PDP-11/10 running paper tape BASIC using a high speed reader / punch
ฝัง
- เผยแพร่เมื่อ 25 ส.ค. 2024
- In this demonstration we show a PDP-11/10 computer connected to a PC05 high speed paper tape reader / punch and a VT100 terminal. The toggle in loader is entered into memory and the absolute loader is loaded into memory. Then the BASIC interpreter is loaded. When running BASIC we load a short program called WEKDAY. (The only more interesting program that was found that didn't make use of string functions which is not part of this version of BASIC). We also show how it is possible to save a small program on paper tape.
![[TH]FFWS SEA 2024 Fall - Knockout Stage Day 5](http://i.ytimg.com/vi/N2tl9xkQyyU/mqdefault.jpg)
![[TH] VALORANT Champions Seoul - Grand Final - EDG vs TH](http://i.ytimg.com/vi/WvObMc9VelI/mqdefault.jpg)
I worked on PDP-11/70's in the mid-80's while in the Marine Corps. Watching the boot sequence reminds me of what a pain it was to boot these things from a "cold" state. That was our worst fear, booting a cold and dark system. We had magnetic tape and disk drives to boot from, which was certainly faster than paper tape but the initial sequence of setting up where it would boot from is very familiar. Thanks for taking me back in time.
Oh my god, I used to repair those papertape reader and punches in the early 80's! Remember standing knee high in used papertape while adjusting the punch mechanisms :-)
The problem is that fan-fold paper tape is hard to come by nowadays. I have a few boxes but I don't really want to punch more than absolutely necessary. Maybe you remember how to adjust the reader? I have one here that works fine when reading continuously, but fails when reading with stalls. For example when the BASIC interpreter reads the tape, it reads a line and then stops for processing a brief moment and then continue with the next line. In this mode it is much more error prone than when reading at full speed. I think I tried most things listed in the manual but it still misbehaves. Ideas?
If my memory is correct, it could be something with the reader stepper motor start time/ramp
oh so many memories…
my first real experience with a real computer was around 1977 with an 11/05 and PC-05 reader punch. LA36 and VT52.
I never looked back! Later joined DEC in support - and remember toggling the bootloader on many systems -- soo many times ?
And that is an actual bit bucket. Somebody asked me what is in the bit bucket once. I said "ones".
Hah, I remember those days… we started at 300 bps (often on 20ma loops), evolved to 1200, RS232 and 9600 was all we could ever need !
It sounds like a Scroll Refrigeration Compressor. What a racket! ;)
All these machines in „cigarette grey“. Far away time of my youth.
I actually typed the demo program you loaded off the tape and loaded it into an Apple II (after making some adjustments and quality of life improvements)
I didn't have "type a BASIC program off a video of a VT100, then port it to FPBASIC" on my bingo card that year! ;o)
oh man, does this bring back memories
This old thech is amazing
Så vackert!
So that's where confetti came from lol.
Yes. This pdp 11 loads a basic program faster from paper strips than a Commodore 64 did from Cassette tape...
Indeed
I remember that, we filled an umbrella with confetti. It took one day for the guy to get rid of them
Yep. I put chad on a ceiling fan ready for the cleaners on April 1st.
In the US Navy "chad" was used as a verb as in "chadding" someone's bunk [bed].
Submarines generally had chadless punches that punched a "U" instead of a round hole.
I can see how punch tape was a step up from punch cards, at least if you drop the tape, you don't have to fuss about getting everything back in the right order again, or run the cards through the collator once more..
worst that could happen, is if the tape was stepped on and torn, then you get to start all over re punching it. 😮💨
Remember to empty the bit bucket
I think I would build a series of bars/sticks with fingers on them to set the address and opcodes for the bootloader. One per instruction. Then just use each bar in order so I don't screw up the loading of the bootloader!
The physical Bit Bucket is at 2:23 !!
Confetti!
Very interesting! Nice!!!
they make great confetti but are a pain if you spill them.
Rip Tayler Lives!! 😅
Schysst, man blir impad, Vilken hastighet fixar hålremsstansen?
"It's not a computer unless the light flicker when you start it":-)
Stansen gör ungefär 50 tecken per sekund, lite beroende på nätfrekvensen. Läsaren läser med 300 tecken per sekund.
I love that quote, it's so funny.
Look at that bitbucket fill!
This is so fascinating! however... why the need of entering the bootstrap loader every time ? they might not had have ROM (silicon) but CORE memory that can hold a program indefinitely . so why the reprogramming every time ?
As soon as you toggled the bootstrap in it can stay resident in the core memory unless it gets overwritten. So it should not be necessary to toggle it in every time. But sometimes it is hard to know if it has been overwritten or not…
Wonder if you still have the original basic games from the 70s? They came from field service and were all .BAS files? There was a good one called trolls and trees.
Other useful paper tape packages were ODT11X and FMPM11S as well as BASIC11
Anyone else here think gatefold paper tape looks cool?
I believe the PC05 cost nearly as much as the computer itself did back then.
I found a price list from 1972 on the net: www.village.org/pdp11/faq.pages/pricing.pdp11-05-15-apr-1972.html . Apparently a PDP-11/05 (OEM version of the 11/10) was $6495 with 8kW memory. Expanding memory to 16kW was $4400 extra. The PC05/PC11 was priced at $3900. Now this particular machine is slightly younger so it is possible that the prices had come down by then.
@MichaelKingsfordGray go away...
Do you still have all of this old stuff? This belongs in a museum... and by museum, I mean my apartment :) I live near Göteborg. You can drop it all off ;)
Do you need hearing protection to work around old computers? I would just for my sanity.
The paper tape punch is of course quite noisy, but the fans in the computers are not much noisier than a normal rack mount server. So, no, I am not wearing any hearing protection while working on these machines.
@@mattislind4443 Punching paper tape is quiet compared to punching and reading punch cards.
Does your 11/10 have 16 or 32KB memory? Assume the Basic you used is without support for strings...?
It has 16kByte. As far as I remember it was Basic with strings.
@@mattislind4443 The ppt for basic with strings barely fits in the input tapeholder. Hmm, need versionnumbers here.
Anybody notice that this machine made in 1971, was 2000 compliant in the date?
Actually I think that the guy that implemented this toy-program once upon a time took the job seriously and made a good job when implementing the date handling routines. Some old DEC operating systems do have problem handling dates beyond 2000.
Actually I am an old DEC Engineer. I ran tests on most of our systems in '93 to see if any TOY clocks or the OS failed. I did not find any that failed. Which ones and what OS do you feel failed? RT11, RSTS, RSX-11, VMS, and OPEN VMS all passed. It was the applications that had issues.
I don't know what versions you tested. I was at least unable to set a four digit year in RT-11 v5.3. I also seen reports that RT-11 (version ?) doesn't handle rollover from 31-DEC-99 very well: www.nycresistor.com/2014/05/15/pdp-11/ Another is a this discussion thread that say that most y2K deficiencies was corrected in v5.7: www.vcfed.org/forum/showthread.php?43632-Y2K-for-RT-11
Wow, when nobody could imagine even a Nokia 3310.
I have an Heathkit H11A, and that thing is quite loud with the fans and motors running.
Question btw, with regard to the 8" drives from DEC, do the motors run constantly? I have one and when I turn it on, the motors never stop. I would say that is abnormal and a clear sign of failure in most drives, but when I looked inside, the drive head appears to be pulled down toward the disk with an elaborate system.
It has an electromagnet, and a plate with a hinge attached to the head. This way the head is always suspended away from the disk unless it is reading/writing. So since I have never seen one actually working and I am still fixing this one, it would be good to know. I can think of no better person to ask. BTW, I know you might not have one of these, but they are extremely similar/almost identical to the RX-01 dual 8" from DEC...
Also, If it turns out that the bootstrap 7400 roms on here are broken, do you know where I can source a copy of the actual roms? Even if they aren't the superior heathkit versions with format and extended modes, it would at least be something to work from. I would settle for even a text with a core dump of the thing.
Also, what kind of paper tape punch would you suggest I use with an LSI-11 with M7270 CPU?, (assuming I will never find an H-10 paper tape machine)?
I was considering one of those NSI2400 RS-232 high speed punches, but they don't look very "DEC". I want to DECk out my machine if you get the pun :D
8 inch floppy motors usually run always. At least those using AC motors.
I am not sure what kind of boot card you have. There are several different boot cards, BDV11(M8012) and REV11 (M9400) for example. You need to tell more about your system. Most boot ROMs are available as images online at different locations. Here for example: www.pcjs.org/devices/roms/dec/
Paper tape on the LSI11 was not very common as far as I understand. At least not using DEC-branded components. But if you can get hold of a PC05 (the same as in my movie) and a M8010 interface card you should be able to get it working. I think the board would be your biggest problem to get hold of. Maybe you could build your own board?
I suggest you to check the forum at www.vcfed.org/forum/forum.php . At lot of people that can help you will be around there..
I'd love to obtain a H11, but I don't know where one could be purchased in the UK. Not to mention it'd probably cost a fortune :(
very CuriousMarc esk....
Very cool! Where in the world did you find an 11/10??
This one came from an ex DEC employee. But they do turn up on Ebay from time to time.
Could these punched tapes only store text?
One inch wide punched paper tape can store eight bit binary data. Not just text.
Early Quotron computers were loaded from paper tape. They used machine language. The plurality of computers were in Manhattan although headquarters were California so technicians would trundle a tape reader (made by Facit) around the NYC subway system from office to office loading computers. This in the '70's.
Hi,
I have a PDP 11/10 but it needs repair. Do you have any good pointer to hardware documentation and troubleshooting?
Thanks
You should take a look on the "Vintage Computer Fourms" if you need help with repair, especially if it's an early PDP-11 like that. Lots of friendly people who know what they are doing.
A lot of documentation are available at bitsavers. Look here for PDP-11/10 : bitsavers.trailing-edge.com/pdf/dec/pdp11/1105/ Then I fully agree that the forum is a great source of help.. Check www.vcfed.org . Good luck and feel free to ask me through mail if you need help.
Thank you! I now have a lot of documentation to read and I also subscribed to vcfed.org forum.
@@mattislind4443 lots of 74/74S-stuff that’s no longer available. Early 80s I say a lot of such machines, I remember one, Prime something, sounded like a jet on takeoff and one of the operators showed me a spare cpu board (measurable in m2). I wondered where the CPU was and found F181s in sets of 4 (16bit) all over the board. For ALU, address calculations, …
toggling in a program been there and done that how old am i
I think we may come back to this after all CPUs and screens and hard disks are DRM blocked signed patented protected company secret intellectual property and you can write into a RAM only if you authenticate to the RAM that you are an app which has been allowed into an app store and the company can revoke your permit at any time without any explanation.
Used to call this "Fat fingering"
i hated octal. 3 bits is not a power of two. I always wondered wg DEC used such an archaic system like that. Even the Altair used octal (which I modified to use hex) the IMSAI made much more sense
Actually I find octal not that bad. I think it is easier to remember octal numbers than hex. But maybe it is just me. Anyhow, it is definitely a heritage from when DEC computer had 12, 18 and 36 bits computers.
@@mattislind4443 the reason I find hex a lot easier is because humans think in decimal (0-9) and its more natural to group things that way and 0-9 can be represented in 1 hex nyble rather than 2 octal nybles. Most Electronic chips also packaged things into power a of two (1,2,4,8 etc) for example a flip-flop chip in 74376 integrated circuits were packed 4 flip-flop in one packagde or dual flip-flops in a 7476 etc. When I built my first personal computer baaed around the intel 8080 the was also and 8 bit computer so I was more efficient to use chips that were based on 2,4,8 bits rather than 3,6,9. Powers of 2 was a more natural progression that than odd sizes that DEC used.
I agree that 0-9 is more understandable for humans, but what about A-F? Most people doesn't recognise them being digits. From what you write it look like you talk about BCD, which of course has its uses . But most data processing takes place on binary numbers which has to be represented using A-F as well as 0-9.
The number of functions packed on to a SSI or MSI chip is mostly a function of the number of pins of the chip. Most of them were 14 or 16 pins. The 7404 packs 6 inverters into a package. A 74368 packs 6 three-state buffers into a 16 pin package. And the 74174 packs 6 flip-flops into one 16 pin package.
My opinion is that I don't find octal that bad because I think they are easier to remember. The octal legacy goes way back to when the computers had word lengths like 12, 15, 18, 24 or 36 bits. The DEC PDP-10 range used 36 bits and could pack six 6-bit characters into one word. I think that the conversion to 8 bit and power of 2 all started with the IBM 360 line which was a 32 bit architecture that had 8 bit bytes. It all fitted well with the EBCDIC character encoding scheme which also was 8 bit. ASCII BTW is 7bit. That would been an interesting mess to base things on. 14 or 28 bit machines?
DEC stayed with octal on the PDP-11 even though it was 16 bit and byte oriented. The VAX-11 however switched to HEX.
@@mattislind4443 yoy would use a lot more inverters and small logic than flip flops etc that's why they packed more small logic on a standard 14/16 pin packages. But most other logic was based on multiples of two. Even the 1401 computer (which predated the 360) was based on BCD logic. The punch card was bcd based as well. It was the punch cards that came before the age of modern day computing suce as in devices like the 407 etc. DEC was late in the computer game with thier first machine being developed in the late 1960s well after the other manufacturers were producing theirs (IBM, BURROUGHS, etc) even Burroughs was BCD based. So departing from the std BCD based way of doing things was odd. Having odd word sizes just did not fit into the standards at the time. Reading BCD in hex is a lot eaiser than reading it in octal. Even IBMs packed notation reads easier
@@rty1955 In regards to TTL chips and the integration level it is easy to check the list (en.wikipedia.org/wiki/List_of_7400-series_integrated_circuits ) There are 1, 2,3,4 or 6 functions per chip among the earlier TTL chips (than 1970) packed in either 14 or 16 pin packages.
IBM very early made the distinction between business and scientific computers. The business computers used BCD arithmetic while the scientific computers used fixed or floating point arithmetic. The advantage of BCD arithmetic is that you can get arbitrary precision without rounding errors. The disadvantage is the lower speed. Now when counting money you want it to be exact but you can probably afford the lack of speed. While scientific computation you can handle lack of precision if you do the math right. And you need the speed.
Prior to the 1410 there were a whole range of both commercial and scientific computers from IBM (en.wikipedia.org/wiki/IBM_700/7000_series ). There were no such thing as departing from BCD. IBM did both. The 1410 (and the tube based 650) was the logical extension of the punched card unit record equipment like the 407 and the like. But IBM had scientific machines like the 701 and 709 at the very same time.
Hollerith cards is dating back to 19th century and has a very peculiar coding with 12 rows. The encoding is nothing like 4 bit numbers BCD we are discussing (homepage.divms.uiowa.edu/~jones/cards/codes.html ) albeit called BCD under some circumstances. But was really BCDIC (Binary Coded Decimal Iterchange Code) 6 bits codes. Later on this became EBCDIC which were 8 bit code. (en.wikipedia.org/wiki/Binary-coded_decimal )
DEC built their first computer, the PDP-1 in 1959. It was all transistor 18 bit ones complement arithmetic machine. (en.wikipedia.org/wiki/PDP-1 )
Most processing is computers are done using twos complement arithmetic. Except for certain business applications where possibly COBOL is used and thus BCD arithmetic for processing infinite length numbers.
Hexadecimal representation of binary numbers has really nothing to do with the arithmetic used. It is just a way of presenting binary numbers to a user. Octal, binary, hexadecimal or BASE64 for that matter can be used for representing binary data. Choose whatever base you like!
Wacky waving inflatable arm-flailing paper tape reader!