5963 (computer grade dual triode) production dates?

Ain't that TRUE! They also made some matching carrier, which was if anything even worse!

Ever seen any Lenkurt UCC-4 (???) carrier? Hermetically sealed, discrete solid state technology from the early 1970's, based on L carrier, and intended for long term operation in the jungle.

I worked on a site that had a Collins MW-108 microwave and Lenkurt UCC-4 carrier. And to top it all off, as late as 1979 there was functional but non-powered O carrier. The USAF had, in the same room, a North Electric Switch (the model I've forgotten), which had been installed in 1948, used. That was at Eielson AFB, near Fairbanks Alaska.

I wouldn't deny a word of that... :-)

Time will tell. I'm too old to care, because I can't imagine I'll be alive in 20 years to see how it all works out. And besides, from my point of view, I lost the war about 15 years ago. I've cared less and less what the idiots do every since... ;-)

That was a 1ESS, not a 4ESS, and the first one was cutover in July 1965 at Succasunna, NJ. Baltimore was cutover in July 1964.

It should also be noted that the 1ESS is an analog switch, not a digital switch. "Electronic Switching System" meant it was computerized, with store program control. The actual network switching fabric was made of analog relays.

The initial version of the 4ESS was introduced in 1976, and the first one was installed in Chicago. It was the first tandem switch produced with built in CCIS (Common Channel Interoffice Signaling, such as CCIS6 and later SS7) facilities as well as traditional inband signaling.

The currently used version of the 4ESS switch was introduced in the 1990's.

What went wrong with a DMS-200 after five years?

That's a strange number for a digital device. Well designed silicon doesn't wear out.

Was it running too hot? ???

The only thing that was wrong after five years was the politics!

"Oops, we should be closer to our good old buddies at AT&T than we thought." ;-) That of course usually happened about the time that a new CEO took over... and had *nothing* to do with technical, operational or accounting concerns.

Actually at the time it was possible to unmaintain a switch right into obsolescence in 4 years. By the late 80's, and then accelerating in the early 90's, there were so many technical changes in the functionality required, that if regular upgrades were not performed on time it would all of a sudden become a

Hmmm... no support when the system is 5 years old??? Sounds exactly like somebody didn't manage the regular upgrades properly. Note that the AUTOVON system was the only one to get away with doing that! But NTI had a totally separate support system for AUTOVON, and they wouldn't let the two systems even rub shoulders. Dealing with the AUTOVON support system in Richardson TX was significantly different that dealing with Raleigh (east coast) or San Ramon (west coast). They would let us do things like monitor what they were doing, because we were a civilian contractor. But on military switches they typically checked *regularly* to make sure nobody was printing out what they were doing. They were absolutely hiding all knowledge of how the switch worked from the military.

The primary motivation for these upgrades in the late 70's and early 80's had been simply better hardware, bug fixes, and continuing efforts at improved software. But in the mid 80's things changed, and ISDN, equal access, and then SS-7 all caused huge, and mandatory, software and hardware migrations. One other change, strictly of Northern Telecom's doing, was when they restructured their licensing to "encourage" LECs to implement a network topology using one main switch, with a million dollar software load, and many remotes with relatively inexpensive software loads costing less than a quarter of a million each. Every company with multiple full featured switching systems immediately converted as many as possible to remotes, served by a single front end.

People I knew that went over the Pac Bell after Divestiture decried the DMS' software problems mostly and lack of support from NorTel for same. Temperature problems were another thing, requiring modifications.

I'd heard that, as well. A visit to a DMS-100 installation they'd put in found the whole floor refrigerated to about 62°F and techs running around in coats. One should note that ESS switches were also adverse to heat, but the HVAC on the floor was designed to cool the switch, not air space.

Correct, which is why they had such a short life span. What the 1A did was allow various associated companies with depreciation schedules favoring retaining fully amortized equipment to finally replace the last panel and SXS exchanges while waiting to see what would develop for full time-space-time local swiching. The 1A improved end office performance and allowed exchange companies to offer unregulated features while cutting power and maintenance costs dramatically while retaining analog interface with existing short haul analog carrier systems without expensive conversion. It was an interim product at best, and the 5E conversions that came later were simplified by already having the end offices "groomed" for ESS. California was one such state in which equipment depreciation rules allowed for retention of both switching and facility equipment long after it was gone everywhere else, which is why SXS offices lasted almost to the '80s and 40 year old tube driven N-CXR systems still carried local toll tandem traffic. GTE retained SXS even longer, since their 1-EAX machine was problematic at best and their comparative lack of resources made waiting for the GTD5 to seem the logical move. Service in those offices was abyssmal, forcing GTE to deploy the GTD5 ahead of schedule with attendant problems along the way.

Although all production 4Es came with CCIS, wholesale conversion of the network didn't really get done until the around 1980. All in place 4A-XBAR machines were converted to CCIS by around 1980, with in-band signalling still left for interface with independents which did not convert at the time. Toll fraud was rampant in the '70s ("phone phreaking" had become something of a national pastime) and a multi-million dollar cost factor, which put the impetus behind conversion of end-of-life XBAR machines to CCIS. The cost/benefit analysis of the 4A CCIS implementation showed that, although remaining

'93, as I remember it. Comparatively few new offices have been built since.

Total nightmare. Hams that like old Collins gear wouldn't be Collins fans much longer if they only saw what Collins built for commercial microwave. It was a total disaster.

I read about it in the "Demodulator!"

Pacific Telephone, for reasons unknown, installed a lot of "ON" carrier in California in mid '50s, supposedly to allow remaining non-carrier open wire routes to be utilized for short haul carrier purposes. Meanwhile, the outside plant people were replacing open wire are a fast rate with 19 ga toll cable. ON would work on either open wire or cable, and would yield 16 channels instead of the usual

A last holdout for O-CXR must've been the Churchill County Telephone System, a county government owned telecom in the Nevada county of the same name, which now calls itself "CC Communications." 6 or 7 crossarms of exchange open wire linked Fallon and Fernley, NV, with O-CXR on a lot of it until around 2002. Although Nevada Bell, United/Sprint/Centel and others have tried to wrest that area away from its taxpayers since the '50s, the vote fails every time, and they enjoy some of the lowest telephone rates in the nation. CC Communications now is in the ISP business, probably the only publicly available taxpayer owned ISP around. There's an obvious synergy there, since the county doesn't have to build and operate its own communications network as do others...they already own the phone company!

dB

It would be correct to say that computerized common control stored program *analog* switches had a short life time, simply because the time lag between development of computers capable of controlling a telephone switch and the later development of an all digital switching fabric was just long enough to make it viable to produce an analog switch controlled by a computer. When the digital switching fabric became available only a decade later the vastly superior technology immediately made all analog switching systems obsolete and uneconomical to operate in most circumstances. (Granted that a great deal of the reason for that was not just the digital switch fabric, but also the vastly more advanced technology for both the controlling computers and for the various peripheral devices too.)

...

It was not designed to be an interim product. That just happened to be a fluke of history.

That is all interesting, but in fact the driving force behind the transition from *all* of the various mechanical and electronic analog systems to the modern digital systems developed in the 70's was cost based, but was not equipment cost or depreciation. Labor to maintain them was the significant cost factor. In areas where that labor cost was highest, digital systems almost instantly replaced everything else.

Which is why Alaska, for example, was virtually 100% digital by the mid-1980's, just over half a dozen years or so after the first digital systems were install. At the same time the entire US was only about 33% digital.

Where do you come up with these absurd statements? The first

In fact, AT&T was caught somewhat flat footed on that one, and the 4ESS had to be a totally new designed to allow use of the all digital network. Bell Northern Research had taken a different approach and designed a previous switch, with an analog network, as if it were a digital network... when the digital network became available they were the first to market simply because it was merely a drop in replacement in their product.

Without a hitch? Go on... It didn't suffer a major shutdown of the entire system, of course. But it was like any other project of that size and scope, and there were *many* local "hitches".

CCIS-6 was first installed in the Bell system starting in 1976.

Common Channel Signaling was developed in the mid to late 60's by the CCITT, and standardized as CCITT No. 6, which was modified by the Bell System and became CCIS-6. (Some CCITT No. 6 was used in the US though, for instance that is what Alascom used here in Alaska.)

CCITT began working on CCITT No. 7, which became SS7, in 1980! It was revised in 1984 and 1988.

...

I'm not sure what your point is, given that virtually *all* offices were converted ASAP.

Why don't you stuff that sort of garbage? That's *bullshit*.

So you are claiming they needed to cool a DMS to 62F to get it to operate? Do you have any other fairy tales? Why do you make up such goofy stories?

What you are saying has *nothing* to do with the DMS design. It says that the HVAC was improperly designed.

However, you are lying! I've seen many DMS installations, but I never saw a single one where any sort of external cooling that was *not* simply directed at the air space, was *ever* used. Not one of those switching systems had heat problems!

And the fact that they didn't was astounding, given that I've seen at least two of them operating with seriously inadaquate air conditioning, with the ambient temperature running as high as 90F! I can't say that was good for the equipment, but in fact it did *not* fail.

Just going by what they said at the time!

Hmmmm...well, I can only go by what people from Pac Bell were reporting in their engineering complaints, etc.

Heat stress testing of all ESS machines is part of the acceptance procedure, where the environment is heated to 95°F for at least 24 hours while the switch is performing simulated capacity operation. This isn't easy to do sometimes, since the latest hardware doesn't produce a lot of heat, and heating plant in telco COs went away many years ago except in the coldest climes.

Interesting. Do remember, though, that this was back in the '80s. Whether there have been design changes to take care of heat problems since then, I do not know. I DO know that Pac Bell was less than satisfied...far less. I also do remember them wrecking out the DMSes after installing 5Es shortly before. There can't be an "in house political" motivation for doing so, since all the RBOCs were by then independent of AT&T, and most, indeed, wanted to stay as far away from T as possible. That didn't last long, and happened before T spun off Lucent. Also, since the GTD5 switch went south, the majority of GTE installations have been 5Es, not DMS100s, although they do have DMSes active in end offices, including two in my area, and a good number of DMS/200 tandems. The latest add near me, though, was a new 5E. The DMS100s replaced SXS back in the '80s, while the 5E was a growth overbuild that replaced a GTD5, which in turn replaced an SXS office that was never fully utilized when new. Supposedly, the GTD5 was redeployed elsewhere, why, I can't say.

After seeing what both NorTel and LU have been through in the stock market in the last decade, I can imagine that every switch job goes into a frenetic bidding war. With Alcatel now gobbling up Lucent (and hopefully getting rid of Pat Russo once and for all time) it's possible Lucent will be the winner overall. Russo has been to Lucent as Carly Fiorina was to H-P..and that ain't good.

Sign of the times: Rumor is that the old Framingham, MA toll center (FRHMMA02), long a hub for L-CXR, radio and always a major toll switch in the metro "Bahhhstun ariyer" met the wrecking ball last month. No confirmation that this was the Waverly St. office..

Repeating fairy tales that half an ounce of logic would tell you are *wrong*.

I think *you* are the one who added the illogic to it. Somebody described something accurately, and you've adjusted it.

Bull. What you said is not *logical*.

...

Horse pucky. The biggest motivations they had were political cronyism. Every time a new CEO showed up, a whole new set of "rules" came down.

I would expect they actually have a lot of DMS-10's. Regardless, as I noted they do both, and it depends on the circumstances. Most likely that circumstance which is most important is what type of switch they must interface with.

That the majority are 5E's is not significant, but that there are a large number of DMS switches blows your other statements away.

Something about *accuracy*, rather than just rants, bothers you?

Which points higher-up towards technical management being unable to coordinate physical plant with other requirements, NOT a good sign! Not that the task is automatic when a center that used to take 3 floors now occupies a little (but much denser) closet. (Admittedly few sites went from one to the other without some intermediate steps.)

Gross mismatches between physical plant and machine room or equipment capacity and demand or software licensing and demand, at the very least bring up the notion of incompetent planning. At the most extreme we have extensive kickbacks for over-purchasing (and a governor who proved that yes, indeed, you can get fired for buying Oracle.)

Tim.

... "Chainsaw" Dunlap

Thanks! Reynolds was a footballer. "Hacksaw," "Chainsaw," "Neutron Jack"...there's a distinct pattern here.

The Collins C-band equipment they sold to CATV and TV stations was pure crap as well. Fixed tuned with a single channel filter at the antenna, and a mechanically tuned 4 GHz oscillator. A poor noise figure, and they ran so hot that the green fiberglass PC boards turned dark brown in a couple years of use.

To add insult to injury, AT&T contracted with Collins to provide GCE (ground communications equipment, in old-time sat parlance) for three AT&T C-band earth stations in the '70s built for the DOMSAT (we knew it as "Dumbsat") project. Although GTE's Lenkurt stuff was inflexible (had to tune 8 cavity waveguide filter sections for a simple frequency change), it at least as pretty good. What WASN'T good were the GTE-Milan FM demods...yoo wanna da noisa figura fora dat? Trying to get parts from GTE-Milan was near impossible, even if you spoke some Italian. They did built a very good cryo-cooled 12°K LNA back then, though. I was the cryogenics guru at my site for years. 12°K, fed by a 105' Cassagrain antenna, allowed us to do radio telescopy on slow weekends.