[time-nuts] LED reliability and sub cables
alan.melia at btinternet.com
Wed Apr 30 15:44:37 EDT 2008
Mmmmmm LED operation is basically a bulk effect whereas the reliability of
transistors depends more on surface effects.
In 1960 AT&T chose to go with Germanium, but the British Post Office chose
the new silicon planar technology for future submerged cables. Previously Dr
Gilbert Metson had written a landmark series of papers on valve (tube)
cathodes. Extensive work was done at the Research Department at Dollis Hill
in North London. The "good" tubes were found totally by accident in a batch
of wartime manufacture (I believe they were an SP41 (I dont know the US
equivalent) but its a 4v heater pentode.) As a result of this probably the
first repeatered telephone cable was laid between UK and Norther Ireland in
around 1944/45. The original repeaters had redundant sets of tubes, but it
was found that this made the system LESS reliable, than just a single string
of reliable tubes. I joined the PO sub cable transistor Group in 1961 and
from scratch we laid the first oceanic cable in 1968 (I think) between
Cornwall and Lisbon. The second was laid the following year between Canada
and Bermuda. The last of the tubed cables was laid about the same time. The
transistors were nothing special technically 400Mhz ft ( bit like the
2N916) but made with care and cleanliness and 75% of the build of each batch
was destructively life-tested ( by my Group). By 1967 we knew we could meet
the the tube system crtiteria of not more than one system failure due to a
component failure in 20 years. In fact I think most of these system
continued carrying traffic for near 30 years.
I am not sure about LEDs, though some work was done in the 1970s on the
efficiency degradation-rate in opto-couplers for the telecomms business.....
that is a difficult one to accelerate easily....sting in the tail MTBFs are
Slightly OT but "life" is measured as elasped time....so I claim immunity
----- Original Message -----
From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
To: "Discussion of precise time and frequency measurement"
<time-nuts at febo.com>
Sent: Wednesday, April 30, 2008 6:12 PM
Subject: Re: [time-nuts] LED reliability
> In message <20080430165656.3C300BE3B at ip-64-139-1-69.sjc.megapath.net>, Hal
> ay writes:
> >> LEDs been viable since at least the early 70's? Talk about planned
> >> obsolescence...
> >How reliable were the early LEDs? When did they start to get used in
> >reliability applications?
> Many of them are still happily emitting their faint red or green light.
> >The best reliability story I heard (many years ago) was about installing
> >another trans-Atlantic telephone cable. They used tubes long after
> >transistors were out. They knew how long the tubes would last. They
> >have much data on transistors yet.
> Actually, that's not entirely correct: They had reliability info
> on transistors and they sucked.
> Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG | TCP/IP since RFC 956
> FreeBSD committer | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> and follow the instructions there.
More information about the time-nuts