I was hired in 1978 by Digital Field service. I was factory trained on the 11/34A you have here. I miss working that world of technology.I retired from DEC after COMPAQ obtained them. I Supported all members of the PDP-11, VAX, and ALPHA's. I still have all my maintenance gear, books, and Micro-Fiche. It is all in storage. I have not looked at it in over 12 years. Thanks for the refresher.
I've never worked on one in person but enjoy the 11/70 emulator that runs in the web browser. Its amazing how many OSes existed for these systems, RT11, RSX, Unix, RSTS, and XXDP to name a few. Most feel very familiar to someone who grew up on DOS and later Unix and Linux.
"Because computers came from IBM." That leaves out the details of the reason DEC called them "programmed data processors:" many companies had policies that one organization was responsible for all computing, and forbade teams from purchasing and setting up their own computing facilities. The approved, central computing resources typically involved IBM equipment. Calling their equipment "programmed data processors" meant that purchase orders for them would fly under the radar of purchasing departments that would disallow purchases of "computers".
Naming the machines "programmed data processors" also meant they were taxed less than "computers" in the US, making them an even more attractive alternative to bigger systems from the likes of IBM.
There’s a YT interview of Ken Olsen, where he said at the time he looked for investors, there where many companies announcing computer and he was told not to use that (toxic) word, because there was need for only a few of them predicted. IBM once thought two where enough, one for the east coast and one for the west coast. Maybe they just wanted competitors not being financed.
I work for DEC has a service tech, from 1976 to 1999. I have worked on all the PDP 11 line , PDP 8, PDP 9. The PDP 11/34 was very popular in those days. It was very pleasant to work on the 11/34. It is bringing me lost's of memory. Thank you for showing it.
I liked it a lot. You mention that the core memory did not work. This memory needs to be supplied by +20V and -5 V. Power supply H754 provides these voltages. In minute 18:17 I can see that this power is missing. Without these voltges the core memory can not work. Congratulations for these vídeo.
Takes me back. My first job in IT was operator for two PDP machines running the admin software for a college - an 11/34 with two RK07 drives and an 11/44 with three RL02's and a tape drive, both running RSTS/E with DECnet. These were pretty old machines even by that time and they were woefully underpowered for the demand on them. It was quite a step up when we moved over to a MIcroVAX-II!
Danke für die Präsentation ! 1978 war eine tolle Zeit. Da gab es noch keine Laufzeit- oder EMV- Probleme und man musste wohl robuste Fingerkuppen für die Mnemonic- Eingabe haben ;-) God save the TTL Chips on Board ;-)
Ich bin damit einverstanden es war eine tolle zeit... Ich habe ein 11/03, aber ich musste es repaurieren. Die terminal ist kaputt. Es ist ein Heathkit H11A, aber innen dies machine es ist eigentlich ein PDP 11/03 mit ein M7270 CPU und 32kw RAM. Auch, es habt dual 8" floppies SSSD 256k/floppy. Also, es habt 512kb online datenspeicher! Wenn ich habe dieser computer werken, Ich habe ein 4000 seite handbuch zu lesen fur RT-11 V4, DEC BASIC V3.2, und DEC FORTRAN IV (FORTRAN 77). TTL forever!
That is one helluva clean BA11s box !! I was factory trained on 11/34, 11/44 and 11/780 way way back in the 1980's. I still fix them today - currently an 11/70. I have not lost the art of fault tracing and chip changing !!
We had two PDP 11/34's in my laboratory at Washington University in St Louis in 1979. Each one had 64k words of memory and two RL-01 disc drives. We did a lot of really great scientific work. One was used for general scientific computing, and the other was used to control a JEOL JXA-733 electron microscope / X-ray microanalyzer.
The last RL-01 I saw was in a dairy where they used two redundant PDP's to control their industrial milk pasteurizing plant. One day when both PDPs crashed at the same time, they had to flush 10'000 liters of milk because it got contaminated with cleaning fluid, because the valves stayed open and it overflowed. Otherwise it was a very reliable system, working for 30+ years.
incredible - I'm so thrilled to have an in-depth look at something I've always want to understand which was a bit before my time... much appreciation and much respect
Great video. Thank you for sharing! 20:00 they may use screws made out of stainless steel only to meet all sort of different certifications (mil, avionic, medical, who knows)
Thats a very early machine. The construction is similar to a very old machine we had at General Instrument Microelectronics back in 1982 - it was ditched soon after I joined - we built a new factory so only the useful kit came with us. Later I joined Digital Equipment in Ayr having used their equipment extensively in the Semiconductor Industry. In the Early 1990 that was when Digital failed yet again to enter the PC market successfully and made strategic errors over UNIX et al. We closed the Galway facility in Ireland and its just so hard to keep a company afloat when some bad decisions have been made.
I worked on Computer Automation kit (SyFA) back in the day - a direct competitor. I always wanted to get the lid off a PDP so I could compare them - curiosity satisfied after 35-odd years! I remember the hand-keyed test programs and the boot ROM options - paper tape, 8" floppies or the 'new' High Capacity disk drive. All contemporary machines had roughly the same options. Excellent video.
I used to chip chase those a long long time ago. Still have an 11/40 front panel (only) and a 16k PDP11 core memory board as art works. And a couple of 110v/230v rack mount transformers. I even owned an 1110, 1134, 1103, 1123 and an 1173 at various times. And still remember when the PDP10's I worked on were sold as scrap to extract the gold off the module pins. Unfortunately got rid of it all later including the dozens of service manuals etc.
I worked in a rental car company where the processing center had 9 PDP 11/23s for the reservation center and a PDP 11/08 for the ARINC or (airport reservation interface). I did the back-ups and file transfers every night on grave yard shift.
It must be difficult for you to junk a piece of equipment that you used to work on. Some people don’t know, the DEC OS team went on to develop Windows NT at Microsoft. DEC had a lot of revolunitary tech.
A blast from the past. I worked on an 11/45, and later an 11/34. DEC had some great engineers, but the short little bus grant continuity cards were an unpardonable sin. I can't tell you how many knuckles I tore up getting those little things in and out. It would have hurt them to put another 8 inches of fiberglass and a handle on top so you could pull them like any other board?
Great! Reminds me of the PDP 11/32 I used end of the seventies, which had a removable disk (maybe even two) for the OS, the application (gamma ray spectography) and a development environment (Fortran, Assembly; it even had Basic). For the gamma spectrums it had a display, the only terminal was a printing teletype (probably TT model 43). The spectrum was stored in core memory, so when the machine was switched on, the last spectrum sampled reappeared on the display. Very noisy altogether. It made me want to buy my first computer, a Commodore PET 2001, which was silent and had a CRT ;)
WOW dude I thought I worked on all PDP-11's I worked on 11/02, 11/03,11/05, 11/10, 11/15/, 11/20, 11/23, 11/35, 11/40, 11/45, 11/50. 11/55, 11/60, 11/70, 11/72, and 11/73, but I never even heard of a 11/32.
Fascinating stuff. Brought back some memories. Nearly 40 years ago, the video facility I worked for in London had this kit: imgur.com/z6DLcGb I think they were PDP11-48s or 70s. Certainly CMOS memory, though I think the first one from the left was earlier and had magnetic core. We had anti-static carpet, as the CMOS was very prone to errors caused by static. I remember learning how to restart the programme while keeping all the existing data intact in memory after a crash! The keyboards had programmable function keys, and I worked out how to programme one to write the edit data to an 8" floppy, and encouraging the editors to use this regularly.
This is the first time I see a PDP as a pure graphics machine. I have seen graphics cards for PDPs but they have been used for a display in a industrial plant. Very basic graphics :-) Do you have a picture of a PDP generated TV image?
"original Channel 4 animated logo" Presumably this would be the BBC logo for BBC 4, before or after a show from that time frame. Knowing the BBC, it would be used for decades afterward as well.
I had asked a guy once who was 'in the know' why on earth they did this in an era when it was completely possible to make etched PCBs, and in fact there were etched PCBs on the same products. The response I got was quite amazing, and actually makes a ton of sense: Board density...you could fit the boards wiring into a smaller footprint by using wire wrap, having the equivalent of a 4 or even 8 layer board with the footprint and cost of a 1 layer board. In addition, you could add or leave out various point to point layers based on the product line. e.g. DMA wires or no DMA wires etc. and reduce or increase the cost as necessary. At that time, even a two layer board was expensive, but once they invested in the equipment, point to point wiring allowed for ease of design, versatile options using the same layout, reduce costs, and a smaller footprint, resulting in boards that were way smaller than they might have been, and uniquely suited to customer needs. When you look at it that way, DEC was way ahead of most computer manufacturers by using this archaic method of circuit design/manufacturing.
@@PlaywithJunk Ich kenn die Dinger nur von einem DEC Handbuch und aus Erzählungen eines Profs, der in Rechnertechnik auch mal in die Bits des Microcodes einer (Schrankwand großen) PDP11 abgetaucht ist. Und dann der Befehlssatz, der so aufgeräumt ist, im Vergleich zu dem des 8086, den wir in Assembler programmieren durften. Wenn Herr Olsen nur geglaubt hätte, daß sich PC‘s verkaufen lassen. Vielleicht würden unsere PCs dann heute im PDP11 Mode starten und über VAX auf den ALPHA-Mode übergeben. Und darauf dann VMS++ aka WNT. (just hanging out my thoughts)
guily6669 usually one would instead program an editor program first which takes input from a teletype. Programming larger programs on the front panels switches alone is silly.
No one created programs at the front panel. You used a card or paper tape reader, or mag tape or hard disk drive packs. The front panel was just to boot the O/S.
Numeric (octal) keyboard and display. That’s a machine for „Warmduscher“(translate says „wimp“, „candy-ass“, „mollycoddle“). Real programmers would use bit switches. 😂
Well, calculators had existed for a long time in 1978, even pocket calculators. They were usually based on 4-bit BCD-arithmetics. The PDP-11 was built more like today's computers. In fact, it was the primary model for the instruction set and/or register structure of several microprocessors.
"mass produced" I don't know how to imagine it in the context that this was in an age when computing was all serious, before computing devices were throwaway consumer products one can get by just buying in the supermarket. Is it far off to assume that such boards were hand assembled one per day?
If you take 1000 workers you can produce higher volumes. They also had some automated wire-wrapping machines for the backplanes. But yes... it took some time to make a complete machine.
I installed over one hundred myself before 1980 and that was just in three US states. The company produced many, many more. I must get my fiche out and working, I have seen it online can't find it now. In there are details like this, including spares counts on projected fail rates. Yes we made all the spares needed to support this product, in the original run, for the life of the product. We even included that cost in the selling cost of the product and the if you were not under a maintenance contact, if it failed we charged you again. A practice I was never comfortable with. We were very good engineering company Thanks to Ken and his associates, expectant engineering. We were never really good at marketing.We made a LOT of money in Field Service.
Too much, the rates were eye watering compared to the hourly rate of any first class engineer. A certain telecoms company had me looking at third party maintainers to reduce these costs.
64-bit :-) LOL.... sorry... We're talking about the year 1978. Those PDP's were 18 or 22 bit machines. They counted some CRC-bits too so they got from 16 bits to 18 and from 20 to 22. Still odd bit counts from todays view. The Cray-1 supercomputer was 64bits (at 80MHz)...
The PDP-11 was a 16 bit machine - 16 bit registers, 16 bit virtual addresses. The physical address size was 18 or 22 bits depending on the model (or 16 bits if you had an old PDP-11/20)
I feel Im missing something, even when the PDP is a piece of history due to NIX develpment. In 1978 there were computer like the Commodore PET with screen, keyboard, basic, tape storage and diskdrive. I mean, for a 78 machine looks a bit outdated, at least when compared with home microcomputers. What is its strongest point ?. Speed ?. Can you proram it using the VT terminal ?. Also, procesors were already in production, like the Motorola 6800 and probably not too far away from the MOS 6502. porbably not as fast as this system. My best guess.
Well... the boards are maximum 4 layers and the center two layers are only for DC supply and ground. So they effectively are just dual layer boards and that was the maximum possible in those days (exept some super expensive military and space stuff) The backplane has maybe 10 or even more layers of wire wrap. So a higher density of wiring was possible. And it was easier to make design changes. They just had to change the wiring lists and the wires were routed to the new position. (sometimes by hand)
Hey did you ever repair the core memory? I have a DEC 11/03 LSI-11 (Heathkit H11A). It has the CMOS memory, lathered with generous splashings of gold topped and legged ICs...These machines are beautiful, inside and out. I should try to track down some core memory...did they ever make a 16 bit core memory for the QBUS backplane?
Just one of our OS's for the PDP/11's product RT-11 it was around much longer than UNIX and was much better developed. Dos was only a look- alike to RT-11. But DOS was just a toy in comparison. Marketing is what has controlled the direction of the world. The World settled on inferior engineering and suffered a technical setback. Thank god for the world there was a Dave Cutter to save Microsoft. (en.wikipedia.org/wiki/Dave_Cutler)
@@MCNOISE666 Those you mention are a look-a-like of RT-11. It was out much earlier. I say that because I learned RT-11 first, that is my bias. But RT-11 was much more powerful and robust.
Even though the date codes on the chips say 1978, and there's even an inspection sticker with 1981 date somewhere, these boards look pretty outdated technology even for 1978. An ALU made up of dozens of separate logic gates ? Magnetic core memory ? are you kidding me ? Microprocessors were already available and inexpensive in 1978. There were even some early 16 bit ones like the TMS9900 and NS PACE. Anyway, thanks for sharing this interesting demo !
For DEC this was a well tested and reliable design. DEC was always a bit slow in using new technology. That's why they missed the advent of the personal computer. And some of their operating systems were made for 18 and 22bit processors. Not so easy to switch to a 16 bit CPU.
I agree, furthermore, DEC processors were fully orthogonal as well as being CISC, with floating point in hardware...unlike those cheesy 8 bit microprocessors of the mid/late 70's from intel and friends. In addition, as Play with Junk pointed out, non-volatile memory is _very_ useful! Consider the downtime of booting up an operating system from tape, and then an interpreter, and then a program Vs. turning the computer on, and proceeding where you left off. You can think of core memory as a hibernation state when the computer turns off. Magnetic states remain in that state until you change them. It is conceivable to turn a computer off, completely unpower it, and then return 50 years later, and proceed where you left off. Try that with DRAM... DEC had some great innovations regarding equalized data/address paths for 16 bit systems, as well as parity memory. Much like the later 80286/7 the address lines grew to allow for higher memory capacity well in advance of intel copying these ideas. So while business machines from the S-100 era were getting on with 32K of ram, and upwards of 64k, DEC had multi-user systems with 1 MB of ram and the ability for many users to log in and use a single machine. Even on the smaller systems though, program sizes, due to orthogonal instructions and CISC architecture allowed for smaller memory footprint, smaller online data storage usage, and faster data processing of the same algorithm compared to any contemporaries such as the 8080A, 6502, 6800 etc. Even when the 68000 CPU came out, motorola made the timing so ridiculous that no one wanted it...until 'DTACK grounded' was discovered...whereupon it became the 3rd generation CPU of choice. It was unfortunate for DEC that DTACK grounded was discovered around 1 year before DECs jump into PCs...without DTACK grounded 68000s, DEC would have blown the computer competitors out of the water. Instead, they were just more expensive 3rd gen computers, and people who might have bought a DEC, bought an Amiga or Atari ST instead for example. Imagine if there was no 68000 with DTACK grounded... Apple would have died an ugly death, and commodore too, and Atari, and there would be 3 less major competitors in the home/business market. DTACK grounded altered the course of computer history for over a decade in its wake.
I was hired in 1978 by Digital Field service. I was factory trained on the 11/34A you have here. I miss working that world of technology.I retired from DEC after COMPAQ obtained them. I Supported all members of the PDP-11, VAX, and ALPHA's. I still have all my maintenance gear, books, and Micro-Fiche. It is all in storage. I have not looked at it in over 12 years. Thanks for the refresher.
I'm sure if you're willing to do it, lots of us would love digital copies of all that literature.
I would have to dig deep for mine, I also left my reader somewhere in a move. I have them for Motorola, 4 - phase, and I have many years with IBM
I've never worked on one in person but enjoy the 11/70 emulator that runs in the web browser. Its amazing how many OSes existed for these systems, RT11, RSX, Unix, RSTS, and XXDP to name a few. Most feel very familiar to someone who grew up on DOS and later Unix and Linux.
PLEASE scan it and upload it to an archive so that the whole community might have help keeping these old machines running!
Ok grandpa
I hope you put that baby back together. A working 11/34 is worth much more than the scrap.
It is well stored in working condition. That one does not go to scrap .-)
Play with Junk thats good to know, never scrap them. That would be like scrapping stratavarious violin
"Because computers came from IBM." That leaves out the details of the reason DEC called them "programmed data processors:" many companies had policies that one organization was responsible for all computing, and forbade teams from purchasing and setting up their own computing facilities. The approved, central computing resources typically involved IBM equipment. Calling their equipment "programmed data processors" meant that purchase orders for them would fly under the radar of purchasing departments that would disallow purchases of "computers".
+David Barts Good information...thank you!!
Naming the machines "programmed data processors" also meant they were taxed less than "computers" in the US, making them an even more attractive alternative to bigger systems from the likes of IBM.
There’s a YT interview of Ken Olsen, where he said at the time he looked for investors, there where many companies announcing computer and he was told not to use that (toxic) word, because there was need for only a few of them predicted. IBM once thought two where enough, one for the east coast and one for the west coast. Maybe they just wanted competitors not being financed.
I work for DEC has a service tech, from 1976 to 1999. I have worked on all the PDP 11 line , PDP 8, PDP 9. The PDP 11/34 was very popular in those days. It was very pleasant to work on the 11/34.
It is bringing me lost's of memory.
Thank you for showing it.
"Every bus needs a terminator"
And I pictured public transport buses with Arnold Shwarzenegger on board lol
not a bad idea. Nobody would dare to ride without ticket...
@@PlaywithJunk
is that 50 or 75 Ooohh...mmy..
can't resist . . !
Those backplanes were fantastic looking. Massive core memory (wow!), and beautiful memory boards.
I enjoyed this, thanks for filming!
I liked it a lot. You mention that the core memory did not work. This memory needs to be supplied by +20V and -5 V. Power supply H754 provides these voltages.
In minute 18:17 I can see that this power is missing. Without these voltges the core memory can not work. Congratulations for these vídeo.
Takes me back. My first job in IT was operator for two PDP machines running the admin software for a college - an 11/34 with two RK07 drives and an 11/44 with three RL02's and a tape drive, both running RSTS/E with DECnet. These were pretty old machines even by that time and they were woefully underpowered for the demand on them. It was quite a step up when we moved over to a MIcroVAX-II!
Deine Videos sind der Hammer; du erklärst kurz und sachlich. Du hast definitiv mehr Zuschauer verdient!
Danke für die Präsentation ! 1978 war eine tolle Zeit. Da gab es noch keine Laufzeit- oder EMV- Probleme und man musste wohl robuste Fingerkuppen für die Mnemonic- Eingabe haben ;-) God save the TTL Chips on Board ;-)
Ich bin damit einverstanden es war eine tolle zeit... Ich habe ein 11/03, aber ich musste es repaurieren. Die terminal ist kaputt. Es ist ein Heathkit H11A, aber innen dies machine es ist eigentlich ein PDP 11/03 mit ein M7270 CPU und 32kw RAM. Auch, es habt dual 8" floppies SSSD 256k/floppy. Also, es habt 512kb online datenspeicher! Wenn ich habe dieser computer werken, Ich habe ein 4000 seite handbuch zu lesen fur RT-11 V4, DEC BASIC V3.2, und DEC FORTRAN IV (FORTRAN 77). TTL forever!
Those system cards are beautiful. Love a Dec computers, RIP.
Yes, the PDP-11 was the first UNIX machine, and the smallest at the time; a PDP-11/20 ran the first complete kernel build in 1970.
That is one helluva clean BA11s box !!
I was factory trained on 11/34, 11/44 and 11/780 way way back in the 1980's.
I still fix them today - currently an 11/70.
I have not lost the art of fault tracing and chip changing !!
That's because this box was never running in a production environment. It was one of our workshop systems. And I cleaned it periodically. :-)
We had two PDP 11/34's in my laboratory at Washington University in St Louis in 1979. Each one had 64k words of memory and two RL-01 disc drives. We did a lot of really great scientific work. One was used for general scientific computing, and the other was used to control a JEOL JXA-733 electron microscope / X-ray microanalyzer.
The last RL-01 I saw was in a dairy where they used two redundant PDP's to control their industrial milk pasteurizing plant. One day when both PDPs crashed at the same time, they had to flush 10'000 liters of milk because it got contaminated with cleaning fluid, because the valves stayed open and it overflowed.
Otherwise it was a very reliable system, working for 30+ years.
incredible - I'm so thrilled to have an in-depth look at something I've always want to understand which was a bit before my time... much appreciation and much respect
Enjoyed your video i have an old micro vax i couldn't resist grabbing always had a fondness for DEC
These old PDP units are worth a lot more than their scrap value!
Great video. Thank you for sharing!
20:00 they may use screws made out of stainless steel only to meet all sort of different certifications (mil, avionic, medical, who knows)
I use to program these things. RSX 11M Plus.... oh the memories....
Thats a very early machine. The construction is similar to a very old machine we had at General Instrument Microelectronics back in 1982 - it was ditched soon after I joined - we built a new factory so only the useful kit came with us. Later I joined Digital Equipment in Ayr having used their equipment extensively in the Semiconductor Industry. In the Early 1990 that was when Digital failed yet again to enter the PC market successfully and made strategic errors over UNIX et al. We closed the Galway facility in Ireland and its just so hard to keep a company afloat when some bad decisions have been made.
I worked on Computer Automation kit (SyFA) back in the day - a direct competitor. I always wanted to get the lid off a PDP so I could compare them - curiosity satisfied after 35-odd years! I remember the hand-keyed test programs and the boot ROM options - paper tape, 8" floppies or the 'new' High Capacity disk drive. All contemporary machines had roughly the same options. Excellent video.
Think of how many housecats that thing could keep warm!
I used to chip chase those a long long time ago. Still have an 11/40 front panel (only) and a 16k PDP11 core memory board as art works. And a couple of 110v/230v rack mount transformers. I even owned an 1110, 1134, 1103, 1123 and an 1173 at various times. And still remember when the PDP10's I worked on were sold as scrap to extract the gold off the module pins. Unfortunately got rid of it all later including the dozens of service manuals etc.
Yeah, you can't keep everything.... I sent this PDP 11/34 to a technical museum. I hope they will keep it working :-)
I worked in a rental car company where the processing center had 9 PDP 11/23s for the reservation center and a PDP 11/08 for the ARINC or (airport reservation interface). I did the back-ups and file transfers every night on grave yard shift.
Real computers... just as I started with. Amazing how history moved.
those boards are gorgeous
It must be difficult for you to junk a piece of equipment that you used to work on. Some people don’t know, the DEC OS team went on to develop Windows NT at Microsoft. DEC had a lot of revolunitary tech.
At the local computer club we drift some VAX systems. Love the huge size and heavy storage devices.
A blast from the past. I worked on an 11/45, and later an 11/34. DEC had some great engineers, but the short little bus grant continuity cards were an unpardonable sin. I can't tell you how many knuckles I tore up getting those little things in and out. It would have hurt them to put another 8 inches of fiberglass and a handle on top so you could pull them like any other board?
Great! Reminds me of the PDP 11/32 I used end of the seventies, which had a removable disk (maybe even two) for the OS, the application (gamma ray spectography) and a development environment (Fortran, Assembly; it even had Basic). For the gamma spectrums it had a display, the only terminal was a printing teletype (probably TT model 43). The spectrum was stored in core memory, so when the machine was switched on, the last spectrum sampled reappeared on the display. Very noisy altogether.
It made me want to buy my first computer, a Commodore PET 2001, which was silent and had a CRT ;)
WOW dude I thought I worked on all PDP-11's I worked on 11/02, 11/03,11/05, 11/10, 11/15/, 11/20, 11/23, 11/35, 11/40, 11/45, 11/50. 11/55, 11/60, 11/70, 11/72, and 11/73, but I never even heard of a 11/32.
Fascinating stuff. Brought back some memories. Nearly 40 years ago, the video facility I worked for in London had this kit: imgur.com/z6DLcGb I think they were PDP11-48s or 70s. Certainly CMOS memory, though I think the first one from the left was earlier and had magnetic core. We had anti-static carpet, as the CMOS was very prone to errors caused by static. I remember learning how to restart the programme while keeping all the existing data intact in memory after a crash! The keyboards had programmable function keys, and I worked out how to programme one to write the edit data to an 8" floppy, and encouraging the editors to use this regularly.
This is the first time I see a PDP as a pure graphics machine. I have seen graphics cards for PDPs but they have been used for a display in a industrial plant. Very basic graphics :-) Do you have a picture of a PDP generated TV image?
"original Channel 4 animated logo" Presumably this would be the BBC logo for BBC 4, before or after a show from that time frame. Knowing the BBC, it would be used for decades afterward as well.
@@peterlamont647 Not BBC, it was indeed Channel 4. A 3D spinning number 4 in fact.
From the picture I see what could only be 11/34, 1134A, and or 1104. No such this as an 1148 and this not a pic of 1170
DEC PDP 11 - very fond old memories!
That's the biggest transformer I've ever seen. I wonder what the efficiency on that power supply is. I suspect not very good.
Dman! IT came a long way in 35 years
Microelectronics came a long way. "Computer science" has not evolved very much since the 50/60s.
Also, most of the logic chips in these machines are still manufactured, still standard components.
there's so much space inside. You could fit an 8" floppy drive :O
Looks in great condition.
Nice. That backplane job looked like quite a mess but I guess it did its job.
The back-plane were wired by CNC devices. It was mostly automated.
Most likely yes. But I have seen similar backplanes that have been made by hand.
It is not "most likely" at DEC I saw the build process machines. DEC did not build them by hand.
I had asked a guy once who was 'in the know' why on earth they did this in an era when it was completely possible to make etched PCBs, and in fact there were etched PCBs on the same products. The response I got was quite amazing, and actually makes a ton of sense:
Board density...you could fit the boards wiring into a smaller footprint by using wire wrap, having the equivalent of a 4 or even 8 layer board with the footprint and cost of a 1 layer board. In addition, you could add or leave out various point to point layers based on the product line. e.g. DMA wires or no DMA wires etc. and reduce or increase the cost as necessary. At that time, even a two layer board was expensive, but once they invested in the equipment, point to point wiring allowed for ease of design, versatile options using the same layout, reduce costs, and a smaller footprint, resulting in boards that were way smaller than they might have been, and uniquely suited to customer needs. When you look at it that way, DEC was way ahead of most computer manufacturers by using this archaic method of circuit design/manufacturing.
When re-programming the back plane the engineer
had to remove so many wires, then replace with new wires. fascinating to watch
Das Teil gehört ins Museum und nicht auf den Müll!
+Jean-Marc Zenhäusern Keine Angst, das landet nicht im Müll :-)
And both these cards later fitted on one chip, to be cloned in the UDSSR to be used in home computers and graphing calculators.
You mean the K1801WM1 chip?
PDP11 is 16Bit (the width of the data path), but depending on the „MMU“ you might have 18/22 address bits or 256k/4M physical memory addressable.
Ja richtig.... alles mal gelernt und viel wieder vergessen.
@@PlaywithJunk
Ich kenn die Dinger nur von einem DEC Handbuch und aus Erzählungen eines Profs, der in Rechnertechnik auch mal in die Bits des Microcodes einer (Schrankwand großen) PDP11 abgetaucht ist. Und dann der Befehlssatz, der so aufgeräumt ist, im Vergleich zu dem des 8086, den wir in Assembler programmieren durften.
Wenn Herr Olsen nur geglaubt hätte, daß sich PC‘s verkaufen lassen. Vielleicht würden unsere PCs dann heute im PDP11 Mode starten und über VAX auf den ALPHA-Mode übergeben. Und darauf dann VMS++ aka WNT. (just hanging out my thoughts)
@@fromgermany271 Ja ja, die "Grossen" irren sich oft spektakulär. Wer braucht schon mehr als 640k RAM...? 😃
I thought I'll see diode boot rom, guess that's only the older models.
There was also a FPP Card Floating Point Processor card that could be put on the 11/34A
A pep is never junk ! 😁
Please make a video of all that stuff you have in those shelves we see in the background. Would love to see a Compaq Systempro server there.
I don't think I have such an old one there.... Maybe a Proliant DL380 Gen1 or a HP Netserver
Very ASMR :D
Nice time you find One, Save it. you'll make more selling Whole..
A pdp is never junk 😎
I totally agree! It has a good place in my private collection.
I was really hoping you didnt recycle this
Recycling a DEC is a mortal sin...
So back then they would just sit in front of that LCD display with the calculator like buttons to program all day long as assembly?
guily6669 usually one would instead program an editor program first which takes input from a teletype. Programming larger programs on the front panels switches alone is silly.
Normally only used to set the address of the bootstrap loader, then press run. and away goes the loader to load your OS.
No one created programs at the front panel. You used a card or paper tape reader, or mag tape or hard disk drive packs. The front panel was just to boot the O/S.
Numeric (octal) keyboard and display. That’s a machine for „Warmduscher“(translate says „wimp“, „candy-ass“, „mollycoddle“).
Real programmers would use bit switches.
😂
Yeah I know... 😃 But I had no other front panel.
nice watch
Doom's day computer
where do you get all thought stuff, it awsome
can you run RSTS/E with DCL on this machine?
UNIX was written on a PDP 11/70 with 2 MB memory.
I think RSTS should work on the 11/34. We had it running on 11/23, 11/44, 11/53. Although most users with an 11/34 had other operating systems (RSX).
How long have you had this PDP-11?
Yo are from schweizerland :)
Jawoll.... schtimmt :-)
How did you find out 😂
Giant calculator :)
Well, calculators had existed for a long time in 1978, even pocket calculators. They were usually based on 4-bit BCD-arithmetics. The PDP-11 was built more like today's computers. In fact, it was the primary model for the instruction set and/or register structure of several microprocessors.
So core memory is virtually the predecessor of flash memory? ;)
Both share the ability to store data when power is off. In that sense, yes.
The direct technological predecessor of flash memory is EPROM, especially if talking about NOR-flash.
Flash does not allow writing individual bytes, but there is something called MRAM/FRAM, which is fast as RAM, but hold data on power off.
i wish i lived back in the day when stuff like this was a thing so i had a way better understanding of it
"mass produced"
I don't know how to imagine it in the context that this was in an age when computing was all serious, before computing devices were throwaway consumer products one can get by just buying in the supermarket. Is it far off to assume that such boards were hand assembled one per day?
Let's say, that 100 new PDP11s were produced a year (were that many?), plus 100-200 for replacement, in a year with ~300 workdays.
If you take 1000 workers you can produce higher volumes. They also had some automated wire-wrapping machines for the backplanes. But yes... it took some time to make a complete machine.
I installed over one hundred myself before 1980 and that was just in three US states. The company produced many, many more. I must get my fiche out and working, I have seen it online can't find it now. In there are details like this, including spares counts on projected fail rates. Yes we made all the spares needed to support this product, in the original run, for the life of the product. We even included that cost in the selling cost of the product and the if you were not under a maintenance contact, if it failed we charged you again. A practice I was never comfortable with. We were very good engineering company Thanks to Ken and his associates, expectant engineering. We were never really good at marketing.We made a LOT of money in Field Service.
Too much, the rates were eye watering compared to the hourly rate of any first class engineer. A certain telecoms company had me looking at third party maintainers to reduce these costs.
When was the last year still machine was still in active use?
I don't know. We had a customer with a similar system that was running until 2000.
Add an M9312 then!
Was this kind of computer a 16, 32, or 64 bit computer?
64-bit :-) LOL.... sorry... We're talking about the year 1978. Those PDP's were 18 or 22 bit machines. They counted some CRC-bits too so they got from 16 bits to 18 and from 20 to 22. Still odd bit counts from todays view. The Cray-1 supercomputer was 64bits (at 80MHz)...
@@PlaywithJunk thank you.
The PDP-11 was a 16 bit machine - 16 bit registers, 16 bit virtual addresses. The physical address size was 18 or 22 bits depending on the model (or 16 bits if you had an old PDP-11/20)
I feel Im missing something, even when the PDP is a piece of history due to NIX develpment. In 1978 there were computer like the Commodore PET with screen, keyboard, basic, tape storage and diskdrive. I mean, for a 78 machine looks a bit outdated, at least when compared with home microcomputers. What is its strongest point ?. Speed ?. Can you proram it using the VT terminal ?. Also, procesors were already in production, like the Motorola 6800 and probably not too far away from the MOS 6502. porbably not as fast as this system. My best guess.
Do you know why they used wire wrap? I mean, they had PCBs for all the other boards.
Quick and easy especially for prototyping.
Stuart Grant
This isn't a prototype.
True.
That's why I am wondering. Wire wrap is (probably) a lot of hand work.Makes it expensive...
Well... the boards are maximum 4 layers and the center two layers are only for DC supply and ground. So they effectively are just dual layer boards and that was the maximum possible in those days (exept some super expensive military and space stuff)
The backplane has maybe 10 or even more layers of wire wrap. So a higher density of wiring was possible. And it was easier to make design changes. They just had to change the wiring lists and the wires were routed to the new position. (sometimes by hand)
Choosed?
12 not 22
I would have thought that a museum of computing would have wanted it.
+DAVID GREGORY KERR If a museum makes a good offer, I will probably sell it.
+chrigel1234 you can donate it as a normal person usually does
Du hast den geilsten Job der Welt. Aber so ein Rarität verschrottet man nicht . Du könntest ein Museum aufmachen. 😢
can i get that sexy core memory?
Emm.... No! :-) Sorry.
:)
Hey did you ever repair the core memory?
I have a DEC 11/03 LSI-11 (Heathkit H11A). It has the CMOS memory, lathered with generous splashings of gold topped and legged ICs...These machines are beautiful, inside and out.
I should try to track down some core memory...did they ever make a 16 bit core memory for the QBUS backplane?
Computergeschichte Schweiz? Wo und von wann bis wann lief das Teil wohl?
Schwer zu sagen. Viele Firmen hatten PDP's und VAXen.
PS: Ciao Paddy :-) Danke für's gucken
Play with Junk : Bitte, deine Videos passen gut zu meiner “Computergeschichte“ auf 20sek.ch.
Hast Du irgendwo auch ein Video mit alten Reeltapes?
Nö, aber ich hatte vor mal eines zu machen. Das wäre eine Idee für ein nächstes Video...
A pdp is never junk 146486
I'm sorry to see you scrap this machine, you could've sold it to an enthusiast?
It was operational and not too huge or complicated to own...
was not scrapped
Did you keep it, no reason to destroy a perfectly working pdp-11
No worries.... it is still there. Well kept...
You realize these things are very valuable and sought after by collectors right?
He is probably a collector himself.
Yes I know. We have our own little museum.
It was the machine that gave us UNIX, thereby the modern world. A more powerful machine might have leapfrogged to just DOS.
Just one of our OS's for the PDP/11's product RT-11 it was around much longer than UNIX and was much better developed. Dos was only a look- alike to RT-11. But DOS was just a toy in comparison. Marketing is what has controlled the direction of the world. The World settled on inferior engineering and suffered a technical setback. Thank god for the world there was a Dave Cutter to save Microsoft. (en.wikipedia.org/wiki/Dave_Cutler)
It was RSX for me. Loved that OS even though it had some bad press.
@@markcummins6571 Isn't msdos a lookalike of CP/M?
@@MCNOISE666 Those you mention are a look-a-like of RT-11. It was out much earlier. I say that because I learned RT-11 first, that is my bias. But RT-11 was much more powerful and robust.
@@markcummins6571 Thanks Mark - I shall read further!
Even though the date codes on the chips say 1978, and there's even an inspection sticker with 1981 date somewhere, these boards look pretty outdated technology even for 1978. An ALU made up of dozens of separate logic gates ? Magnetic core memory ? are you kidding me ? Microprocessors were already available and inexpensive in 1978. There were even some early 16 bit ones like the TMS9900 and NS PACE. Anyway, thanks for sharing this interesting demo !
For DEC this was a well tested and reliable design. DEC was always a bit slow in using new technology. That's why they missed the advent of the personal computer.
And some of their operating systems were made for 18 and 22bit processors. Not so easy to switch to a 16 bit CPU.
BTW... magnetic core memories have the advantage not to lose stored contents when power is off. For some purposes, core memory was used even later.
I agree, furthermore, DEC processors were fully orthogonal as well as being CISC, with floating point in hardware...unlike those cheesy 8 bit microprocessors of the mid/late 70's from intel and friends. In addition, as Play with Junk pointed out, non-volatile memory is _very_ useful! Consider the downtime of booting up an operating system from tape, and then an interpreter, and then a program Vs. turning the computer on, and proceeding where you left off. You can think of core memory as a hibernation state when the computer turns off. Magnetic states remain in that state until you change them. It is conceivable to turn a computer off, completely unpower it, and then return 50 years later, and proceed where you left off. Try that with DRAM...
DEC had some great innovations regarding equalized data/address paths for 16 bit systems, as well as parity memory. Much like the later 80286/7 the address lines grew to allow for higher memory capacity well in advance of intel copying these ideas. So while business machines from the S-100 era were getting on with 32K of ram, and upwards of 64k, DEC had multi-user systems with 1 MB of ram and the ability for many users to log in and use a single machine.
Even on the smaller systems though, program sizes, due to orthogonal instructions and CISC architecture allowed for smaller memory footprint, smaller online data storage usage, and faster data processing of the same algorithm compared to any contemporaries such as the 8080A, 6502, 6800 etc. Even when the 68000 CPU came out, motorola made the timing so ridiculous that no one wanted it...until 'DTACK grounded' was discovered...whereupon it became the 3rd generation CPU of choice. It was unfortunate for DEC that DTACK grounded was discovered around 1 year before DECs jump into PCs...without DTACK grounded 68000s, DEC would have blown the computer competitors out of the water. Instead, they were just more expensive 3rd gen computers, and people who might have bought a DEC, bought an Amiga or Atari ST instead for example. Imagine if there was no 68000 with DTACK grounded... Apple would have died an ugly death, and commodore too, and Atari, and there would be 3 less major competitors in the home/business market. DTACK grounded altered the course of computer history for over a decade in its wake.
Peter Lamont great comment
Actually, there IS a microprocessor in that system! It was visible when the boards were shown...
First
OMG i can buy this machine !!!!