RS-232 standard -- the most non-standard ever! Only (?) TI used that reverse channel. I spent a 3 hour WTF session attempting to determine the same issue early in my career, so I sorta knew where you were going. Nice to see it working!
I worked for a telco, we used every pin, synchronous clocks, back channel, the lot. Then IBM cut it back to 9 pins, WHAT! Back then, with what we would see from IBM etc, I was almost waiting for them to move from a 25pin connector to say a 37pin job - yet they did the exact opposite.
I got a bin full of console cables for various brands of networking equipment. They all use the RJ45 connector but are wired differently. Even the same brand like cisco or Dell might have 3 different wiring configurations. Its so infuriating 😄
@@zaca211 That's one downside to using RJ45's for serial, rather than the original D25, Even going 9 pin D has it's issues as there are many twists, IBM PC, Centurion, and many more.
I used to use one of those TI 810 printers 40 years ago in my first job. It was a beast. I recognised it immediately, before you even showed it close up. I also remember the joy of creating a serial cable that satisfies both ends to go from the serial port on a computer (pre PC compatible, CP/M usually) to a printer, often without a manual for the printer, and sometimes without a technical manual for the computer either.
Aside from the corrosion, this thing is an absolute tank! I didn't have to replace any parts, just fix bad connections is all. I was digging through a manual looking for any indication on pinout on the printer side, but turns out, I was looking in the wrong manual. I can't imagine the battle of trying to do it without a manual at all though!
@@joeybruins I interfaced several printers to CP/M and Unix machines over the course of a few years. Most were happy with a very simple cable, pin 2 or 3 (on a DB-25) as the data, pin 7 as the common ground, and pin 20 to do flow control. Occasionally, one end or the other needed its flow control pins (usually 6, 8, and 20) connected together to all signal the same thing, like Usagi did. We had several TI810s which we gradually sold to customers. I'm not old enough to have used a chain printer, but the 400cps dot matrix one that we had was the fastest printer I experienced until the advent of laser printers.
ti 810s were real workhorses. I’m surprised they’re hard to find. Were used for several decades by all airline reservation systems, car rental agencies, etc. My first job was coding COBOL on ti’s minicomputer systems. 880 printers were faster than the 810s, we had them everywhere.
It's possibly they were scrapped along with the systems when/if upgrades to newer tech happened. But who knows, maybe there's a long forgotten pile of these things in some storage room.
Brings me back to my childhood with my Parent's Mac and Apple ImageWriter, the only thing better sounding from that time was the big honkin' Wang daisywheel printer that my Dad got with the old decommissioned Wang OIS that he got from work. That Wang system isn't something you see at all these days, the closest I've ever seen is the Wang setup on this channel.
@@SeishukuS12 I remember those daisywheel printers… there was one in my data processing class in high school. Apparently there was something wrong, because when you powered it up, it would just spin the daisywheel like an angle grinder. And we didn’t know what the heck was going on and tried to open where the daisywheel was, and it would catch one of the letters and break off and go flying in the room somewhere. So we had a few daisywheels missing teeth 😂
@@UsagiElectric I seem to recall "back in the day" there were files you could send to dot matrix printers, that mad them hum at different pitches etc, and play rudimentary songs. THe other cool thing was a bit later then postscript laser printers started to happen (DEC LPS40 and LN03R I used a lot) - a lot of people would code custom postscript forms for calendars, diary inserts etc, but postscript being a general programming language, we coded up a calculation of Pi to n digits program - Tell it 10 digits and it would think for about 20 seconds, and print, 100 digits was about 10 minutes - we gave it 1000 digits, and it sat there thinking for around 3 days, and then spewed out an accurate page of pi.
FYI - The Centurion printer type PRTQ is the nane of tbe Centurion print spooler program. The print spooler lets you send a print job to the PRTQ at full system speed and not be limited to the slow printer speed. Once the print spooler program receives the print job the printer starts printing the job, and the system user to do other jobs as the spooler prints the last job. In the case of this CPU5 system OPSYS gen. PRT0 is setup with the TI-810 printer driver.
My thoughts as well. I worked a different system/ OS that also used a spooler. You could send print job to 'prt' (printer) or 'slo' (spooler listing output). Standard practice was to send to 'slo' to have job run quick without waiting on printer. But if the spooler task dies, had to have sysop/ admin restart it. Wonder if there's some startup script that needs to start PRTQ spooler on Centurion?
Early in my career I spent many many hours listening to a TI-810 hammering out reams of green-bar paper. In my application for the job they stated that I had to be able to lift 45 pounds, an oddly specific number, and also the weight of one box of green-bar paper.
RS232 IMO is the best example of a non-standard 'standard' interface. When I was dealing with RS232 connections back in the 1970/80's I used to first use two straight-through ribbon cables with a female/male connector on each end connected via a RS232 breakout box. Then I would sort out the required pin connections via the box and then when it's working I would make up a proper cable with the required pin connections. Back in the day some terminals required straight through, some required x-over for pins 2 and 3 - and then there's handshaking! There was one particular computer I worked with which required a high on a certain pin to work but didn't set any other pin high until it saw that particular pin high. Nightmare as you couldn't just connect control pins together in the plug to get it to work. There was this one serial printer that a customer wanted to use that required a high on a pin for it to become ready and before it would raise any other pin high. Needles to say that getting that printer to work with that computer was not easy. Ah the bad-old good-old days......
The problem mainly was that manufacturers did not understand that their equipment usually was a DTE, and that the standard said that DTE equipment should have the male connector. They put female connectors on their equipment, but wired them as DTE. That meant it would usually not work on the first try. When they adhered to the standard (like later with the IBM PC), they put a male connector on their DTE equipment, and all you needed was a straight male-to-female cable to connect it to a DCE, possibly combined with a "null modem" block or cable that had female on each end and crossed all connections.
Ugh. You are 100% right. I used to carry an RS232 breakout box in my toolkit, plus a ribbon cable with M & F DB25 connectors on both ends. Such a pain in the arse. Eventually, you learn to be able to spot baudrate & other errors by what kind of garbage came out of the printer. I do not miss those days at all!
Nice job! A trick we used to do with old ribbons on printers to refresh them was to squirt a little bit of WD40 on them. This redistributes ink from the edges of the ribbon which never see any action, down to the "battle zone", so to speak where the pins contact the ribbon, and it extends the useful life of the ribbon. We used this as a cost-saving measure because me and my friends couldn't afford ribbons that often. All good wishes.
Hi David, great work again! I used to repair similar printers for Digital Equipment in the 80s and immediately thought of an encoder problem when the printhead did not move to its left margin at power up. A word of advice: if you ever come across a dot matrix printer from DEC, do NOT remove the encoder PCB from the carriage motor unless you have the manual describing the correct encoder adjustment procedure. On those printers, you must usually power the motor with some 5V or so, hook up your scope to the photo sensors, and then physically move the block housing the photo sensors only fractions of millimeters to produce the correctly timed pulses. You need a scope and very calm hands, it's a really fiddly task.
I have to add: Wow! Considering this might be the only extant operating Bendix G15, you're doing the Lord's work! Thank you, bless you, and we fervantly hope you feel appreciated, because you certainly are! All good wishes and can't wait to see more about this project!
The TI810 was the very first dot matrix printer I worked with in 1979. I hooked the Centronics interface to our own Commodore IEEE to Centronics Parallel interface box and produced lots of output on it. I do remember how loud it was. Great to see it turn up at your place and restoring it! I always found it an awesome printer.
Honestly, I'm surprised the corrosion wasn't more of an issue. The fact that it only caused fatal damage to a socket and flex PCB is a testament to how well those boards were made. Great work getting it going again!
This brought back memories. In the late 80's I worked with all sorts of dot matrix line (?) printers connected to mainframes. The sound is like music to my ears.
Living in The Netherlands and owner of 23 pinball machines I sure loved your 2 in the background. Good to hear your brother will take good care of it (mayby them soon). Understandably you made a wise decision from your hobby's point of view. Love it. Keep up the good work! Cheers from Amsterdam
Memories of me refilling ribbon cartridges with ink squeezed out of fountain pen cartouches... And driving that cartridge with a drill to help distribute the ink. Saved me quite a buck back then :D
I had a TI-810. Don't remember where I got it, but I used it for quite a while and was very happy with it. It was fast, but definitely not quiet. Nice printer. I *think* it had a parallel interface. I believe there were options for the interface. I gave it up when I no longer had a source of free tractor feed paper :-) Lots of businesses (and maybe airlines?) used it for multipart carbon or carbonless forms. Very heavy duty machine.
I spent years fighting the RS-232 interface. A patch box was my friend! Never saw a Pin 11 handshake, I thought it was going to be xon/xoff missing as no return transmit path. Still love the noise of a dot matrix. Another great video. Thanks.
We had these at an auto parts store rebranded as Triad. Came with our triad point of sale system. They ran for 20+ years printing carbon copy receipts and reports daily without fail.
Finishes with a listing that has code fixes dated 1977. 47 years ago... Love it. Back about then, I was a Field Service Tech for DEC in a remote site - everything from PDP8 to 10s, LA-series printers and old teletypes - when my boss told me I didn't need an oscilloscope to fix things anymore, I moved into SOftware - I bought one of their excess inventory Tek475 scopes for $50, which is still in my lab and still working.
Ironically, an oscilloscope is probably essential to diagnosing, repairing any serious problem with a computer... unless you are okay with just throwing money at the problem (I.e. replace whatever is faulty).
At least you didn't have to repair the print head by replacing broken pins with bits of guitar strings. I did it once on a 24-pin Oki Microline printer and it was a, uhm, satisfying nightmare :D
Strange that when Centurion customised the printer they apparently left the badge recess empty and printed their logo on the centre of the panel instead of having a badge made, the empty space is very obvious - and I see you decided to put the TI badge back which looks much better.
8 years ago i was sent to fix a printer at a logistics company. It was an epson dot matrix that printed all the slips for the drivers. When i got there there was 20 trucks backed up and two frustrated older IT guys. 26 year old me wsltzed in like nothing and was immediately turned around. They said the printer was older than me and there was no way i could fix it.
Well, they're 3x the age of the printer and they hadn't fixed it. I _HATE_ morons like that... there's no way anyone could ever know something they don't.
Yikes! I supported one of these for years and years. It was fast and loud! As far as I know the reason they are hard to find is the airlines still use them.
@@Stuart-AJC That could be, it's been years. I remember buying one for our warehouse and having to drive to some Airport equipment place in Newark to pick it up! Seemed like they literally purchased everyone of them at the time as "stock".
I've seen a number of dot matrix printers used for passenger manifests but I don't believe any of them were wide format like this one. They tend to be 80s Okis
@@fcguy7okidata - I remember the name, can’t remember what they looked like. I remember when there was actually a job called “travel agent”. Their office almost always had one of these 810s. Just one example, I remember the plastic daisy wheel printers to the really messy chain printers that would burn holes in the paper when they’d jam. Had this 18” (?) wide sheet of cloth ink that was hell to change.
The pinout is probably optimized for direct connection to an old 300/1200 baud modem that didn’t support flow control. Dispatch would call the remote site with the printer attached to the modem and dump a bunch of printouts, such as work orders or bulletins.
I used to work with those machines in the airline I still work today. They were used to print tickets, the actual airline tickets that would have 4 or 5 ways, with red carbon (anyone remember that?). I joined the airline in 1999 and they were already old. Surprisingly they were still being used all the way until 2007 when paper tickets were ultimately phased out and the airline adopted e-ticket. Other Airlines would have been using other kinds of paper tickets back then, but we used the regular paper ones, they were called T.A.T. tickets. The printer is an absolute TANK, and nothing could stop it. The only things I ever saw breaking in those printer's were the sprocket belts, and that was it.
Great work Dave and grats to the Discord crew for the assists. Ken Romaine comes in clutch again. I love a happy ending story involving the Centurion in any way.
When I taught assembly language programming long ago, one of the more satisfying programs was to read a keystroke and send it to the dot matrix printer. Everyone was hoping that tap tap tap on the keyboard would become zap zap zap on the printer. It didn't quite work like that because the old "Proprinter" was bidirectional and would queue up input characters. But in the end we found a way to make it flush its buffer and more or less print instantly as we wanted.
Paint it BLUE!!! We musts paints it blue! But seriously if it were me I would recreate it by painting it blue and then getting some vinyl printed Centurion logo recreation and have the whole system be classic and matching and everything. In any case thanks for all these great videos!
The magic! YUP! The magic of making old stuff working is so great! That printer is older than all my kids, Maybe older than me. And it still works. Can we say that planned obsolescence was not in the mind of the maker of that product, on the contrary, they made sure it would last for a lifetime! Yes some component did not lasted that long. But who knew that this type of chemical reaction would happen after decades... So happy that you have that system fully working. Congrats!!
What a great piece of nostalgia, love this sound! I never worked on these types of line printers, but was responsible for repairing and maintaining several dozen Okidata line printers, amongst lots of other tech, in a distribution center for a number of years.
9:41 The LM339 is a quad comparator with open collector outputs. It is not an op-amp. It works down to its minus rail. It takes about 1.5uS to react to its inputs It pulls down with about 6mA when the (+) is below the (-) It releases its output when (+) is above the (-)
THANK YOU! It's an easy mistake to make, because the symbol for an opamp is the same as a comparator is the same, but jeez, the 339 is the classic comparator!
Very nice. It brings back old memories. I serviced many different makes and models of dot matrix printers. Specifically the centronics ACT-900 which used a similar motor to produce return feedback pulses failed. Fixed the jammed motor to make the printer to work. This happened some time in 1987
I had an 810 Ro back in the day (1985 ish) when I was in college. I had it hooked up to a Commodore 64 originally and then a 10mhz AT clone. I could not figure out the handshaking and ran it at 1200. I only had the upper case roms. Used it mostly for printing computer listings when I dialed into college (I was a computer science major). Don't get your hand inside the machine when the carriage is moving... it will mess up your fingers. Ask me how I know 🙂
I was shouting at the screen when you were saying cabling issues. RS232 is easy when you understand the different layers of it. Each of the following are at opposite ends of the connector i.e. a end -> b end GND -> GND, since you always need a ground reference TXD -> RXD and RXD -> TXD so you have a data flow for the printed data, this is just data, no flow control, unless you use X-On / X-Off software flow control. The pins meanings are Transmit Data / Receive Data. This is the absolute minimum needed for communication. RTS -> CTS and CTS -> RTS, so you have hardware flow control. The pin meanings are Ready to Send and Cleared to Send. This is needed to slow down the sender at the right time. DTR -> DSR + DCD at one end and then and DCD + DSR -> DTR so you have modem flow control. The pin meanings are Data Terminal Ready, Data Set Ready, Data Carrier Detect. This is for modem control, so another layer of flow control for the modem, its only relevant if you have a modem in a link. If you need a full modem support for incoming call, then you also need Ring Indicator (RI), in this case, then connect RI -> Data Carrier Detect (DCD), since the receiving machine is being told its detecting a carrier signal. Do the opposite at the other end DCD -> RI Things will then work properly at all layers. When some of the other layers are missing (i.e. no modem / null modem), this is what's results in the mangled cabling that connects between the different lines above. Generally this is a bad thing, but sometimes its necessary to get the right levels on the right pins to make sure that each end thinks that the other bits of the RS232 protocol see the correct levels on the correct pins.
The problem here is that the printer doesn't have CTS/RTS, the flow control is buried away on Pin 11. 24:50 The DCD/DTR/DSR was only saying the printer was on-line, not acting as a flow control.
I worked at Texas Instruments in Cypress, Texas and my desk was about 20 feet from the 810 and 825 printhead lifecycle test room. 40 or 50 or maybe even more printheads running 24 seven. Try and imagine the racket.
great work, I enjoy watching your channel. as an old school dot matrix printer owner, I learned that you can spray a little wd-40 on the ink spool to get the ink darker or 'refreshed'. I used that when the printer company no longer made my print ink spools for my okidata
My father had a cp/m s-100 and a printer that looks just like this one, though it sounded different when it printed. (I think it was a daisy wheel rather than dot matrix.) He spent DAYS troubleshooting that damned printer. Not just once, but again and again first for the physical connections, then for each GD program that came in. But it was the physical connection steoy that was funniest. He spent so long on trying to get the pinout right that he had just given up and decided to get drunk instead. And whike he was thoroughly buzzed but not quite in his cups (you know that moment where you still know its a bad idea but you're fine with doing it anyway?) he decided it would be fun to throw darts at the cable. I mean, it already didnt work, whats the worst that could happen? 😅 Well, he had forgotten that the computer was still sending signal to be printed and on the third or fourth dart throw, he severed exactly the right wire and the damned printer started spitting out characters. Random dart throwing FIXED THE PROBLEM!! Of course, subsequent investigation made him realize he hadn't read the instructions fully and completely. But thats a different side of the story. 😂
Back in the around 2002 to 2003 a local business I worked for had 2 of these TI printers, still working and being used to print receipts for transactions. They upgraded to new printers and so the guys took the printers as well as old monitors, loaded them on a boom lift, and launched them from 40 feet in the air to the concrete below. Then the mess was gathered up and chucked in the dumpster. I guess back then, they were just considered old printers.
I used two of these at work One was a KSR and one was an RO. The only thing that ever happened was the hub the carriage cable was connected to the motor cracked and I had to replace that. Otherwise never a problem. Stone reliable and they ran 24/7 as a administration terminal and a call detail recording logging printer on a Nortel SL1 VLE phone system.
I really feel when testing the Centurion and printer you should have been wearing a shirt and tie if not a full suit :D Enjoyed the video and the genuine excitement when things work is great!
I fix modern computers.......... I don't have the love for the old ones....... BUT I totally appreciate and am motivated by your "don't give up" attitude with these projects.
Wow!! My first job was testing 810's for CADO Systems......back in 1979. We swapped out the character-generator ROM because CADO had a few custom characters. I definitely recognize the internal guts. Well-built machine.
Had not heard the CADO System computer name for over 40 years ago when I was working at Warrex / Centurion. We had a Centurion dealer in Atlanta, GA that also sold CADO system ( I think for Typesetting uses ).
@@kenromaine2387 It's possible you are referring to the CAT photo-typesetter, which was not associated with CADO. Most of the systems sold by CADO were for small business accounting, payroll, inventory, purchasing, and billing. The 810 was CADO's first-supported dot-matrix printer, and was replaced by a much cheaper (and lower-quality) unit from Dataroyal.
Worked on these TI printers back in the early 1990's when hp had just taken over the Texas Instruments printer division. Mainly board swaps and cable changes, I can imagine the customers reaction if I had gotten as excited after repair 😇. Those cable data lines were never any fun that is for certain! Sort of nice to live in a plug and play world of today but I do miss those days too.
It would be a great idea to put that printer in the bedroom and use it as an alarm clock, (printing out a weather report from NOAA every morning). 😄 Regards. 👋
This carries a scent memory for me. I can smell the fresh folded paper packs and the plastic packs the tape reels came in. 40 years and it was like the thing was sitting behind me again wobbling away on some 2 by 4s.
I once had a job calibrating and modifying these kind of printers before they were sold and that was in the nineties, so still then they were used. I also remember a cabinet below it to house the chain paper.
Oh, yea, even as late as ‘96 I had to maintain several that were used as report printers. Most had been replaced with high speed, quieter, ones capable of multi-part forms (some with automated stuffers), but these TI’s were the workhorses for general report printing.
Reverse channel. That really takes me back. I knew it existed but I never worked with it. You used to be able to get half duplex lines from the telephone company. If you were only using it for a printer or if you used polled Bisync, they worked great and they were cheaper than a full duplex line. You could used them for voice but if you didn't have control of the line, you would have a dropout. Reverse Channel was used to turn the line around so data could be sent the other way. By the 70s and 80s it was pretty well phased out but it appears that a few were still supporting it. I would have expected CTS/RTS flow control but the description for the printer clearly indicates that it's not supported.
One of the tricks we used to use back in the day to extend the life of dot matrix printer ribbons was to give them a spray WD40 and let them sit over night. This lets the ink from the top and bottom edges of the ribbon diffuse through the rest of the ribbon.
Back in the day I had an ancient PC with a monochrome display and an old dot-matrix form feed printer that I had poached from various different people's trash. I somehow managed to get the serial port working so that I could transfer GW-BASIC from a laptop onto the PC, then I went digging through the printer's manual and started experimenting with it. I was able to put the printer in "high-density" graphics mode, at which point I could send the printer arbitrary bytes to drive all of the pins in its column independently. The ribbon was pretty weak, so I compensated by overdriving each pin 4x to make the image darker before moving the carriage, but I was able to successfully print monochrome images. I suppose if I varied how many times I drove each pin, I could have even printed different shades of gray! Those were fun times. I no longer have the equipment but I should go digging through my old stuff and see if I still have a few of the sheets that I printed.
9:34 the LM339 is a quad comparator with open-collector outputs, note that it's used with positive feedback (for hysteresis) rather than negative feedback that you'd typically apply to an op-amp.
I was a Static RAM Product Engineer for the AM 9101 at AMD in the 80's. The 9101 was pin compatible with Intel's 2101. It is a 256 by 4 bit RAM. Most likely a buffer for the serial input. Imagine having to design with 1K RAM chips today!
With apologies to JFK: "We choose to go to the Moon in this decade and do the other things, not because they are easy, but because we _thought_ they would be easy." 🤣
Those TO-220 transistors with the solid metal backs (the TIP150s at around 6:45)... when my company was still servicing intercom boards for a major brand, we were told to watch for these and replace when found. Never really found out why that directive.
The quick brown fox says HELLORLD!
😆
So that's what the fox says...
@@JohnSmith-xq1pz and the bunny
@@SeekingTheLoveThatGodMeans7648And the strange man on the screen
Your enthusiasm on getting this old tech working is why I continue to follow your channel.
Thank you so much!
Also, the bunnies.
Came here to say exactly the same thing. Your enthusiasm always makes me smile.
cannot agree more on comments, dedication here seems a really small word
RS-232 standard -- the most non-standard ever!
Only (?) TI used that reverse channel. I spent a 3 hour WTF session attempting to determine the same issue early in my career, so I sorta knew where you were going. Nice to see it working!
I worked for a telco, we used every pin, synchronous clocks, back channel, the lot. Then IBM cut it back to 9 pins, WHAT!
Back then, with what we would see from IBM etc, I was almost waiting for them to move from a 25pin connector to say a 37pin job - yet they did the exact opposite.
RS-232 is so flexible. That's what makes it seem complicated.
I got a bin full of console cables for various brands of networking equipment. They all use the RJ45 connector but are wired differently. Even the same brand like cisco or Dell might have 3 different wiring configurations. Its so infuriating 😄
@@zaca211 Why bother with a standard if they won't follow the standard?
@@zaca211 That's one downside to using RJ45's for serial, rather than the original D25, Even going 9 pin D has it's issues as there are many twists, IBM PC, Centurion, and many more.
There's something very satisfying about the sound of a dot matrix printer whirring away.
Combined with the sound of a floppy drive
I used to use one of those TI 810 printers 40 years ago in my first job. It was a beast. I recognised it immediately, before you even showed it close up.
I also remember the joy of creating a serial cable that satisfies both ends to go from the serial port on a computer (pre PC compatible, CP/M usually) to a printer, often without a manual for the printer, and sometimes without a technical manual for the computer either.
Aside from the corrosion, this thing is an absolute tank! I didn't have to replace any parts, just fix bad connections is all.
I was digging through a manual looking for any indication on pinout on the printer side, but turns out, I was looking in the wrong manual. I can't imagine the battle of trying to do it without a manual at all though!
how did that go to work? as it probaply wasnt easy to just measure the pins with a multi meter
@@joeybruins I interfaced several printers to CP/M and Unix machines over the course of a few years. Most were happy with a very simple cable, pin 2 or 3 (on a DB-25) as the data, pin 7 as the common ground, and pin 20 to do flow control.
Occasionally, one end or the other needed its flow control pins (usually 6, 8, and 20) connected together to all signal the same thing, like Usagi did.
We had several TI810s which we gradually sold to customers. I'm not old enough to have used a chain printer, but the 400cps dot matrix one that we had was the fastest printer I experienced until the advent of laser printers.
ti 810s were real workhorses. I’m surprised they’re hard to find. Were used for several decades by all airline reservation systems, car rental agencies, etc. My first job was coding COBOL on ti’s minicomputer systems. 880 printers were faster than the 810s, we had them everywhere.
It's possibly they were scrapped along with the systems when/if upgrades to newer tech happened.
But who knows, maybe there's a long forgotten pile of these things in some storage room.
Printers are just disobedient little robots, but it was a lot of fun watching you get this one to behave.
24:00 Please no, my eyes hurt. That was a flashbang on my eyes.
Ahhh... I love the sound of a busy dot-matrix printer.
I'm a bit partial to the rhythmic chunka chunka of the Data-100 chain printer, but the TI-810 at full chat is definitely a cool sound!
Brings me back to my childhood with my Parent's Mac and Apple ImageWriter, the only thing better sounding from that time was the big honkin' Wang daisywheel printer that my Dad got with the old decommissioned Wang OIS that he got from work.
That Wang system isn't something you see at all these days, the closest I've ever seen is the Wang setup on this channel.
I used to hear them all the time when i was in banks,medical labs,offices etc... I miss the sound too.
@@SeishukuS12 I remember those daisywheel printers… there was one in my data processing class in high school. Apparently there was something wrong, because when you powered it up, it would just spin the daisywheel like an angle grinder. And we didn’t know what the heck was going on and tried to open where the daisywheel was, and it would catch one of the letters and break off and go flying in the room somewhere. So we had a few daisywheels missing teeth 😂
@@UsagiElectric I seem to recall "back in the day" there were files you could send to dot matrix printers, that mad them hum at different pitches etc, and play rudimentary songs.
THe other cool thing was a bit later then postscript laser printers started to happen (DEC LPS40 and LN03R I used a lot) - a lot of people would code custom postscript forms for calendars, diary inserts etc, but postscript being a general programming language, we coded up a calculation of Pi to n digits program - Tell it 10 digits and it would think for about 20 seconds, and print, 100 digits was about 10 minutes - we gave it 1000 digits, and it sat there thinking for around 3 days, and then spewed out an accurate page of pi.
FYI - The Centurion printer type PRTQ is the nane of tbe Centurion print spooler program. The print spooler lets you send a print job to the PRTQ at full system speed and not be limited to the slow printer speed. Once the print spooler program receives the print job the printer starts printing the job, and the system user to do other jobs as the spooler prints the last job. In the case of this CPU5 system OPSYS gen. PRT0 is setup with the TI-810 printer driver.
My thoughts as well. I worked a different system/ OS that also used a spooler. You could send print job to 'prt' (printer) or 'slo' (spooler listing output). Standard practice was to send to 'slo' to have job run quick without waiting on printer. But if the spooler task dies, had to have sysop/ admin restart it.
Wonder if there's some startup script that needs to start PRTQ spooler on Centurion?
@@mikefochtman7164 It should have a JCL file with a the JCL commands ( guess you could call it a script ) to start up the Centurion print spooler.
Yeah, as soon as I saw 'Q' in the name, I figured that it was a spooler. lol
Seeing the printer action at the end gives Tech Tangents vibes. You just need to get the bunny staring at the printer.
Early in my career I spent many many hours listening to a TI-810 hammering out reams of green-bar paper. In my application for the job they stated that I had to be able to lift 45 pounds, an oddly specific number, and also the weight of one box of green-bar paper.
I might still have a 132 print position template around here. Remember when the screen was 80 x 24?
I never thought I'd witness someone getting so excited about a printer light. Your enthusiasm is infectious.
RS232 IMO is the best example of a non-standard 'standard' interface. When I was dealing with RS232 connections back in the 1970/80's I used to first use two straight-through ribbon cables with a female/male connector on each end connected via a RS232 breakout box. Then I would sort out the required pin connections via the box and then when it's working I would make up a proper cable with the required pin connections. Back in the day some terminals required straight through, some required x-over for pins 2 and 3 - and then there's handshaking! There was one particular computer I worked with which required a high on a certain pin to work but didn't set any other pin high until it saw that particular pin high. Nightmare as you couldn't just connect control pins together in the plug to get it to work. There was this one serial printer that a customer wanted to use that required a high on a pin for it to become ready and before it would raise any other pin high. Needles to say that getting that printer to work with that computer was not easy. Ah the bad-old good-old days......
The problem mainly was that manufacturers did not understand that their equipment usually was a DTE, and that the standard said that DTE equipment should have the male connector.
They put female connectors on their equipment, but wired them as DTE. That meant it would usually not work on the first try.
When they adhered to the standard (like later with the IBM PC), they put a male connector on their DTE equipment, and all you needed was a straight male-to-female cable to connect it to a DCE, possibly combined with a "null modem" block or cable that had female on each end and crossed all connections.
Ugh. You are 100% right. I used to carry an RS232 breakout box in my toolkit, plus a ribbon cable with M & F DB25 connectors on both ends. Such a pain in the arse.
Eventually, you learn to be able to spot baudrate & other errors by what kind of garbage came out of the printer. I do not miss those days at all!
Nice job! A trick we used to do with old ribbons on printers to refresh them was to squirt a little bit of WD40 on them. This redistributes ink from the edges of the ribbon which never see any action, down to the "battle zone", so to speak where the pins contact the ribbon, and it extends the useful life of the ribbon. We used this as a cost-saving measure because me and my friends couldn't afford ribbons that often. All good wishes.
Watching that thing print out the list of patreon members and supporters would be a great finish to an episode.
Hi David, great work again! I used to repair similar printers for Digital Equipment in the 80s and immediately thought of an encoder problem when the printhead did not move to its left margin at power up. A word of advice: if you ever come across a dot matrix printer from DEC, do NOT remove the encoder PCB from the carriage motor unless you have the manual describing the correct encoder adjustment procedure. On those printers, you must usually power the motor with some 5V or so, hook up your scope to the photo sensors, and then physically move the block housing the photo sensors only fractions of millimeters to produce the correctly timed pulses. You need a scope and very calm hands, it's a really fiddly task.
LA120
I have to add: Wow! Considering this might be the only extant operating Bendix G15, you're doing the Lord's work! Thank you, bless you, and we fervantly hope you feel appreciated, because you certainly are! All good wishes and can't wait to see more about this project!
The TI810 was the very first dot matrix printer I worked with in 1979. I hooked the Centronics interface to our own Commodore IEEE to Centronics Parallel interface box and produced lots of output on it. I do remember how loud it was. Great to see it turn up at your place and restoring it! I always found it an awesome printer.
RS232 - the standard that isn't. 25 pins to do whatever with!
Well, there is a documented standard, but few actually follow it.
There definitely is a standard, but I'd be curious to if anyone ever actually fully implemented or used it.
Honestly, I'm surprised the corrosion wasn't more of an issue. The fact that it only caused fatal damage to a socket and flex PCB is a testament to how well those boards were made. Great work getting it going again!
This brought back memories. In the late 80's I worked with all sorts of dot matrix line (?) printers connected to mainframes. The sound is like music to my ears.
Living in The Netherlands and owner of 23 pinball machines I sure loved your 2 in the background. Good to hear your brother will take good care of it (mayby them soon). Understandably you made a wise decision from your hobby's point of view. Love it. Keep up the good work! Cheers from Amsterdam
Memories of me refilling ribbon cartridges with ink squeezed out of fountain pen cartouches...
And driving that cartridge with a drill to help distribute the ink. Saved me quite a buck back then :D
I had a TI-810. Don't remember where I got it, but I used it for quite a while and was very happy with it. It was fast, but definitely not quiet. Nice printer. I *think* it had a parallel interface. I believe there were options for the interface. I gave it up when I no longer had a source of free tractor feed paper :-) Lots of businesses (and maybe airlines?) used it for multipart carbon or carbonless forms. Very heavy duty machine.
Oh WOW! That brings back memories. I remember fixing quite a few TI 810 & 820 printers back in the early '90s. They are workhorses!
I had one of those printers in the late-'70s-early-'80s. It was a great printer.
I worked for TI also. You may want to change the ribbon as the print head requires the lubricate present ink the ink.
Say what?
I think he is saying that the print head needs lubrication to stay in working order. And also that it gets that from the particular ink used here.
I spent years fighting the RS-232 interface. A patch box was my friend! Never saw a Pin 11 handshake, I thought it was going to be xon/xoff missing as no return transmit path.
Still love the noise of a dot matrix.
Another great video. Thanks.
We had these at an auto parts store rebranded as Triad. Came with our triad point of sale system. They ran for 20+ years printing carbon copy receipts and reports daily without fail.
was a time when nearly every auto parts store had a parts system that used Omni 810's for printing recpeipts
Finishes with a listing that has code fixes dated 1977. 47 years ago... Love it.
Back about then, I was a Field Service Tech for DEC in a remote site - everything from PDP8 to 10s, LA-series printers and old teletypes - when my boss told me I didn't need an oscilloscope to fix things anymore, I moved into SOftware - I bought one of their excess inventory Tek475 scopes for $50, which is still in my lab and still working.
Ironically, an oscilloscope is probably essential to diagnosing, repairing any serious problem with a computer... unless you are okay with just throwing money at the problem (I.e. replace whatever is faulty).
At least you didn't have to repair the print head by replacing broken pins with bits of guitar strings.
I did it once on a 24-pin Oki Microline printer and it was a, uhm, satisfying nightmare :D
You mentioned JCL and a shudder went through my soul from 30 years ago.
Different kind of JCL, Centurion uses it's own home spun Job Control Language, but still definitely shudder worthy, haha.
I still have boxes of green tractor feed paper that I printed at the university. That sound really takes me back.
Strange that when Centurion customised the printer they apparently left the badge recess empty and printed their logo on the centre of the panel instead of having a badge made, the empty space is very obvious - and I see you decided to put the TI badge back which looks much better.
Usually, holding down LF while powering on a dot matrix printer will start the test print.
8 years ago i was sent to fix a printer at a logistics company. It was an epson dot matrix that printed all the slips for the drivers. When i got there there was 20 trucks backed up and two frustrated older IT guys. 26 year old me wsltzed in like nothing and was immediately turned around. They said the printer was older than me and there was no way i could fix it.
Well, they're 3x the age of the printer and they hadn't fixed it. I _HATE_ morons like that... there's no way anyone could ever know something they don't.
As a former IT bod, you never throw out a fresh set of eyes.
Usagi never fails to make me happy. Congratulations on the win 🎉
Thank you!
Carcinogens and Bad Decisions would be a good band name. Hilarious Collection Of would be their best of album title.
Yikes! I supported one of these for years and years. It was fast and loud! As far as I know the reason they are hard to find is the airlines still use them.
We used to print tickets on them, but I don't think anyone does that now?
@@Stuart-AJC That could be, it's been years. I remember buying one for our warehouse and having to drive to some Airport equipment place in Newark to pick it up! Seemed like they literally purchased everyone of them at the time as "stock".
I've seen a number of dot matrix printers used for passenger manifests but I don't believe any of them were wide format like this one. They tend to be 80s Okis
@@fcguy7okidata - I remember the name, can’t remember what they looked like. I remember when there was actually a job called “travel agent”. Their office almost always had one of these 810s. Just one example, I remember the plastic daisy wheel printers to the really messy chain printers that would burn holes in the paper when they’d jam. Had this 18” (?) wide sheet of cloth ink that was hell to change.
I supported about 20 of these, we had TI business systems and 990/4 systems in our regional offices.
The pinout is probably optimized for direct connection to an old 300/1200 baud modem that didn’t support flow control. Dispatch would call the remote site with the printer attached to the modem and dump a bunch of printouts, such as work orders or bulletins.
Wow used 2 of those printers on a TI 1500 unix system in the 90's they where bullet proof but so loud they had their own room.
I wrote code for the TI minicomputers before they added the unix os. Was the os called DX10? Can’t remember, but we used RM COBOL. Fun days.
I used to work with those machines in the airline I still work today.
They were used to print tickets, the actual airline tickets that would have 4 or 5 ways, with red carbon (anyone remember that?).
I joined the airline in 1999 and they were already old.
Surprisingly they were still being used all the way until 2007 when paper tickets were ultimately phased out and the airline adopted e-ticket.
Other Airlines would have been using other kinds of paper tickets back then, but we used the regular paper ones, they were called T.A.T. tickets.
The printer is an absolute TANK, and nothing could stop it.
The only things I ever saw breaking in those printer's were the sprocket belts, and that was it.
Great work Dave and grats to the Discord crew for the assists. Ken Romaine comes in clutch again. I love a happy ending story involving the Centurion in any way.
I had fun watching this!❤ Back in the eighties a d nineties I worked as it-engineer and had to build sooooo many serial cables...
When I taught assembly language programming long ago, one of the more satisfying programs was to read a keystroke and send it to the dot matrix printer. Everyone was hoping that tap tap tap on the keyboard would become zap zap zap on the printer. It didn't quite work like that because the old "Proprinter" was bidirectional and would queue up input characters. But in the end we found a way to make it flush its buffer and more or less print instantly as we wanted.
Paint it BLUE!!! We musts paints it blue! But seriously if it were me I would recreate it by painting it blue and then getting some vinyl printed Centurion logo recreation and have the whole system be classic and matching and everything. In any case thanks for all these great videos!
Well done the non standard rs232 usage was rampant in them thar days! 😊
The magic! YUP! The magic of making old stuff working is so great! That printer is older than all my kids, Maybe older than me. And it still works. Can we say that planned obsolescence was not in the mind of the maker of that product, on the contrary, they made sure it would last for a lifetime! Yes some component did not lasted that long. But who knew that this type of chemical reaction would happen after decades... So happy that you have that system fully working. Congrats!!
Great job David! Another great fix for another old piece. I saw a lot of those used at work….😄
What a great piece of nostalgia, love this sound! I never worked on these types of line printers, but was responsible for repairing and maintaining several dozen Okidata line printers, amongst lots of other tech, in a distribution center for a number of years.
9:41 The LM339 is a quad comparator with open collector outputs. It is not an op-amp.
It works down to its minus rail.
It takes about 1.5uS to react to its inputs
It pulls down with about 6mA when the (+) is below the (-)
It releases its output when (+) is above the (-)
THANK YOU! It's an easy mistake to make, because the symbol for an opamp is the same as a comparator is the same, but jeez, the 339 is the classic comparator!
Just want to say thank you for your great channel. I love to see that old tech getting back to life and the community spirit to solve these puzzles.
Wow! The cabling issues and printer codes transported me back to the days of using my Commodore 64 with an Epson printer. Excellent.
Very nice. It brings back old memories. I serviced many different makes and models of dot matrix printers. Specifically the centronics ACT-900 which used a similar motor to produce return feedback pulses failed. Fixed the jammed motor to make the printer to work. This happened some time in 1987
I had an 810 Ro back in the day (1985 ish) when I was in college. I had it hooked up to a Commodore 64 originally and then a 10mhz AT clone. I could not figure out the handshaking and ran it at 1200. I only had the upper case roms. Used it mostly for printing computer listings when I dialed into college (I was a computer science major). Don't get your hand inside the machine when the carriage is moving... it will mess up your fingers. Ask me how I know 🙂
4:34 holy hand grenade of a capacitor
its not a modern 'capacitor plague' afflicted thing
Yes, and at 8:20 he is really calling its bluff by counting to... gasp... 5!
I was shouting at the screen when you were saying cabling issues. RS232 is easy when you understand the different layers of it.
Each of the following are at opposite ends of the connector i.e. a end -> b end
GND -> GND, since you always need a ground reference
TXD -> RXD and RXD -> TXD so you have a data flow for the printed data, this is just data, no flow control, unless you use X-On / X-Off software flow control.
The pins meanings are Transmit Data / Receive Data. This is the absolute minimum needed for communication.
RTS -> CTS and CTS -> RTS, so you have hardware flow control. The pin meanings are Ready to Send and Cleared to Send. This is needed to slow down the sender at the right time.
DTR -> DSR + DCD at one end and then and DCD + DSR -> DTR so you have modem flow control. The pin meanings are Data Terminal Ready, Data Set Ready, Data Carrier Detect. This is for modem control, so another layer of flow control for the modem, its only relevant if you have a modem in a link.
If you need a full modem support for incoming call, then you also need Ring Indicator (RI), in this case, then connect RI -> Data Carrier Detect (DCD), since the receiving machine is being told its detecting a carrier signal. Do the opposite at the other end DCD -> RI
Things will then work properly at all layers.
When some of the other layers are missing (i.e. no modem / null modem), this is what's results in the mangled cabling that connects between the different lines above. Generally this is a bad thing, but sometimes its necessary to get the right levels on the right pins to make sure that each end thinks that the other bits of the RS232 protocol see the correct levels on the correct pins.
The problem here is that the printer doesn't have CTS/RTS, the flow control is buried away on Pin 11. 24:50 The DCD/DTR/DSR was only saying the printer was on-line, not acting as a flow control.
I worked at Texas Instruments in Cypress, Texas and my desk was about 20 feet from the 810 and 825 printhead lifecycle test room. 40 or 50 or maybe even more printheads running 24 seven. Try and imagine the racket.
That’s the stuff of nightmares! H3ll modernization that added a new circle for IT. 😱
@@ChrisSmith-tc4df I can still hear it. That was 1984.
great work, I enjoy watching your channel. as an old school dot matrix printer owner, I learned that you can spray a little wd-40 on the ink spool to get the ink darker or 'refreshed'. I used that when the printer company no longer made my print ink spools for my okidata
The sound of a dot matrix printer..... man I don't miss that one bit! 😀
That accoustic foam when new is similar to the dense PU foam in packing cases. I'm not sure if it did much for the noise back in the day
My father had a cp/m s-100 and a printer that looks just like this one, though it sounded different when it printed. (I think it was a daisy wheel rather than dot matrix.)
He spent DAYS troubleshooting that damned printer. Not just once, but again and again first for the physical connections, then for each GD program that came in.
But it was the physical connection steoy that was funniest. He spent so long on trying to get the pinout right that he had just given up and decided to get drunk instead. And whike he was thoroughly buzzed but not quite in his cups (you know that moment where you still know its a bad idea but you're fine with doing it anyway?) he decided it would be fun to throw darts at the cable. I mean, it already didnt work, whats the worst that could happen? 😅
Well, he had forgotten that the computer was still sending signal to be printed and on the third or fourth dart throw, he severed exactly the right wire and the damned printer started spitting out characters.
Random dart throwing FIXED THE PROBLEM!!
Of course, subsequent investigation made him realize he hadn't read the instructions fully and completely. But thats a different side of the story. 😂
Back in the around 2002 to 2003 a local business I worked for had 2 of these TI printers, still working and being used to print receipts for transactions. They upgraded to new printers and so the guys took the printers as well as old monitors, loaded them on a boom lift, and launched them from 40 feet in the air to the concrete below. Then the mess was gathered up and chucked in the dumpster. I guess back then, they were just considered old printers.
LOVE your channel. Nice to know there are bigger geeks in the world than me!! (Mean that in the most respectful way)
I used two of these at work One was a KSR and one was an RO. The only thing that ever happened was the hub the carriage cable was connected to the motor cracked and I had to replace that. Otherwise never a problem. Stone reliable and they ran 24/7 as a administration terminal and a call detail recording logging printer on a Nortel SL1 VLE phone system.
Makes me nostalgic for my long departed dot matrix... Nice work
modern printers still print by a matrix of dots
Now that was quite interesting.
I have forgotten mostly how serial comms operate.
And it was nice hearing the song of a dot matrix printerer...
I really feel when testing the Centurion and printer you should have been wearing a shirt and tie if not a full suit :D
Enjoyed the video and the genuine excitement when things work is great!
I fix modern computers.......... I don't have the love for the old ones....... BUT I totally appreciate and am motivated by your "don't give up" attitude with these projects.
Makes hooking up my ImageWriter II to my Power Macintosh G3 look like child's play. Love to see a community effort pay off!
That title sums up soooo many idea's I've hqd in my life lol
Wow!! My first job was testing 810's for CADO Systems......back in 1979. We swapped out the character-generator ROM because CADO had a few custom characters. I definitely recognize the internal guts. Well-built machine.
Had not heard the CADO System computer name for over 40 years ago when I was working at Warrex / Centurion. We had a Centurion dealer in Atlanta, GA that also sold CADO system ( I think for Typesetting uses ).
@@kenromaine2387 It's possible you are referring to the CAT photo-typesetter, which was not associated with CADO. Most of the systems sold by CADO were for small business accounting, payroll, inventory, purchasing, and billing. The 810 was CADO's first-supported dot-matrix printer, and was replaced by a much cheaper (and lower-quality) unit from Dataroyal.
Gotta say, the only thing better than the sweet song of a dot matrix printer is the sweet song of a daisy wheel printer.
Great work! You are definitely persistent and determined. Great to see the old printer working again. This old stuff can really challenge us.
Awesome fix! That setup looks great with the blue printer!
Worked on these TI printers back in the early 1990's when hp had just taken over the Texas Instruments printer division. Mainly board swaps and cable changes, I can imagine the customers reaction if I had gotten as excited after repair 😇. Those cable data lines were never any fun that is for certain! Sort of nice to live in a plug and play world of today but I do miss those days too.
It would be a great idea to put that printer in the bedroom and use it as an alarm clock, (printing out a weather report from NOAA every morning). 😄
Regards. 👋
Spanish Eyes are crying for the loss of that Spanish Eyes pinball, but what great detective work and repairs!
That printer looks very pretty in its new paint scheme. Like it belongs there.
This carries a scent memory for me. I can smell the fresh folded paper packs and the plastic packs the tape reels came in. 40 years and it was like the thing was sitting behind me again wobbling away on some 2 by 4s.
I once had a job calibrating and modifying these kind of printers before they were sold and that was in the nineties, so still then they were used. I also remember a cabinet below it to house the chain paper.
Oh, yea, even as late as ‘96 I had to maintain several that were used as report printers. Most had been replaced with high speed, quieter, ones capable of multi-part forms (some with automated stuffers), but these TI’s were the workhorses for general report printing.
Thank you, for sharing! Love it... Sorry we couldn't make it to the LSSM tonight.
Reverse channel. That really takes me back. I knew it existed but I never worked with it. You used to be able to get half duplex lines from the telephone company. If you were only using it for a printer or if you used polled Bisync, they worked great and they were cheaper than a full duplex line. You could used them for voice but if you didn't have control of the line, you would have a dropout. Reverse Channel was used to turn the line around so data could be sent the other way. By the 70s and 80s it was pretty well phased out but it appears that a few were still supporting it. I would have expected CTS/RTS flow control but the description for the printer clearly indicates that it's not supported.
Of course it wouldn't be easy. Printers are never easy!
Awesome vid and congrats on getting it going.
Printing always was (and still is) and ADVENTURE
Back in the day, we restored ink ribbons by spraying with WD-40 and putting in a ziploc bag overnight. Works amazingly well.
I was just thinking that ribbon could use a little WD40. (modern WD40 is not the stuff to use, 'tho.)
The ribbon is actually in very good shape for its age.
Mountains were climbed. Printers were conquered! Great video 🙂
That printing...is such a nostalgic noise 👍😀
One of the tricks we used to use back in the day to extend the life of dot matrix printer ribbons was to give them a spray WD40 and let them sit over night. This lets the ink from the top and bottom edges of the ribbon diffuse through the rest of the ribbon.
Back in the day I had an ancient PC with a monochrome display and an old dot-matrix form feed printer that I had poached from various different people's trash. I somehow managed to get the serial port working so that I could transfer GW-BASIC from a laptop onto the PC, then I went digging through the printer's manual and started experimenting with it. I was able to put the printer in "high-density" graphics mode, at which point I could send the printer arbitrary bytes to drive all of the pins in its column independently. The ribbon was pretty weak, so I compensated by overdriving each pin 4x to make the image darker before moving the carriage, but I was able to successfully print monochrome images. I suppose if I varied how many times I drove each pin, I could have even printed different shades of gray!
Those were fun times. I no longer have the equipment but I should go digging through my old stuff and see if I still have a few of the sheets that I printed.
9:34 the LM339 is a quad comparator with open-collector outputs, note that it's used with positive feedback (for hysteresis) rather than negative feedback that you'd typically apply to an op-amp.
Also notice that they are using the forward drop of the LEDs to bias the (-) pin of the comparators. This is a bit clever as it saves some resistors.
I was a Static RAM Product Engineer for the AM 9101 at AMD in the 80's. The 9101 was pin compatible with Intel's 2101. It is a 256 by 4 bit RAM. Most likely a buffer for the serial input. Imagine having to design with 1K RAM chips today!
For future reference, using white vinegar to neutralise the battery leakage helps.
With apologies to JFK: "We choose to go to the Moon in this decade and do the other things, not because they are easy, but because we _thought_ they would be easy." 🤣
That connector was a damned mess. Great hard wire repair. Feel your pain with the crazy comms wiring.
that felt so good. kick ass work and community!!
I bought an NEC Spinwriter to use with my Compupro 8/16 for similar reasons
Those TO-220 transistors with the solid metal backs (the TIP150s at around 6:45)... when my company was still servicing intercom boards for a major brand, we were told to watch for these and replace when found. Never really found out why that directive.
Fantastic series. Well done. Would like to see how you would tackle re-inking the ribbon. 👍