing but isn't is puzzling me.
The teletext chip is a Plessey MR9735, datasheet here:
http://www.binarydinosaurs.co.uk/mr9735.pdf
It can work in either 'off hours' or 'on hours' meaning effectively
'day' or
'night' mode when there's no broadcast TV signal at night, this is the 80s
remember.
That was quite common. I remember pins in one of the Philips chipsets
with names like 'After Hours Sync'. Basically if there is a television signal
it will be synced to it (so you can, for example, overlay teletext on a
normal picture -- subtitles, etc), if there isn't then it can run stand-alone as
a display device.
The surrounding circuitry forces Off Hours by pulling high all the incoming
TV signals apart from Line Flyback which is pulled low. The datasheet says:
"When the incoming transmission is turned off, (i.e. Goes 'Off-hours'),
this
is recognised by the [On Hours Detector] after at least 300ms of missing
sync. Pulses. An internally generated Composite Sync signal is then switched
to the Composite Sync Out pin."
Since Sync In is pulled high there are never any negative sync pulses (I've
watched this on a logic analyser) so after 300ms Sync Out should become an
internally generated pulse, but this doesn't happen and Sync Out remains a
steady 5V meaning the TV picture is unsync'd.
I know the MR9735 itself is fine as I have a pair of Tandata viewdata
terminals which also use this chip in Off Hours mode and I can swap them
around. The chip itself is receiving a steady 6MHz clock to pin 21 and the
clock divider outputs at pins 20 and 19 are working.
Argh!. My first thoughts were 'defective chip' (but it isn't),
'missing master clock'
(but you have checked that) and 'Needs to be programmed for out-of-hours
sync' (Some Philips ICs had a bit in one of the internal registers for
this, this
one doesn't that I can see). I assume both power supplies are present
and correct
at the chip?
-tony