My new paperweight

[vasrc]

Worked for Royal Dutch Phone company in the early 90ties, and was part of the AT&T/Philips ESS-5 group, which replaced all those old Dutch mechanical stepper switches for the electronic version, and from one of the centrals I helped convert I got one of the stepper switches, looked like:

Had some interesting talk with those maintaining them, and a lot of curse words were part of that conversation indeed.

And to think that only took care of a single digit.

They used bikes to ride through the building section that contained all those switches for a "city", whereas it was all replaced by a few ESS-5 racks.

Buying one from eBay is not the same to me as having the one I tore out myself, but it was nice to be reminded on them by your post on history.

Yup, that's the evil switcher

Luckily, it was a small switch in a small town in Wyoming, so other than making a lot of noise (and maintenance) it didn't require any ladders
Was soon replaced by an ESS as well once they started to lay fiber (many of the Railroads here in the US were some of the first to install it since they had so much right-of-way.. , ie SPRINT (Southern Pacific Railroad Internal Network Telecommunications))..

Z

[ericg]

Very cool. How was the data distributed afterwards? Any used for real time or printed reports to individual departments?

That was a bit before my time. I worked mostly with DEC PDP and VAX systems (Pipeline SCADA control)

Z

Data distribution depended on lots of variables. For launches from Vandenberg AFB, we could pick up real time in-flight telemetry with our rooftop antennae in Huntington Beach, CA. From there we stripped out FM and PDM data channels to banks of strip chart recorders in real time. And yes, I've seen engineers unrolling strip charts on the floor down the halls a few times. Eventually we got a large dedicated room for strip chart display. After a launch, it was a horizontal wallpapering operation.

Most of our launches were from Cape Canaveral. So we had to wait for analog tapes recorded at ground stations located there, Bermuda, Antilles, Tananarive, and other strange places I've forgotten.

The strip charts were mostly for people in a hurry. We digitized the FM and PDM data, and stripped the PCM channels. It all went to computer readable magnetic tape reels -- thousands of them. Every rocket has thousands of sensors, and every sensor has unique certified calibration data. A major task for the digital computer was to apply each sensor's calibration info to convert raw bit counts to engineering units. (The strip charts didn't do calibrations.)

Most commonly, calibrated data went to high speed (1000 lines/minute) drum based impact line printers. During busy periods we typically printed 5-10 boxes of 11" x 15" fan-fold paper per day. Our printouts went to individual departments -- Propulsion, Electrical, Thermal, Mechanical were big customers.

The Huntsville Saturn engineer was right -- often, the most important frame of data in the entire transmission was the one you only got half of, because the rocket was exploding....

Our CDC 924 (every line of code written in assembly language, assembled and committed to punched paper (or Mylar) tape!) eventually got phased out after we acquired our PDP-10 (DECSystem-10) computer, for which I was sysadmin. The new machine was a game changer. Switching from assembly language to Fortran (and an OS!) greatly increased our coding efficiency, and time sharing made it possible for individual engineers to gain direct access to their calibrated data. Most importantly, time sharing suddenly allowed software development to proceed concurrently with production jobs -- a major benefit! We eventually installed some Tektronix 4010 graphic terminals that allowed them to plot their data and print it on attached thermal printers.

Which PDP computers did you use? I've also worked with PDP-6, PDP-11, PDP-15, and a few flavors of VAX. I enjoyed working with none of them as much as the PDP-10. Were you a member of DECUS?

[ericg]

One day, somewhere in the '60s, I dropped by a circuit lab in my company. A friend showed me a a DIP chip he had just sampled. He told me it was an Intel memory chip, implemented entirely in an IC. I asked him how many bits it held, and he replied, "Eight." I opined that, if it were ever to become useful, a chip would have to hold a lot more than 8 bits. He said, "They're working on it."

[BlairG]

My first programming was on an IBM 360-40. Coding in COBAL, RPG and my favorite Assembly ALC

80 column punch cards. I remember a platter hard drive looked like a washing machine with a plastic dome on top.

Wow those were the days.. moving/storing all that data..

[vasrc]

Data distribution depended on lots of variables. For launches from Vandenberg AFB, we could pick up real time in-flight telemetry with our rooftop antennae in Huntington Beach, CA. From there we stripped out FM and PDM data channels to banks of strip chart recorders in real time. And yes, I've seen engineers unrolling strip charts on the floor down the halls a few times. Eventually we got a large dedicated room for strip chart display. After a launch, it was a horizontal wallpapering operation.

Most of our launches were from Cape Canaveral. So we had to wait for analog tapes recorded at ground stations located there, Bermuda, Antilles, Tananarive, and other strange places I've forgotten.

The strip charts were mostly for people in a hurry. We digitized the FM and PDM data, and stripped the PCM channels. It all went to computer readable magnetic tape reels -- thousands of them. Every rocket has thousands of sensors, and every sensor has unique certified calibration data. A major task for the digital computer was to apply each sensor's calibration info to convert raw bit counts to engineering units. (The strip charts didn't do calibrations.)

Most commonly, calibrated data went to high speed (1000 lines/minute) drum based impact line printers. During busy periods we typically printed 5-10 boxes of 11" x 15" fan-fold paper per day. Our printouts went to individual departments -- Propulsion, Electrical, Thermal, Mechanical were big customers.

The Huntsville Saturn engineer was right -- often, the most important frame of data in the entire transmission was the one you only got half of, because the rocket was exploding....

Our CDC 924 (every line of code written in assembly language, assembled and committed to punched paper (or Mylar) tape!) eventually got phased out after we acquired our PDP-10 (DECSystem-10) computer, for which I was sysadmin. The new machine was a game changer. Switching from assembly language to Fortran (and an OS!) greatly increased our coding efficiency, and time sharing made it possible for individual engineers to gain direct access to their calibrated data. Most importantly, time sharing suddenly allowed software development to proceed concurrently with production jobs -- a major benefit! We eventually installed some Tektronix 4010 graphic terminals that allowed them to plot their data and print it on attached thermal printers.

Which PDP computers did you use? I've also worked with PDP-6, PDP-11, PDP-15, and a few flavors of VAX. I enjoyed working with none of them as much as the PDP-10. Were you a member of DECUS?

Sounds both impressive, stressful and busy. I can imagine what direct access to data did for the engineers, let alone displaying/graphing it.

Fortran was what we wrote our first SCADA system in first on PDP10's and then VAX's. We also used PDP11's in the compressor stations. At the time they were impressive but I can't imagine trying to do so nowadays. We (relative to what you were gathering) were only pulling in hundreds of pts/sec, tiny by today's standard, but done over 9600 baud (woo hoo!). I spent a lot of nights in the middle of nowhere, Wyoming/Utah/Washington/Colorado sleeping in a compressor station control room trying to debug DEC's Analog and Digital I/O boards or trying to explain to the station mechanic over the phone how to press (Ctrl-C) on the keyboard (What, you want me to push them both at the same time??)..

Also, remember DECnet and hardline Ethernet

Yup, was a member of DECUS.. Still have one of their coffee mugs with the Chesire cat.

Good times..
Z

[RoChess]

"They're working on it."

In a way they are still working on it

[RoChess]

ie SPRINT (Southern Pacific Railroad Internal Network Telecommunications))..

Never knew their company name was an acronym and was related to railroad. Always figured it was tied to the English word for running/sprinting, which obviously it is a play on, but acronym makes much more sense with their history as it was much easier for them to lay down foundation network alongside their tracks.

But those stepper switches sure would make a noise that was ear deafening when you stood inside the rooms. The rooms I had access to had about 500,000 of those switches in them, hence the need for a bike during maintenance. And it got all replaced by ESS-5 equipment that fitted in 1/1000th of the size with more capacity and much faster. Maintenance was mainly done from a console with the occasional walk to replace a faulty module that was highlighted by LEDs.

[Meapilot]

That is ALL so interesting to me. In some ways it does "blow my mind". Interesting that this was all in the 60's basically and when the "home computer" become available to the public. Depressing that the modem speed was same as my first dial up connection ( SLOW ) ha

[dibble9012]

That’s a cool find! I love old tech like that.

I know these are not the greatest pictures (all the parts are currently wrapped in boxes). I plan on eventually doing new cleaner shadow boxes to protect them. My grandfather worked at IBM.

[vasrc]

That’s a cool find! I love old tech like that.

I know these are not the greatest pictures (all the parts are currently wrapped in boxes). I plan on eventually doing new cleaner shadow boxes to protect them. My grandfather worked at IBM

Very nice collection, you definitely should display it formally. The first image is gold. Your Grandfather worked for IBM when they were famous for taking care of their employees.
https://www.marketplace.org/2016/06/13/profit-ibm/

Z

[dibble9012]

Very nice collection, you definitely should display it formally. The first image is gold. Your Grandfather worked for IBM when they were famous for taking care of their employees.
https://www.marketplace.org/2016/06/13/profit-ibm/

Z

I found a whole lot of other stuff when he passed away. Lithography masks. Silicon wafers. IBM product presentation slides. And a few programming punch cards. I thought about lending them to a museum but most museums want them donated and I’m not willing to let them go completely.

[Mark42]

Ran across this thread and decided to add something from my archive..

This a core memory card from a General Electric (GEPAC) 4020 Process Control mini computer probably built circa 1967-ish. It was retired from petrochemical service around 1995.

The card has 4096 24 bit words plus a single parity bit per word for a whopping 102,400 ferrite cores on the card. I can tell you from personal experience a single parity bit is NOT enough.

32K word memory was pretty standard on those machines, but if you were living large then you could get 64K.

The original DDC machines came with drum mass storage, the supervisory machines would have (tiny) multi platter disk drives. The cooling fans were 1/8 horsepower and would remove an errant finger. And lets not forget the IBM selectric I/O typers, card reader and card punch! All in all a lot of noise....

Best Regards

[logbuilder]

Wow, blast from the past! I started with NCR in 1977 fresh out of college. Core memory like this was still in use. Early on, these were all strung by hand so therefore very expensive. The first systems I worked on only addressed 64k since they were 8 bit processors. You would be amazed at how much you can do with 64k. Early systems based on octal (base 8). I miss my old TI Programmer calc. Disk drive technology was new too. A 5MB drive on some of our mini computers would be the most you could get and cost thousands. We backed up to tape. Mostly programming done in ASM but then RM/COBOL eventually took over. I worked in COBOL for years and then Pascal (Turbo), Basic, C, Perl. My first personal computer ran CPM. I could PIP with the best of them. My first modem was a 300 baud acoustic couple. When I got a Hayes 1200, I thought it just couldn't get any better. I ran a BBS from home. It ran on a MS-DOS PC using DoubleDOS. My career was in retail technology so I was on the leading edge of scanning. Bank cards still required the knuckle buster and a carbon sales slip. If you wanted to make two computers talk, better get familiar with RS232 or 485 and the UART. Later became specialized in 2780/3780 comms and SNA. Got real familiar with datascopes.

Thanks for the trip down memory lane. I'm now retired and my main contact with technology is HomeSeer. Thanks to the community for allowing an old fart to participate.

Oh, those were the days my friend!

[Mark42]

Glad you enjoyed the trip down memory lane..

I started in the process control trade in 1975 as a systems engineer for a large petrochemical company working on the GE/Honeywell process control computers.

Honeywell bought GE's process control division in Phoenix in the early 1970's and continued the development of the 4000 series of computers, i.e. 4400, 4500, 45000. In the early days each machine was a 'snowflake', based on the same hardware and OS software but customized to every customers requirements.

They shipped as compiled binaries with listings and required 24/7 hardware and software support. As a customer systems engineer, you had to deal with bugs and design/install patches to fix the bugs. At the systems level everything was done in assembler, and since you could not reassemble the programs in the field, patches were installed directly into the binaries (steady nerves and fingers required). Things did get better in the field in later years, but I like to think that early 'hard' training paid off down the road.

Over the years I programmed in whatever language was required, Fortran, Pascal, various flavors of Basic,some C; my last contract involved writing a lot of javascript.

Like you I am retired and am currently migrating (and de-kludging) my HS2 scripts to a shiny new HomeTroller HS4 system.

But the little HomeTroller rectangular box just doesnt have the personality of the old machines where you could literally see what was happening by watching the lights on the console panel, not to mention warming your hands on them (no LEDs here) in the (always) massively over air conditioned computer rooms!😀

Best Regards

[vasrc]

Ahh the good old days. I suspect there are more than a few 'old farts' hanging around this board, especially some from the automation/control fields. Process control/Automation is a lot more complex nowadays.. Sort of miss not being able to play with all the new, sparkly toys that are being used now..

Probably lots of "war stories" as well.. Be fun to hear some of them.

For me, I'll always remember working in remote compressor stations, alone, late at night after a 4 hour drive through snowdrifts, hoping not to hear all of the engines go quite if you pulled the wrong wire (I only did that once, which was more than enough). Still better than venting the station however

Here's what the internet looked like when I was a freshman at the Univ of Utah. I remember a PDP in the Physics dept in the corner, not sure if it was this one..