In message ,
Terry Casey writes
In article ,
says...




I recall that certain systems insisted that the four set-top
bypass channels had to be close to the standard off-air broadcast
channels, because some TV sets would not tune to anything but these.


They would have to be very unusual TV sets!

I don't know about 'unusual', but they were a problem. I think there
were only couple of budget brands which only tuned 'spot-on' to the UHF
channels (xxx.25MHz, in 8MHz steps). One might ask indeed "Why would you
need them to do otherwise?" Of course, even our cable set-top boxes
could normally only tune in 125kHz steps, but at least that got you to
within +/-63kHz of the correct frequency - and that was more than close
enough.

It was more likely to be, in the case of Westminster, that, when CATV
systems rarely went above 600MHz, there was nowhere else to put them,
coupled with the fact that the off-air channels were left clear, so it
was convenient utilise to this for the n + 2 arrangement by straddling
the otherwise blank off-air allocation.

A comb of 7.988636364MHz would allow E25, E27, E29 and E31 to be used
with +/-34kHz error (off-airs being 23, 26, 30 & 33)

Of course, this was all long before the Channel 5 debacle - I can't see
a way of interleaving 5 channels around Crystal Palace without involving
the allegedly taboo n + 5 scenario - although I've never seen a problem
with any set I used directly connected to a CATV network

Sets generally seemed to improve a lot in later years. I think that the
change of IF from 39.5MHz to the European 38.9MHz made quite a
difference to N+/- problems. What surprises me is how well some sets
could tolerate having direct inputs of 48+ channels (without them going
through the converter UHF bypass filtering). Certainly, in the olden
days, when faced with more than half a dozen channels, some sets tended
to sag a bit at the knees.

But, as you have said, there used to be so many embargoed channels on a
cable TV system - no adjacent, no N+/-5, no N+/-9, no sums or
differences (with single-ended amplifiers) etc. It's a wonder anyone was
able to get more than two or three channels!


But wasn't it at the same place which ingeniously used a not-quite-8MHz
comb reference which was actually derived from one of the UHF off-air
channels?

Well, dividing E26 by 64 or E30 by 68 would do the trick. I based my
comb frequency on E28, being the centre channel but an off-air lock
would certainly produce a very stable result, and the offsets would
still be reasonable - +58/-11kHz or +11/-58kHz, depending on choice of
off air channel.

If this is what they did, they could have used either of those channels
from Crystal Palace. Next time I see him, I'll ask the man who will
almost certainly know (if I remember!).

As the headend equipment was largely supplied by the 'other'
company, I doubt if I would have been involved with tinkering with it
(although I'm pretty sure that I did swap one or two of the modulator
SAW filters because of the problems which arose when NICAM started).
Maybe 'my' comb generator was a replacement.

My involvement with the Westminster system was at the time of the DTV
roll-out (or possibly Broadband Internet, I can't remember which) which
coincided with the transfer of the system from BT to ntl, so I never saw
the BT headend but I did see the documentation related to it, complete
with frequency details.

I had little to do with the system in London (I think I only went there
once - underground, near Shepherds Bush IIRC). As I said, my involvement
was among the concrete cows and the roundabouts.
--
Ian