EEVblog #1169 - TI 1972 Computer Interfacing

[boffin]

SLOW = 110 baud = 10 characters/sec of (start + 8 data + 2 stop)

A lot of 110 baud systems used 1.5 stop bits...

Indeed, but I didn't want to complicate things further....   After posting last night I went and did some reading about 20mA loop; as I had completely forgotten about it.

Now the big question, how many remember using a TELEX ?  (FIGS+WHO ARE YOU)  -- how did I get so old ?

[bsudbrink]

how did I get so old ?

You haven't died yet.   
My father (age 90) always says that he hates his birthday but that "It beats the alternative".

[HwAoRrDk]

No need to mock the device, for some applications a physical print was preferred. For example systems I worked on often the console was a printer/keyboard.

I recall a blog post from Microsoft's Raymond Chen talking about how they used to use serial terminals like these when doing kernel-level debugging on Windows 3.x and 9x. The advantage being that you could spit out debug information that wouldn't need to be logged in memory and would persist (because it's on paper) in the event of a crash.

Perhaps I can find the link later. Couldn't find it.

[WN1X]

When I was co-op engineer in college, we used these terminals.  Ours were the version with two tape drives.  The application was to generate configurations for microwave relay system racks.  One drive would prompt the operator for various options.  The responses were written to the second drive.  That tape would then be downloaded to a Univac 1108 mainframe that would punch a card deck that was sent to an IBM mainframe that produced another card deck of all the parts that would be needed to build the system.

Ahh, the old days I spent many hours writing code and running it on a Univac 1108 for the Navy. 36 bit words, core memory,  tape drives and drum (not disk) storage.

[HKJ]

Ahh, the old days I spent many hours writing code and running it on a Univac 1108 for the Navy. 36 bit words, core memory,  tape drives and drum (not disk) storage.

The 36 bit meant 6 bit characters in filenames. It could also use disk storage, big removable disk stacks. It may have been a later model, I am sure it had solid state memory and it could handle about 100 users with 512k memory (I believe). Input was mostly from VDU's, but some people did also use punched cards with output on a some line printers.

[texaspyro]

Ahh, the old days I spent many hours writing code and running it on a Univac 1108 for the Navy. 36 bit words, core memory,  tape drives and drum (not disk) storage.

They also used disk drives.

Anybody experience the subtle wonders of a bearing going out on a FASTRAND drum?   A couple of tons of spinning iron running at around 1000 RPM...  bearing goes out, drum launches through concrete walls... great fun.   Unless they were firmly bolted to the floor, the first FASTRAND systems would walk around the room as the earth rotated under them.  Later units had two drums spinning in opposite directions.

I have a couple of core memory stacks from the 1108.

[helius]

See https://multicians.org/low-bottle-pressure.html
This site has lots of great stories from early mainframe engineers.

[bsfeechannel]

Lastly to the loser that in here said "archaic piece of junk", we wouldn't have the stuff we do, without innovative stuff like the Silent 700.

True.

This terminal is more generically called a teleprinter. Teleprinters were invented long before computers.

Dave got amazed that the terminal didn't even have a screen, but still in 1972, computers were expected to print out their results. Video terminals became an affordable alternative a few years later.

We owe to this "archaic" technology the existence of the Web.

Tim Berners-Lee created the first Web browser, but it could only run on NeXT OS. So he commissioned Nicola Pellow to write a portable version: the Line Mode Browser. it emulated the behavior of a teleprinter, but it is what made the Web known to the general public for the first time.

[DTJ]

I used on of these Silent 700's as a 12 year old in around 1978. My father would borrow it from work for me to play around with. I'd put together programs in BASIC and run them and chew through a pile of paper roll.

I think it was a PDP 11/70 that I used to log into.


The home phone was near the kitchen.

When using it after dinner cutlery being clinked together in the kitchen would be picked up by the acoustic coupler and cause random character to be generated.

[coppice]

Dave got amazed that the terminal didn't even have a screen, but still in 1972, computers were expected to print out their results. Video terminals became an affordable alternative a few years later.

In 1972 even the most basic video displays made the silent 700 look cheap. It was very rare to see one, and when you did it only displayed 8 or 12 lines of upper case only text. In 1975 DEC launched the VT52, which was one of the first decent 80x24 video displays, capable of displaying upper and lower case text. It was rather large, ran hot, and cost a fortune. It was largely made practical by advances in semiconductor RAM.  In 1978 DEC launched the VT100. By this time they were putting some sophistication, like smooth scrolling, into the terminals, and they were much more compact. They were still several thousand dollars, though.

[GeorgeOfTheJungle]

Dave clean the printhead with alcohol, the side that touches the paper, with a cotton swab.

[boffin]

Dave got amazed that the terminal didn't even have a screen, but still in 1972, computers were expected to print out their results. Video terminals became an affordable alternative a few years later.

In 1972 even the most basic video displays made the silent 700 look cheap. It was very rare to see one, and when you did it only displayed 8 or 12 lines of upper case only text. In 1975 DEC launched the VT52, which was one of the first decent 80x24 video displays, capable of displaying upper and lower case text. It was rather large, ran hot, and cost a fortune. It was largely made practical by advances in semiconductor RAM.  In 1978 DEC launched the VT100. By this time they were putting some sophistication, like smooth scrolling, into the terminals, and they were much more compact. They were still several thousand dollars, though.

Ahh the VT52, great device, and one of the best keyboards ever.  The keyclick was generated by a big relay that just clunked closed when you typed.  The baud adjust was two multiway switches underneath which was annoying, but most of the time ours were 2400 or 9600 fixed.

The VT100 on the otherhand was much more modern, but the keyclick/feel and the smooth scrolling were just crappy by comparison.  The VT102 is when DEC started to really integrate everything on a single board.

[floobydust]

8080 development started late 1972 and it was released in 1974. Not sure why the product ads seem premature 

Dot-matrix impact-printheads were king in this era. Ribbon and ink everywhere, quite the mess fixing them. A silent thermal printer is way ahead of its time here.

I wouldn't expect any of the electrolytic/tantalum capacitors to be working. It looks like it runs, the CPU. Wow.

[German_EE]

At the first place I worked at we needed a serial terminal but the Texas Instruments devices were way too expensive. So, the command from on high was to build one.

We started with an Epson MX-80 which was programmed via its DIP switches to print one character at a time and not do a line feed and carriage return afterwards, this had a parallel input. We then found a keyboard in the junk pile that had an 8-bit parallel output. Finally we raided stores for a pile of 74LS devices including lots of shift registers and spent a week wiring it all together including RS-232 done through transistor level shifters.

Now for a lesson in economics.

One nearly new Epson MX-80 printer plus four man weeks of time probably cost more than one of these Texas Instruments terminals. We tried to tell them before we even started but they got their 'terminal' anyway and the men in suits were happy.

[Silveruser]

SLOW = 110 baud = 10 characters/sec of (start + 8 data + 2 stop)

A lot of 110 baud systems used 1.5 stop bits...

Indeed, but I didn't want to complicate things further....   After posting last night I went and did some reading about 20mA loop; as I had completely forgotten about it.

Now the big question, how many remember using a TELEX ?  (FIGS+WHO ARE YOU)  -- how did I get so old ?

Not only do remember TELEX, spent a few years fixing them.  Also worked on 720C, a version of the Silent 700 used for CPIC ( Canadian Police Information Computer).  It was a much heavier version full die cast case. Here is was coded on a plug in board with a diode matrix - you'd cut the leads on the diodes to code it. Weak points were the print head and end of line mech which used grain of wheat bulbs.  The 720c were replaced with an early 8080 based Inteligent Terminal pre PC days for sure.

[Quarlo Klobrigney]

I used one of those back around 1982-83 to send in field reports from equipment dumps (via the modem) in far flung remote monitoring sites.

I later found out about BBS's. I would log in, and a menu system would print out. The bugger of it all is that if you selected a menu item, and if you went back to the main screen, it would print it all over again. Roll after roll of thermal paper that I'm sure was not cheap back then. The fact I could speak to a mainframe 3000 miles away was sheer magic.

So after all these years, I get to see what's under the hood. I was really afraid to open it as seen by the price of it back then. Old happy memories of being a self-autonomous employee. Show up at the site once in a while, make sure nothing caught fire or blew up, and send the reports and all was good.

[westfw]

8080 development started late 1972 and it was released in 1974.

Huh.  If the first silent700s shipped in 1972, well before the 8080 was available, then I think that implies that as a product line, this "terminal" went through quite a bit of product redesign over its lifetime.  1972 till "mid 1980s" is actually a pretty impressive life for a computer product during that timeframe (pre-8080 to post IBM-PC!)

I used on of these Silent 700's as a 12 year old in around 1978. My father would borrow it from work for me to play around with.

Hey, me too!  Though more like 1976 and 16 years old.  I also played a lot of Star Trek.  Used to be something like $4/hour for the phone call (which dad complained about. $4 was a lot in 1976.  Never complained about the paper use, though.  Oil Companies!  It didn't come home THAT often, though.)

The advantage being that you could spit out debug information that wouldn't need to be logged in memory and would persist (because it's on paper) in the event of a crash.

It was "standard practice" to have a printing terminal connected to the "console" of your mainframe, well into the 1980s (until mainframes themselves started to die out, essentially.)

No need to mock the device, for some applications a physical print was preferred.

In those days, print was pretty much all that was available.  What would you count as the first portable "screen-based" thing?  The Osborne 1 (in 1981) - no modem, a 52 column display, and $1795?  Maybe the TRS80 Model 100 (1983) - 8x40 display WITH a modem - $1100?  Maybe the Toshiba T1100 in 1985 (Full 80x25 display!)

Dave got amazed that the terminal didn't even have a screen, but still in 1972, computers were expected to print out their results. Video terminals became an affordable alternative a few years later.

Quite a few years later.

In 1972 even the most basic video displays made the silent 700 look cheap.

I think we tend to forget just how insanely expensive it was to do anything with computers back in those days.

Yeah, that.  Lots and lots of that!  $1995 was quite a bargin for a Terminal AND a MODEM.  Contemporary CRTs ran over $1000, plus close to another $1000 for the MODEM.  Display cards for CPM systems that did 80 columns (the "standard" from punch cards, you know)  were expensive, and required special high-resolution CRTs with high-resolution phospors that were difficult to find and quite pricey.  It wasn't until after widespread IBMPC Clones that "computer monitors" became common.  (Original IBM PC Monochrome display adapter + (12inch) hidef green monitor was about $700.)

[texaspyro]

I still have and use a Toshiba T1000 and know a company that still runs a production line with over 20 of them.

[westfw]

I still have and use a Toshiba T1000

Heh.  I bought one when they got cheap.  The battery died a long time ago, it wouldn't run without it, replacement batteries were expensive and/or hard to obtain, and I never did get around to hacking it.  I think it's still in my garage, though...

[floobydust]

You'd splash on some Hai Karate, toss the TI terminal in your car and drive to work. Perhaps in a car of the era, a 1972 Mercury Monterey for example.
Cost around $5,000 curb weight: 2,015kg/4,442lbs, 7L 429CID engine, and it had maybe 15 transistors.

[texaspyro]

The battery died a long time ago, it wouldn't run without it, replacement batteries were expensive and/or hard to obtain, and I never did get around to hacking it.  I think it's still in my garage, though...

The battery is just 4 NiCad cells in an open-frame plastic holder...  trivial to rebuild.  I switched to using LSD NiMH cells long ago.  Every 2-3 years that company that runs a production line on T1000's opens them up,  vacuums out the fluff, and swaps out the cells.   The damn things are still going strong.

[blacksheeplogic]

No need to mock the device, for some applications a physical print was preferred.

In those days, print was pretty much all that was available.  What would you count as the first portable "screen-based" thing?  The Osborne 1 (in 1981) - no modem, a 52 column display, and $1795?  Maybe the TRS80 Model 100 (1983) - 8x40 display WITH a modem - $1100?  Maybe the Toshiba T1100 in 1985 (Full 80x25 display!)

I'm old enough to remember using a lugable in the mid-80's commercially.

My point was primary the way the was portrayed without any f'in idea of what it was to work in the industry at that time, the limitations/constraints, expense, the applications being developed & run and what was actually required/needed commercially. These did the job the needed to, were beautiful in their time and did not deserve the stupid ranting and laughter looking at it with 30+ years of technology in between.

[westfw]

The [Toshiba T-1000] battery is just 4 NiCad cells in an open-frame plastic holder...  trivial to rebuild.

Doesn't it have some sort of "Sense" terminal as well?  (I'm pretty sure I've lost the actual pack.)  I was disappointed that it failed to run with no battery installed :-(
I would've mailed it to Dave for a tear-down, if it was such an expensive PITA to mail things internationally from the US.  :-(

[German_EE]

Didn't Brother also produce a typewriter that was a serial terminal?

Edit. found it, the Brother EP44
https://typewriterdatabase.com/1984-brother-ep44.5403.typewriter

[coppice]

Didn't Brother also produce a typewriter that was a serial terminal?

Edit. found it, the Brother EP44
https://typewriterdatabase.com/1984-brother-ep44.5403.typewriter

There were a lot of typrewriter + terminal machines, but not many portable ones. Quite early on IBM started making versions of their Selectric golf ball typewriters with a computer interface. These were mostly for mainframe console use, but offered one of the few high quality print options as well.

That Brother machine was a later generation of portable devices. I think it dates from 1984.