IBM System/360 Front Panel

I recently acquired this front panel from an IBM 360 mod 50, with lots of switches and over 250 blinkenlights. The IBM System/360 was an iconic mainframe computer introduced in 1964 (see promotional video here: • Computer History: IBM ... ). The system was a huge development gamble for IBM, with all the "models" from small to huge sharing mostly the same overall architecture and instruction set, but with radically different hardware implementations depending on the cost and performance point. A key novelty was the introduction of "SLT" ceramic hybrid modules, pre-dating the integrated circuit, but with denser packaging than individual transistors. This front panel is from a model 50 (in IBM speak a "mod 50"), a larger business system. And yes, I'll take you through most of the buttons. Take this with a grain of salt, this is very complicated and I am new at it. Understanding the meaning of every one of the 1,152 bits that the lights and rollers can display will take some serious further study. I am sure the IBM CEs (Customer Engineers) that spent their careers on these panels will chime in the comments section and correct all my mistakes.
Try Ken Shirriff's emulation of a running 360 driving this panel:
static.righto.com/360/
Ken Shirriff's article about the 360:
www.righto.com/2022/01/ibm360...
Chapters
00:00 Panel unboxing
00:53 First look, front side
02:28 First look, back side
04:17 Repairs
06:52 Every part of the panel explained!
07:43 Power panel
09:00 Memory panel
09:49 Emergency pull
10:11 Auxilaliary memory power panel
11:05 Channel test panel
11:34 Main indicator panel
12:47 Data and address entry panel
14:21 Error lights panel
15:41 Debug / program flow control switches
17:51 Time counters
18:22 Manual operation and debug panel
20:50 FLT control (firmware test routines) panel
21:51 Start/boot (IPL) panel
23:02 Overall view, flipping rollers, future work
Our sponsor for PCBs: www.pcbway.com
Support the team on Patreon: / curiousmarc
Merch on Teespring: teespring.com/stores/curiousm...
Learn more on companion site: www.curiousmarc.com
Contact info: kzbin.infoa...

[@Journeyman-Fixit]

Night shift computer operator 1976 Cleveland, Ohio. Worked for CMI (Computer Management Incorporated) from 1976 to 1980. We had an IBM 360/20 and a IBM 360/40 - Man they were workhorses, we worked two shifts X 80 hour weeks at that service bureau taking any job we could to be profitable. Finally after a few years we got some traction with SC Johnson company and the Department of the Navy. Those were the days, I'll never forget there was one job I thought was going to do us in it was: 500,000 cards to tape, my hands still hurt when I think about it today. We had two 1403 N1 printers that every once in a while someone would leave a stack of tapes on top of one of them, when the printer ran out of paper the top would automatically raise up and the tapes would fall off smashing them on the floor! Thanks for uploading! 😊

[@pilgrimm23]

a joke ONLY an IBM Operator would get: How can you tell a mail dataset from a femaie? //SYSIN DD PGM=IEBGENDER

[@measl]

*I worked on the 360s,and then the 370s, from 1965 through 1978 (with a few System 3s thrown in for fun), before moving on to Data General C/330s and MV8000s in the early 80s. What I loved about the panel based computers is that everyone was terrified of touching them: the CE would show the operator what to do to get started,and that was it for most people. I worked overnights most of the time, and to amuse myself while the humongous printouts happened, I would examine different memory locations - when the day shift people would come in they'd be all freaked out that the rollers/switches weren't the same as when they left! Good Times!*

[@gsp49]

Yes, we would be up all night calculating stuff that would take less than a minute today.

[@James_Knott]

I used to be a computer tech and worked on Data General Nova & Eclipse computers, along with DEC PDP-8, PDP-11, VAX 11/780, various PR!ME models and Collins 8500 computers. the Collins was a mil spec version of an IBM system. The 2nd front panel I worked on was for my IMSAI 8080 computer, which was the first computer I owned. The first was on an ancient Teleregister system, located in the Toronto Stock Exchange. It used vacuum tubes and relays!

[@measl]

@@James_Knott *The IMSAI was a **_gorgeous_** machine, but I was no fan of the S100 bus. My first homebrew was an ALTAIR 6800, but with a pair of 6502 CPUs on a modified SS-50 bus (I used the inverted clock for the second CPU, so each processor was unaware of the other, and they shared physical resources). I worked on the Eclipse lines a LOT - I even wrote their first (and probably only, since the MV8000 replaced them with a *nix like system) multitasked Print Spooler. I have the service panel from the last S/370 I worked at (at Blue Cross in NYC) hanging on the wall behind me (I'd send a pic if they'd allow it). I worked on the Muppets with Jim Henson on a S/32, and I write the first Dental Practice system (including a universal insurer interface) on a Micronova 2! Damn I miss those days!* *I wanted to work on the VAX 11/780 so bad I could taste it, but the VAX owners were snobs, and once they knew you were a fan of either DG or Onyx (remember **_them???), your interview was over! I tried so hard that I actually interviewed for a job o a 780 that was a 2.5 hour drive, in a perfume manufacturer (as an asthmatic, that place would probably have killed me, but that would have been a small price to pay at that time). Alas, it didn't happen until I saw a 780 being [literally] tossed in the trash (lord knows what they upgraded to that they felt a multicabinet 11/780 was only worthy of the dump) - I grabbed it before the trash people could, and had it running in my home about a month later. That thing ate more power than a C/330 with a pair of Zebra drives and a bank of 1600bpi vacuum drives all by itself: the power company was quite impressed with the bills they were sending me for those few months: I was making _really_** good money for the 1980s (over $200K/yr before taxes, and I had a hard time keeping up with the power bills (not to mention the 4 tons of A/C it needed as well)!*

[@James_Knott]

@@measl I know the S-100 wasn't the greatest, but it was improved for the IEEE spec. In some ways, it was better than the IBM PC bus, for example level triggered rather than edge triggered interrupts in the PC. I remember the Onyx name, but never saw one. I used to work for a major telecom and the computers I worked on were for message switching, including telegrams. The Collins computers were for the Air Canada reservation system. On the Eclipse computers, mostly S/130, I would even get into the microcode, when trouble shooting. Incidentally, I can thank the VAX for my first modem. One day my wife visited me at my office and I showed her the Adventure game.on the VAX. She wanted to know if she could play it on my IMSAI. I said no, but if we had a modem... I was soon the proud owner of a 300B manual modem. 🙂 I realized the end was near for that sort of computer when I read the new Intel 80386 CPU was as powerful as the VAX CPU. I then moved into planning, where I planned the installation of telecom equipment in the company's main office and customer sites in downtown Toronto. Incidentally, when I was a tech working on the Eclipse computers, I hand wired a couple of Ethernet controllers on prototyping boards for the Eclipse. At the time we had a DECnet LAN connecting the VAXes and the programmers wanted to connect some Eclipses to it.

[@measl]

@@James_Knott I agree. The IBM PC took the industry backwards by almost a decade. Every single thing about it reflects their design philosophy: make it fast, make it cheap, and don't worry about compatibility. Our projections show we won't even sell a half million of these, so just get it done as fast and dirty as possible.* *They succeeded. It was a **_terrible_** design, from start to finish. What made that so sad was that it came from the first company to understand that both forward and backward compatibility was the key to success. They abandoned every one of the philosophical positions that made IBM the powerhouse that it was - really, it marked the death of IBM as a serious player in their own field. The Compaq 386, as awful as that design was, it blew the doors off the 370/168 (the biggest, baddest, sexiest machine IBM ever made for the civilian market)*

[@pentlandite3651]

As a student, I was a 360/50 operator at the University of Waterloo in 1971, and your outstanding presentation sure brought back memories. The UoW machine had 512KB of fast memory installed, but a portion of it never worked reliably, so the IBM CEs disabled part of it to the 384KB that did work. This machine also had 1MB of slow (LCS) memory that you mention. The 360/50 was mostly used to provide APL ("A Programming Language") services for around eighty 2741 terminals, all connected via a 3705 communications controller. Thus the 360/50 was normally running round the clock except on Friday's when the IBM CEs did their PM. That's when I got to IPL the machine and then bring up APL as a job. The OS job dispatcher loaded jobs in memory starting in LCS, which was too slow for running APL. So I had to run a dummy job aptly named 'Corefill' which just hogged all of LCS but did nothing, then load APL which would now load into 'fast' memory', then cancel 'Corefill' on the 1052 console. There was also fun stuff like mounting tapes on the 2400 drives, mounting the 2314 removable disk spindles, and hauling boxes of green-bar paper. The 'infamous' red 'do not pull' switch was a popular place to hang surplus rubber bands that you would need for keeping the various card decks together...not a good place to store them!

[@JiveDadson]

I drove an IBM 360/44 at the University of Houston in 1971/72. It had core main memory, but a small amount of that new fangled transistor memory in the console. It also had a drive for those enormous old floppy disks, but no-one used it. I was given the job of system programmer too. FORTRAN IV, learned from the IBM manuals. There were no CS courses of course. I remember a few of those controls, but not many.

[@lawrencedoliveiro9104]

“Corefill” ... I had to do something similar on an IBM 5170 in the mid-to-late 1980s. This was one of the early ones to have its full 640KiB of RAM, and the WordStar installer was doing a signed comparison on available RAM, and deciding that there wasn’t enough for it to install. To fix, this, I created a RAM disk which reduced the amount of free RAM to below 512KiB. Then WordStar would install. I guess some things about IBM never really change ...

[@gorillaau]

What did these system cost?

[@kylecoulombe2755]

pentlandite hey my high school computer science teacher went to uwaterloo in the 70’s and used the 360/50! He gave me a punch card from one of his programs for it when I graduated

[@BrunswickHeadsRevisted]

We had two 44s. Hard-coded rather than microcode and a reduced instruction set (e.g. no storage-to-storage instructions). Though you could load the auxiliary storage with an emulator when you wanted to use all the 360 instructions when doing offline processing.

[@RussHolsclaw]

Well, I'm a retired IBMer who was once a CE trained to maintain the 360 Model 50 CPU's. I enjoyed your video on the front panel. I've got quite a few comments and information I can pass along about this machine, and its front panel. I hardly know where to start, so I'll start by writing a few things offline... so I can compose my thoughts, and then post them here as I complete different parts of them. I dare say there's a lot of information that has been denied to you due to the fact that the wires were cut out behind the panel ... they would have informed you in areas you could only speculate about. (Actually a half-century later, I'm delighted to realize that I remember a lot more than I might have expected... so watch this space over the next few days.

[@Mrjbroo11]

Anything?

[@johncrunk8038]

I, too, am an ex-IBMer and worked on most of the 360's -- 30,40,45,50,65 and 370's. He will need to install a new power panel to do a lamp test on that beast! I remember that the 360/75 would trip a breaker if the lamp test button was pressed.

[@blackbird8632]

1 year ago...

[@MatthewPherigo]

Would've loved to hear what you had to say!

[@satan3090]

Are you an Alzheimer's patient?

[@johnopalko5223]

When I was a student at the University of Illinois in 1972, I had a part-time job as a 360/75 operator. Now *that* was an impressive front panel. I've always had an especial fondness for the IBM 360 mainframes. My phone has more processing power but they sure looked sweet.

[@KameraShy]

Probably the same machine I ran programs on there in '72. You may have run my program.

[@bayareapianist]

As a sophomore EE Student, I took a Fortran course in Iran in 1986. I never saw how this computer looked like. All I had to do was to punch some of my 100 ration cards, add some special cards and give them to student behind the counter. Then the day after I would get the result of my program! I had to choose my variables and goto addresses carefully so that I could reuse my cards, otherwise I would finish my cards before the end of term. I also got some used card from a person who had taken the course before me and named my variables like his. Making the story short, I immigrated to Canada in 1988 and I start my first job at a computer store. I loved how fast i could type on an IBM PC compatible and see the results immediately. After a couple months I got laid off. Then a few months after I got a job at IBM. I felt I had landed on the moon specifically my pay and over time that I was making. Now after over 35 years, I am a part of history!

[@cygil1]

Submitting jobs on punch card, that's old school, Iran was very behind the times (unsurprisingly). Multiuser terminal systems were standard by the mid 1970s in the west.

[@jerry3890]

The CE key also enabled some of the function switches on the front panel. The power section could not bias voltages without that key turned on. Most customers would run the computer with the check control switch in stop. If there was a hardware error the check light would come on and the system would freeze. The operator would then push the logout button, which dumps a number of internal registers to main storage. The operator would then load (from cards) a program called SEREP (System Error Record Editing Program) which would print a formatted dump of the errors on the system printer (the 1403). Teach operators Error light on, press logout, run SEREP. I serviced mostly the 65. Both the 65 and the 40 were rock solid, I always did feel that the 50 was a bit flaky.

[@HebaruSan]

I think I understand now why command line parameters are sometimes called "switches."

[@paullitwinas]

I was a CE who repaired the entire 360 line.This brought back memories from the 70's working out of NY Brokerage office in the Wall street area of lower manhattan.My favorite system was the 360/65.It had 1 meg of core storage.I remember useing a hammer to tap on arays to clear dust put of them while running the "3A1" diagnostic....thanx for the video.

[@cyprusman5908]

This bought back some memories ! We had FOUR 360/65's and a Model 360/40 at BOAC (now BA) in 1971 when I started as a junior computer operator. They were the great days of computing, and I loved every minute of it !

[@alancordwell9759]

Fantastic! And thanks for the walkthrough, very interesting and informative. Looking forward to any developments on it!

[@Platypi007]

My grandfather was an engineer for IBM back in the late 40s until the 80s. Not sure what all systems he worked on but I often wonder when I'm watching videos like this if he ever worked on the particular system in the video. :)

[@geoffcrisp7225]

We had three 360/50's on the floor at Centre-File in The City of London in the late 60's then upgraded them to 370/155's running OS/ ASP. It was a large bureau operation running remote systems for Stockbrokers and Building Soci plus Payroll. Lots of Peripherals and strings of disks.

[@ComputerHistoryArchivesProject]

Outstanding video! Best view of the control console that I have seen. Excellent clarity and just fascinating to watch your unboxing and exploration of the switches behind the panels. Thanks for posting this! ~ CHAP

[@CuriousMarc]

Thanks for the compliment! I watch your period videos quite often, you are doing a great service to computing history by preserving and posting these.

[@johnrickard8512]

I'm going to keep an eye on this one. I've always been fascinated by early computing tech, and have always wanted to see the almighty s/360 in action! I guess it helps that it's before my time. Regardless of my reasons, you sir, have gained a subscriber ;).

[@pilgrimm23]

Awesome! I was a operator on a Dual IBM 360/65 system Shared DASD, Shared Spool running the NIH variant of HASP. That was in the late 1970s in Albuquerque NM at the University of NM. We had a lot of surplus hardware from America Corporations. This brought back memories. I recall the last time I saw a face plate was the Knoxville TN World's Fair American pavilion. I reached over the red ribbon and put it into ROS READ mode. for old time sake.

[@PaulRiecke]

I also worked on a shared 50/65 using HASP. We had four banks of tape, three banks of HD, two printers, card reader, card reader/punch, and a 9,000 reel tape library at HQMC. After setting up the shared system we had quite a few visits by the CEs working out the bugs. Our experience there was a good intro into what we work with now.

[@BrandonNedwek]

I've wanted a walk through of one of these panels for a long time. Thanks for posting!

[@luthmhor]

I wish I understood any of this, I watch all your videos but it's all gibberish to me, I still find it fascinating. Your passion is infectious! Greetings from Canada!

[@Alan_UK]

An enjoyable reminded of the start of my career. I started as a trainee programmer for a London bank. The bank had a number of model 50's but also had model 30's as backup and a rule was that all programs had to be capable of running on the 30's in an emergency. They also had a room full of 20's running TOS (Tape Operating System) rather than DOS (Disk Operating System). These 20s were solely for running cheque sorters and team of girls (the bank was very sexist) completed for achieving the most throughput each day. So there were dark rooms of the model 30s and I would often go into one, switch on the lights and power up the machine setting the booting dials to the hard disk with the System 360 OS. A very Starship Enterprise feeling. One day testing didn't go well and I was so frustrated I pressed the power off (or did I press the big red button!). There was a huge bang and I thought I had broken the machine and would loose my job and career. I have forgotten to power down the disk drives and I guess they simultaneously retracted the heads and put the emergency brakes on. I removed the disks and slip out turning the lights off and telling nobody! I soon moved on to another organisation implementing the first European IMS DB/DC system with an incredibly complex database. We had an IBM systems engineer assign to the team and one day he came in with a big grin on his face. He had persuaded IBM to let him have a model 50 front panel like yours. He would put it in his living room along with his collection of pin ball machines. He said he loved the tactile nature of the model 50 front panel. Enjoy your front panel!

[@KameraShy]

In the mid-80's a location of the company I worked for was still running its business on a 360 with TOS - probably the model 20. In a weekend they cut over to a 4331 (I think). Shop was so well run it went flawlessly.

[@doalwa]

I wouldn't know what to do with, I still want one :-) all those switches and lights... blinkenlights for days :-)

[@Loquax1975]

This was really interesting! :) When I was a kid, I kept reading the "computer" article in our encyclopedia. It had an image of a 360 and I so much wanted to get to know that and understand all those flashing lights!!!

[@jms019]

Thank you for going into so much detail and good luck with bringing it to life !

[@brucerosner3547]

I worked as a systems programmer at IBM Poughkeepsie in the 60's. I was something of a maverick as I preferred colored over white shirts so they put me on the night shift. At night I had an entire 360 to myself. I still remember single stepping through code using the front panel switches and lights to debug. Hard to believe today but 4 kilobytes was considered a lot of memory then.

[@PauldeSwardt]

I trained on one of these at CDC (Control Data Corporation) just off Oxford Street in Central London(UK) (1980). We had a 360 up in the main computer centre in East Barnet - which as the story went had been dropped into the harbour of a South American country were the correct bribes had not been paid. Trainee engineers were given the recovered 360 - to clean it up (which could not be re-sold as the insurance had already been paid). So from time-to-time it would just stop. There was a procedure to instruction step to the next step using the toggle switches shown around 13-19mins. There was a 50/50 chance that this would fix the bug - and normal service would be resumed! -

[@hermassi1234]

Paul de Swardt i

[@dentrecasteaux1523]

I worked as a CE for IBM in London in the 70's. I remember working on a nasty bug on that Mod 50 on Oxford St !!

[@AureliusR]

Why would CDC have an IBM mainframe? They made their own machines.

[@Bhakti-rider]

There was a component of IBM named Service Bureau Corp. (SBC). As a result of some lawsuit, IBM had to transfer (for a price, I believe) SBC to CDC. I worked at that element of CDC in the 80s; we had MVS/HASP, running on 370s. @@AureliusR

[@luisluiscunha]

Thanks a lot. I really liked the guided tour. In the end, it helped me to understand what an impact something as an Altair 8800 had in people who were used to see those panels for computers, and what it cost to do those operations. Imagine being able to have the equivalent "fun" in a system you could afford, as a person. That also says a lot about the "Personal Computer revolution". Thank you, again. This was great.

[@James_Knott]

My first computer was an IMSAI 8080, which was a better quality Altair clone.

[@WaltonPete]

It looks great! Surprisingly, it matches the decor remarkably well.

[@dragonheadthing]

A very nice explanation of the front panel. Thank you for the video!

[@bborkzilla]

My first experience with big iron was an IBM 370. I would totally reanimate a front panel from one of those beasts if I ever came across one!

[@citizen240]

Like seeing an old friend after a long time! Sigh, brings a tear to me Eye.

[@jyeager2881]

Remember setting load address with the dials to the card reader 1501 ??? then pressing the load button to IPL the system. Was fun watching the card deck read through the machine!

[@lionlinux]

I had read much about this machine and haven't seen it yet .. Thank a lot to show it!

[@zzz13zzz17]

Hey, I was work with soviet clone of this called ЕС ЭВМ in 1990-92 as operator and programmer. Vms, rexx, pl/1. That was great time.

[@BrunswickHeadsRevisted]

I visited the facility that used to make the clones on the outskirts of Moscow. I think it's occupied by NICEVT these days (or at least in 2003).

[@kevincozens6837]

I had run programs on an IBM 360 and 370 back in the day but I never got to see the actual front panel. It really is incredibly complicated. I noticed the book on display beside the front panel. I have that same book in my collection from the days I did a bit of IBM 360/360 assembly language programming.

[@James_Knott]

I don't think I ever saw a 360. but back in the early 70s, when I worked as telecom technician at the Toronto Stock Exchange, I regularly saw a 370.

[@fredblonder7850]

Regarding the “Emergency Pull” on the 360 and 370: As it was explained to me, you pull it if there are flames coming from the machine. You do not pull it if there is only smoke coming from the machine. This would result in an expensive service-call to IBM, so it was used only in dire circumstances. Actually, I never heard of anyone using it.

[@jasonmurawski5877]

Fred Blonder what exactly did it do? It seems like you would pull it in an emergency, but what does it do?

[@fredblonder7850]

@@jasonmurawski5877 I’ve never spoken with anyone who actually pulled it, so I don’t know. For all I know, it could play music, have a Jack-in-the-Box pop-up and throw confetti. I assume it shut-down the power-supply and shorted out a few bus-bars and possibly released some chemical fire-suppressant.

[@normhoback5545]

I was also taught to use the emergency pull IF I was certain a CE was getting electrocuted. I don't know what the 360's Emergency Pull really did, but in in the Amdahl 308x look alikes, the EP switch was attached to a toggle switch under the console that was a main disconnect for all power. Actually easy to reset, the CE just took the skin off the console after the issue was fixed and flipped the toggle back on.

[@LickorishAllsorts]

When a 360 20/25 was scrapped I got to use the BRB as a final goodbye. The machine room itself had power kill switches every fifteen feet or so around the ouside wall - got to use one of those when a transformer in a brand new 370 actually blew up, small copper beads everywhere, one hell of a panic and the whole machine room down for a day!

[@Bhakti-rider]

When I started as an operator trainee in 1969, what was said about Emergency Pull was that if we went to war, you pulled it.

[@colinstu]

can't wait to see progress updates on this!

[@kurtdobson]

Very cool... The very first computer I learned to program was the IBM 7090 when I was a cub-scout and in 8th grade in Fortran. My first job was a software engineer in Advanced Development at Sperry Univac and we had a mainframe series 9300 which had a micro-coded instruction set. My group wrote an IBM 360 emulator and we were able to boot the 9300 as an IBM 360, then load and run IBM native programs. This is what inspired my career. Thanks for sharing the good old days.

[@rscgln]

Sorry, no! The 9300 was not microprogrammed as they were not microprogrammed the 9200 and the 9400. The first Sperry Univac, IBM-Compatible mainframe microprogrammed computer was the 90/30, followed by 90/25 and 90/40 (same hardware, dummy micro-instructions to slow down the 90/25, 500nS clock and microcode prefetch for the 90/40). I started to work on it in 1977 as a CE but I think it was released in 1974 or so. By the way, the 90/30 already had the 360 mode of operation support in its microcode that allowed to run native IBM code.

[@JGunlimited]

Those switches are so satisfying

[@judmcc]

I'd love to have one of these. The first computer I used was an IBM 360/65.

[@MrGeocidal]

I used an IBM System/360 (remotely) from 2015 to 2019. That particular unit is still oporating to this day.

[@mfkman]

The ability to set breakpoints as well as data access breakpoints impressed me.

[@mmadmic]

About the size of the box, it remind me a funny anecdote. In the early 90s, I was a teenager and went to an electronic market, they were discounting all their old stuffs and my friend saw a box with a few cables and what seemed to be phone accessories, he bet 50 Belgian francs (+- 1.5€) and won the auction, but he didn't carrefuly read and the auction was not for a few telephone cables but for a full Burroughs mainframe computer from the 60s, with terminals and everything not working (of course), and he had to carry all this garbage to his parents home ... you'd image his mother's face when she discovered it. It was quite fun. Personnaly, I bought an IBM printer for arround 5 bucks but it never ever worked neitheir on my PC nor my other computers, it was an EBCDIC machine.

[@KameraShy]

Very interesting! I always wondered what all those lights, switches and buttons were for. I wrote my first program in Fortran in 1968 at the University of Illinois on one of those machines. Of course, the closest we ever got was the check-on window where we deposited our little deck of punch cards and picked up the output the next day. Never saw the inside of their inner sanctum. Had to wait to touch one until later in the 70's when the 370's had arrived.

[@James_Knott]

Back when I was in grade 12, I had a Fortran class and we used pencil mark cards, which the teacher took to the school board office, to run on a computer there.

[@jims6056]

Wow that is a piece of history right there!

[@pondbearflyer1193]

Wow, nice job, very impressive, Bravo.

[@GH-oi2jf]

This is so old school. I worked with Control Data mainframes of the same era. The console had no switches or lights, only two CRTs and a keyboard.

[@gmac1109]

Wow! Where did you find this panel? I'm a former IBM CE from Brazil, and I wish I had one of these in my living room... Model 40 had the same kind of toggle switches as the 50, they're beautiful. The 50 was easily recognizable by the voltmeter.

[@ipekshev]

Thanks for an amazing video!

[@knightwatchman]

My first job as a mainframe operator was on a 360/40. The IBM CE(s) always told me, "Never touch those switches and dials!".

[@VileStorms]

these are way more complex than modern computers, people must have been smarter in the 60's because i have no clue what the heck this does even after you explained it. I'm just a dumb kid lol you did a good job, its just above my head.

[@allartvogelesang316]

Thank you! I learned to program on a 360/50 at Simon Fraser University in BC, Canada.

[@allartvogelesang316]

I wish I could post a picture of me at the console.

[@Ashjuk]

Wow! I remember constructing those stacking rotary switches back in the 1960s when I was an apprentice - not seen one for years.

[@CuriousMarc]

Most of mine are broken. Do you know who was the manufacturer? How to get new ones? Repair them?

[@Ashjuk]

@@CuriousMarc I really wish I could help you, sadly not. I did my apprenticeship with the Ministry of Defence so anything we used would have come from central stores and just be marked with a MoD part number. Pretty much all I can recall is that they came as a kit of parts with the base plate, wafers, rods and spacers. You used as many wafers as needed and then cut the rods to suit. I have just Googled for rotary wafer switches but the modern ones don't look like the ones I used at the time. I suspect the only source now would be from vintage equipment or surplus sales but as we are talking the 1960s they are probably thin on the ground. Is it just the wafers that are broken? If I recall the wafers were made from a material like Tufnol so have you not considered making some new ones yourself? A long task I know but possibly the only way you will be able to restore the panel. And thank you for a fascinating and enthralling channel - it's great to see this old equipment being put to use again.

[@CuriousMarc]

@@Ashjuk Well thanks anyhow, really appreciate your answer!

[@DihelsonMendonca]

I prefer that beautiful grand piano on your room !

[@mikemcgonegal1616]

I operated a 360/50 in the early '70s. Other than the power switch, I can't say that I ever touched any of those gazillion others. Didn't even know what they were for. I remember, though, that it was considered really bad form to pull out the big red emergency power switch.

[@GATMachine]

How about a CuriousMarc Q & A? You're an interesting guy. What's your story? How do you know so much about all this tec? How do you make a living? Is it your home/Lab in the videos? I could think of many more questions.

[@elen5871]

dude's been the CTO of like a bajillion tech companies, he is fluuuuush with caaaaaaaaash, guessing it's his house in most of these

[@Lazarus0357]

I started in computers working as a 360 operator, then moved up and after a while ended as shift supervisor. One fine day I moved to programming (on a 360 also) and could sleep nights again. But bugger if I remember what all the switches are for! Regards

[@Lindsay5137]

Well Lazarus, as someone who worked on 360/30s plus a few other CPUs (as a CE) then 15yrs as software support I can't remember many (if any) programmers who knew what ANY of the switches were for - so don't feel bad about not remembering ALL of them.

[@Lazarus0357]

Thanks mate! But I don't feel bad about not remembering details like the meaning of all those switches. It was at the start of my career, one of many steps upward. Regards

[@Lazarus0357]

Yes, now I remember about using the LOAD button when I was an operator. Regards

[@mytouch-id2407]

9D00000C 47700300 9C00000C 47F00300 And to make things more realistic be sure to blink the Master Check light.

[@BixbyConsequence]

Those were the days! Most programmers don't sleep nights anymore, sadly.

[@sumonarosner4674]

I worked as a programmer in IBM Poughkeepsie in the late 60's on the night shift. Sometimes I would force a red light to get a field service friend to bring in a pizza. Some interesting challenges at that time were the shortest stable program you could enter with switches and playing a tune through AM radio from memory emission noise. The infamous HCF - Halt and Catch Fire instruction comes from a parody of the assembly instruction cheat sheet that every programmer carried with them.

[@measl]

*Just a note about the HCF instruction. But first, a nomenclature correction. Those green (360) and yellow (370) reference cards you referred to as a "cheat sheet" was almost universally referred to as "The Bible" by the programmers and operators, since without it, you were in deep trouble if things went haywire and you landed a 4 inch core dump for dinner!* *As for the HCF instruction itself, that did NOT come from the IBM world: it was a REAL instruction, implemented on the late versions of the Motorola 6800 - in response to a request from the air force. The Air Force had decided to use the 6800 in some new avionics package, and the request for HCF was made because of the fear that this new tech might fall into the wrong hands in the same way the U-S tech did. I have a LOT of 6800 experience, and have personally tested the HCF instruction: out of 20 tries, about a half dozen of them halted with a big "loud, pop. No real fire, but the chip was genuinely unrecoverable. The HCF instruction became public by accident: either Byte or Popular Electronics (probably Byte, but it's been so long that I can't remember any more), decided to see what would happen if they tried the "unused opcodes". They clearly had some of the late model 6800's, because their testing resulted in the famous popping sound, along with the smell of freshly seared silicon."* *Since it became public knowledge, I have seen it attributed to all kinds of crazy things, from originating in the 360 architecture, to a "special instruction" that required all kinds of crazy prelude code to enable it before it would work. I've seen it attributed to almost every silicon based platform, and I've heard it described as the cause of all manner of awful mishaps (from causing the Challenger disaster to being a Russian planted change to the lithographic masters for the platform(s) of the day. HCF has taken on a life of it's own.*

[@thisman1906]

Merci Marc !

[@karlramberg]

So, will you implement an intermittent faulty CPU emulator ? :-D

[@CuriousMarc]

I'll have to remember to put that feature in :-)

[@karlramberg]

Route the data and address buses trough a radiation chamber so bits get switched at random for a realistic experience :-)

[@jerry3890]

Run it in check stop and when the red light comes on press "logout" and run SEREP

[@johnlister]

Every time I see photos like this I have to ask: did anyone ever work at a place where the panels weren’t blue? All the IBM promo literature had red panels, but the default was blue (though there were feature codes for a variety of colo(u)rs including green and yellow if I remember correctly. But in the five datacenters I worked in, everything was blue, until it all went white with the 3081 and successors.

[@Boemel]

This is awesome, i just got a 3750 :D Mine is still wired up and working. I am looking for replacement Lamps and Indicators, do you have an idea ? I find none on ebay.

[@macieksoft]

Are you gonna wire it to some emulator? Wish we had S360 or S370 emulator that simulates the control panels. Would be fun to be able to input instructions by the switches and then watching what happens when CPU get started.

[@gsp49]

I used to work on a 370 mainframe, every day, for years. One of the first things I was taught, the blinking lights did not mean anything.

[@johnlister]

I’m guessing you were an operator. I was a systems programmer for the UK bank and we could book one of the machines for testing/debugging. I have used the single step switch to try to find an error in a systems program. Do you remember the console on a /370? Instead of a raster screen, it used something that drew each character using up/down left/right changes to the electron beam on a CRT that had a long decay time. Then the 303X machines came with an auxiliary processor to run the machine which had a regular raster display. More powerful and easier to use, but no more blinkenlights!

[@kathyquinlan5922]

I am slowly catching up on your videos after seeing the on with the IBM (forgot the model number) Valve based debounce circuit. Is this project finished or not started ? If it is not started can we the internet help ? (I have made a fair few props for TV mini series for kids here in Perth Au that have flashing lights etc.) I would love to help build a working IBM 360 CPU emulator (one that could actually run real code) Funnily enough I just recieved a Xilinx Arty Dev board from Santa ;) to learn how to write Verilog on (I have been meaning to learn for years now, but uC' just keep getting more powerful and keep taking jobs that used to be the domain of FPGA's ) Regards, Kat.

[@CuriousMarc]

This is a long term project, it will be measured in years. Not started yet besides light fixing of the front panels and removing the bad foam from the tapes. Plus I need to get a 3 phase installation just to start...

[@dimbulb23]

I was an IBM CE, started in '64. I worked at and around a large 360 model 65 installation at an major airline. I was new and never worked on the 65s (I think there were six at this airline customer). The only time I actually worked on the 65 was fixing its console, an IBM Selectric I/O. In my 30+ years at IBM, I never saw anything more impressive than that computer room. I agree with what's been said here about the practically of owning one of these monsters. I know a few guys who could probably get it working but for what? It's place is in a museum someplace.

[@CuriousMarc]

Let's see. We have two IBM 1401 running at CHM, there is a 360/20 running and a 360/40 almost running at the LCM, Carl has two running IBM 1130 (a real one and a gate for gate FPGA replica), his friend has a running IBM 1800 in Finland, there are two working IBM 360 FPGA gate for gate replicas with original front panels, a /65 and a /30. And we are seriously considering helping with an IBM 7090, but that one will require a very large team and budget. Don't know of a working 360/50, hope this will be the one. Multi year, team projects. Much better than just using panels as static decoartion, if you ask me.

[@dimbulb23]

I agree, getting it working is worthy goal. Good luck.

[@Ashene64]

"It belongs in a museum"

[@AK-vx4dy]

Protect is for "keys" to memory pages (some sort of memory protection)- i read about it book about 360 assembly

[@johnlister]

IBM has had memory protection keys since I don’t know when. Areas of memory could be assigned keys and there was a 4 bit key associated with each block of memory. (2K or 4K depending on the model/settings). If I remember correctly, there were also another control bit associated with the key. In the Program Status Word for the running program, there was also a key. This had to match the key of the storage to allow any writes. The other bit allowed read-only access to the storage or not. (IBM’s operating systems had pointers from low storage locations to lots of routines that programs could use, hence the access but not modify status)

[@frankkoslowski6917]

Very nice demo. Although a Debouncer using valve technology might seem almost ludicrous.

[@robertgilhooly341]

I was a final test technician on this machine in the late 60s. Traveled all around the country fixing them. The full system had 512k of core memory and cost $1M. Even the most simple technology today like a smart thermostat in you house is by far more powerful than this machine but at the time it was the height of technology. In those days when one failed the field technician would spend a day trying to fix it and then call the area tech who would spend a day and then I would go out to fix it on the third day. The field techs were plenty smart but they worked on a lot of different machines. I only did the 360-50 so I knew all the quirks. If a mainframe today went down for several minutes it would be a major disaster.

[@CuriousMarc]

Awesome, thanks for the note! Do you have any documentation for the 360/50? We have the hell of a hard time to find a copy of any ALDs. We have some for the /40, the /65, but not the /50. You can contact me by going on my channel about page.

[@chrisbragdon5901]

These machines had a weird love hate relationship with data center operators back in the day. “Love”: Mr. Manager strolling down the hall with a future client to the big fish bowl windows where all the system operations were proudly displayed. CPU lights ablinking, tapes ataping, printers aprinting, folks running around feeding printer paper, tab cards, mounting tapes, swapping disk packs in a smooth, orchestrated symphony of...”Hate” when it all comes to a screeching halt, data checks, equipment checks, CPU master checks, console printer out of paper, even though it isn’t, on and on and on.

[@Obladgolated]

Ordinary programmers would not interact with this front panel in any way other than to look at its beguiling blinking lights through a glass window, from a hallway outside the "computer room," which was usually protected by elaborate locked doors. The people who did interact with the panel were almost always IBM CEs, except those who turned the machine on and off. Ordinary programmers (of the 360 era) interacted with the machine only by submitting "jobs" in the form of decks of punched cards; this was done using card punches and card readers that were in another room (or rooms), separate from the "machine room." During times of normal operation, an "operator" had a seat in front of the panel, which he didn't usually touch. A nearby terminal (such as the IBM 2741, with its thrilling, whirling golf-ball type element) told the "operator" what tape to mount on which tape drive, or what disc pack to mount on which disc drive, or that the line printer had run out of fan-fold paper. The "operator" often didn't know anything about computer programming, being more like a human robot actuator that accomplished mechanical tasks that the mainframe couldn't do by itself. The debugging aides available from the front panel (some of which were almost like early logic analyzers) were used in conjunction with test software, of which the CEs were the custodians. Ordinary programmers debugged their code with the help of voluminous "core dumps," which took up reams of computer paper and were typically looked at once and then never again. Often hundreds if not thousands of jobs depended on the continuous and fault-free operation of the computer behind that front panel, which had considerably less power than the CPU in an iPhone. There is no commodity that has fallen more quickly in price than that of information processing power.

[@Bhakti-rider]

Operators indeed did not need to know any programming language, and mostly didn't, but they ("we" for a time) did more than mount tapes and disk packs and put new paper in the printers. One would always be sitting at the console and monitoring what was happening in the system, and often entering replies or commands.

[@TiagoFernandes-ro6ck]

A classic... Using the kitchen knife to unbox goodies...😃

[@clemstevenson]

+graphic equaliser & Dolby mode select :-)

[@davesmith9325]

Brilliant

[@leeacton9572]

I recently bought a front control panel from an IBM S/360 Model 65 that was part of an IBM 9020 system from one of the FAA's Air Route Traffic Control Centers. Mine is in excellent condition and everything appears to be original and all the wiring is intact. However, a large number of the bulbs are burned out. I’m going to need to replace those if I want to get it blinking again. In your video (@<a href="#" class="seekto" data-time="274">4:34</a>) you say “I found all the lights” Do you have a source for these bulbs or any specs for them? Thanks!

[@thcoura]

What a house !

[@DoRC]

I can't imagine how much wiring there was behind that panel. There was 1100 connections to the output lights alone!

[@CuriousMarc]

Yes, the wiring was a work of art that could have belonged to the Apollo spacecraft. Very sad it was taken out of this panel, but these are rare enough that I could not be picky...

[@Lindsay5137]

Don't really remember any problems with the flat cables but what did cause problems was "cold flow" when the yellow wires were tight around the pins on the back of the boards, even just looking at them made the bug go away for a week or two. I remember replacing a card in memory on a Mod40, the fault disappeared immediately. only to return 5 hours later - turned out to be a faulty fan.

[@willharry6791]

@@Lindsay5137 One of the problems with the Mod 50 was silver migration at the connectors of the flat cables. Silver spurs would grow and short to an adjacent pin and cause an error (Master Check). The current flow would in the spur would cause it to open up. So when attempting to troubleshoot the problem it was no longer there. In the account I serviced we had this problem and the system was up and down for about a week before we finally realized what was going on. We had to replace all of the flat cables.

[@RobertDeloyd]

I worked on one of those in the early '70s :)

[@EvilSandwich]

The address and data panel kind of remind me of the front of an Altair 8800. Did you also have to use those switches to manually input a bootstrap loader? Or was the machine more sophisticated than that?

[@RobertDeloyd]

@@EvilSandwich yes, doing a cold start on the 360 we flicked the switches and turned the dials. The 50 is one of the machines I worked on. I changed jobs and went from a 370/155 to working on a 360/30, and when that one finally died the company purchased the 50. I believe I learned more on the 360s than I ever did on the 370. They have core memory! www.righto.com/2019/04/a-look-at-ibm-s360-core-memory-in-1960s.html

[@stevenpam]

I don't know why I watched this, but it's cool

[@OfficialiGamer]

so just curious how did you find one of these?

[@dusterdude238]

the switches look like either "Switch-craft" Corp. or a similar style.

[@JoeGoesx]

Was it raining when you filmed this?

[@truettneathery4358]

We had one of these at R.M. Parsons !!! In the basement !! OPerators from India !!

[@TheZooman22]

What type of wire would you need ? I can only imagine terminating the wire at various points would be a big challenge, since contacts like that are not made anymore.

[@CuriousMarc]

Yes, the wiring harness was a work of art (try to see if you can find a picture of the back of one of these panels), and is going to be a difficult thing to reproduce. Since I don't have the original lamp sockets and plan to hook it to an emulator anyhow, i am probably going to go for some kind of high speed serial protocol to make the wiring manageable.

[@deaustin4018]

You know, I wrote programs for these things back in the early 70s, but I've still never really seen one up close. You just gave the punch cards to the girl at the window, then got a printout back a day or two later (a thin one if your program was ok, a thick one if your program was full of bugs).

[@jmafoko]

hehe

[@citizen240]

//SYSABEND DD SYSOUT=A Would get you a VERY thick printout if the program blew up, aka ABnormalEND. Actually, the thinnest printout came if you had a job control language (JCL) error because that meant you were unable to even start the job/program.

[@Bhakti-rider]

Ha... no one (application programmers, I mean) used SYSABEND; SYSUDUMP, yes. You needed to know assembler to properly read a dump, but even if you didn't you could often find the trouble. I learned to put an 8-byte 01-level data item with a specific set of special characters at the beginning of Working-Storage so I could find it on the right side of the dump. Then I could go from there to find the values in other Working-Storage fields. That often revealed what the situation was.@@citizen240

[@taters5586]

Man, you have a very nice house.

[@user-hw7dm9ik4p]

my 1st machine which i was teen, and wrote FORTRAN and COBOL for that

[@scalamasterelectros3204]

Bro i love mecanicl swiches i can play with them for hours

[@moow950]

Nice to use with a modern (Rasperry based?) emulator!!

[@theoldman947]

On the upper left is a red light called "Master Check." We would scroll through the rollers looking for the actual error to see what happened. Once some young lady asked a customer engineer (repair person) what the Master Check meant. He said that there was two kinds of faults: alphas and betas. Master Alpha faults were easy to fix but those Master Beta's were really bad. I laughed my head off but the young lady pretended not to get it.

[@JGunlimited]

Wasn't expecting the wood case lol

[@roachtoasties]

How long did it take for IBM to arrive after you pulled the Emergency Pull switch? :)

[@willharry6791]

It might take an IBM CE a while to arrive, but the data center manager would be in the computer room faster than a New York second.

[@johnlister]

It depended on the situation. At a big UK bank in the 1970s-early 1980s, we had an IBM CE on site from early Monday to early Saturday.

[@elen5871]

i hope it's not too gauche to ask, but how much does something like this set you back? having something this cool is kind of a dream of mine lol

[@tickertape1]

Couple of thousand

[@moezboubaya9755]

How much it cost, if you wanna buy one historical machine like this one ?

[@Abr3200]

What could you do with something like this?How powerful was? Can you give us some Specs? :) Tnk!

[@CuriousMarc]

The System/360 line spanned everything from entry level (model 30) to supercomputing (model 195) and helped consolidate IBM domination of the industry. The Model 50 was a bigger mainframe for large business applications. Very competent machine at the time. Up to 512k of core memory, then up to an additional 8MB with the slower Large Core System. 500ns cycle time, 32 bit machine, rated at 0.2 MIPS on Wikipedia. But that really does not tell the true story because these mainframes were so strong in IO. They really had separate channels and processors for IO, so lots of IO capability: 16 tapes, many disks, lots of terminals, a new version of the 1403 printer that could do 1200 lines per minute, etc... It had an OS with virtual memory and time sharing. Perfect for database, accounting and time sharing terminals. Bank and insurance were big customers: the model 50 was delivered to Bank of America. This one came from Hong-Kong, so I suspect also the mainframe for a large bank or insurance company. The SABRE airline reservations system (still in use today) famously ran on IBM 360s. The mod 50 also had decent floating point performance, so you could do scientific work.

[@johnlister]

@@CuriousMarc The 360/67 was the prototype for virtual storage, but it really was the 370 series of computers where IBM fully implemented the feature.

[@Bhakti-rider]

The 360 didn't employ virtual storage; that was introduced in the 370.@@CuriousMarc

[@CuriousMarc]

@@Bhakti-rider Sorry virtual machine.

[@lawrencedoliveiro9104]

<a href="#" class="seekto" data-time="693">11:33</a> Specifically, channels are *DMA* I/O. The idea was to take the burden of low-level byte-by-byte I/O transfer off the CPU onto dedicated peripheral controllers.

[@johnlister]

And in those days, a channel was an actual computer in its own box, running a program of Channel Command Words (CCWs). Another topic for discussion!

[@station240]

Ex Movie prop I take it.

[@jimechols4347]

Miss all these old electromechanical computers. Today everything is so solid state, not even the mouse is electromechanical anymore it's just all solid state. But yes I agree solid state components are way more reliable than electromechanical and give way less problems.

[@z_polarcat]

Thats one giant Alter

[@Gr8thxAlot]

Very cool! A colleague has one of these in mint condition. If anyone has interest, let me know.