[BC] Getting phones on the air

[Broadcast List]

Keep in mind that once we decided on a single pair driven phone 
system (at least for the connection between the Central Office and 
the telephone set) phones have had hybrids in them.  A few countries 
use four wire systems all the way to the set, but hybrids are much 
less expensive than wire, and for voice communications the compromise 
is worth it.

Imagine though if we did have a four wire system.  Not only would on 
the air phone calls sound better, but the caller could be fed higher 
levels and actually hear what is going on.  (For the non-technical; 
broadcast hybrids feed the phone line at a lower than normal level, 
sometimes as much as 10 dB lower, in order to have less leakage back 
into the hybrid output/console input.)

Modems could be much faster as all frequencies could be used for both 
sending and receiving.

I recall our early experiments in improving call quality.  We did all 
kinds of stuff to try to make them sound better.  One of the things I 
did was to put one hybrid on each line, then weekly go into the 
studio and call a auto-answer silent number (in a CO, we had a list 
of many) and nulling each hybrid on each line.  I had discussions 
with the phone company about putting my hybrids in the CO and 
operating them remotely.  There was no Tariff so it was too 
difficult.  The Mass Calling Exchange (choked service) made good 
caller quality very difficult.  Each call to the Mass Calling 
exchange was treated as a long distance call and placed on a limited 
number of trunks between the caller's CO and he Mass Calling 
exchange.  This is how they prevented -- generally prevented -- 
contest calls from overloading the phone network.  Calls placed to 
the local CO circuits (hot line, warm line...) always sounded much better.

Anyway, it wasn't THAT long ago that we had to null our own hybrids.

MUCH later, the Telos 10 made all of this fooling around with hybrid 
transformers.

Regarding shipping audio around the country, it has been addressed, 
but even during my time at Mutual and NPR (the first two networks to 
distribute programming via satellite, by the way) we had loops of 
AT&T conditioned lines around the country.  These loops allowed us to 
monitor what we sent out in real time coming back.  The board 
operator in Master Control always had the responsibility of 
confirming the loop was working.  Mutual had a tic every second when 
the network and NPR fed classical music.  From time to time someone 
would use down time to be a "DJ" and record another reel of music (no 
announcing though).  NPR did many switches during the day.  Only a 
few stations could be switched in, and most feeds from stations were 
not sent live to the network, though any one between them and 
Washington, DC could hear it.  The recording technician would know 
AT&T had done the switch by hearing the music go away and start
recording.  The exception was NYC.  NPR had a SMALL office/studio 
there.  They had a direct conditioned line from there.  BBC had an 
office, and lines went between them.  NPR took some BBC feeds or at 
least feeds from England, and shipped them to DC.  It turned out that 
was against some law or tariff.  One day AT&T or NY Telephone was in 
the office and the show from England had to be fed over dial-up.

Both NPR and Mutual were 8 kHz circuits as I recall.  Some stations 
only got 5 kHz feeds off the loop.  It is amazing how much things 
have changed in 30 years!

One last thing, as hard as it was to keep these loops in spec, and 
even 15 kHz stereo conditioned STL loops, at least our amps were 
solid state.  Imagine doing that in the 1930s using hollow-state devices?

--chip

>Date: Tue, 01 Jan 2008 11:54:59 -0500
>From: Bob Tarsio <Bob at Broadcast-Devices.com>
>
>I suspect it was done much like it can still be done today by taking 
>the caller output from one side of the hybrid in a standard 
>telephone and feeding it to console input. Back then it would have 
>likely been transformer isolated and fed to a console input. The 
>idea of two way communication through a phone could have been 
>accomplished as well at that point as there were hybrid transformers 
>available to divide the caller and sender audio into discrete paths. 
>The isolation wasn't as good as can be done today with digital 
>techniques but it probably worked well enough.