21

In my childhood I had one of these:

https://en.wikipedia.org/wiki/Electronika_BK

Note, that the CPU of this is based on LSI-11. This is rather surprising because PDP-11 took a room (I was lucky enough to be able to tinker with it at parent's work) but BK was a home computer / microcomputer.

I was fascinated by what I used to call "DEC architecture", mostly meaning the set of assembly command, registers, memory addressing, etc and I liked that they are the same at parent's work and on my small home PC.

Now, were there any microcomputers with similar processors (for the mass market) outside of Soviet Union? It does not have to be the same processor, but it should feature similar command set / registers / etc. Presumably it should be either by DEC or based on their design.

The most popular ones, like commodore, amiga, zx spectrum are all have very little to do with DEC and their processors.

  • 2
    Some people might have had a DEC Professional, VAXstation or even something bigger at home, but I think the Heathkit H11 is the only thing that would qualify as a "home computer". – Ross Ridge Dec 17 '17 at 4:03
  • 2
    Funny you should dismiss the Amiga, but the 68000 instruction set is clearly inspired by the PDP-11. – pndc Dec 17 '17 at 11:45
  • 1
    Yea, the 68k, much like PDP-11 and VAX, had a very programmer-friendly architecture. Lots of registers, rich and highly orthogonal instruction set, flat address space. The spirit of the design was much the same. – Alex Hajnal Dec 18 '17 at 5:19
  • Don't know if it completely fits your criteria, but Dec did their own range of "Micro" PDP (alongside Micro Vax): see PDP-11/23 and PDP-11/53 up until 1990. Not "home PC" size, but a lot smaller than a room! – TripeHound Dec 18 '17 at 12:54
18

The Heathkit H11 was available either as a kit or pre-assembled. It never became really popular in the West, but it was one of the most powerful PCs available in 1978. It used the LSI-11 small format of the PDP-11, and came with 4 kwords of memory for $1295. (That is 8 kbytes, but DEC preferred to refer to memory as register size, which was 16 bits.) It could be expanded to 32 kwords. IIRC, the LSI-11 CPU could actually handle up to 128 kwords (using bank swapping and/or memory-tunnel registers), but any memory above 32 kwords would have to be off the motherboard.

I agree with Andrew that the PDP architecture was fascinating and conceptually easy to understand. I used and programmed one at work for many years.

----- added a day after initial answer -----

I apologize to Alex Hajnal and others for using my own terminology, "memory tunnel registers". I was getting tired and chose not to look up the details on something that had very little to do with the question. The PDP-11 family offers several ways to access memory beyond 64 kbytes. One is the traditional memory mapping sort of deal where pages of memory can be mapped into and out of the accessible 64 kbytes. I was not referring to that.

The Q-Bus (used by LSI-11) offered 18 bits of physical address space, spanning 256 kbytes total. (Although Ken Gober indicates that on Heath's H11, the upper 2 bits of addressing were not wired, which would limit that machine to just 64 kbytes. I'm speaking of a fully addressed machine now.) Without memory mapping, the full memory space could still be accessed via 2 bits of the Processor Status register (PS), which provided address bits 16 and 17. Four instructions existed to access the full memory space.

  • MTPI -- move to previous instruction space
  • MTPD -- move to previous data space
  • MFPD -- move from previous data space
  • MFPI -- move from previous instruction space

What these instructions would do is push or pop a 16-bit word, accessing full physical memory via R6, the stack pointer (SP).

I didn't remember all the details, just that you get access any memory location directly without using memory management. Since it was one word at a time, the phrase I came up with was "memory tunnel registers". I guess the terminology almost fits, because the action does use both the SP and 2 bits of the PS.

  • Can you explain what you mean by "memory-tunnel registers"? I'm not familiar with the term and can't find any reference to it online. – Alex Hajnal Dec 17 '17 at 9:04
  • @AlexHajnal maybe it means indirect addressing. – RonJohn Dec 17 '17 at 12:11
  • I was told years ago that while the LSI-11 processor card might be able to handle more than 128Kwords, the wiring of the Q-Bus backplane in the H11 did not actually allow it to work. – Ken Gober Dec 17 '17 at 15:33
  • Someone told me a story that the designers of the DEC VT240 terminals used an LSI-11 because they were ordered to, and that a foreign company copied the VT240 including firmware and board layout; according to this person, outside demand for the LSI-11 dropped off severely when that company was shut down. Do you know if that was plausible, or was there substantial outside demand for the chip (other than VT-240 knock-offs)? I've used a VT-240 and was unimpressed with the performance, but I don't know if that was the LSI-11's fault. – supercat Dec 17 '17 at 16:58
  • maybe they were remembering story of C.Itoh CIT-220, which resulted in a lawsuit against C.Itoh (articles.latimes.com/1985-01-12/business/…). That said, Soviets also did produce a clone of VT240 itself, most likely an unlicensed one. – sendmoreinfo Dec 17 '17 at 17:58
8

It's possibly a stretch, but the General Instruments CP1600 which was in the Intellivision, though otherwise unsuccessful, was based on the PDP-11 architecture.

The Intellivision was a product of Mattel, not GI, so it's the one commercial machine that opted to use the chip rather than being the machine the chip was designed for. General Instruments designed the chip for general use and likely sought other purchasers. It was a genuine attempt to provide something patterned after the PDP-11 for micro use.

With either of a couple of official accessories (only one given a wide release) the Intellivision becomes a computer, but very far from a common one.

7

There was Terak 8510/A - a graphic workstation with the LSI-11 compatible processor, a graphical frame buffer (hardware-scrollable the same way as in the BK-0010), and a text mode with downloadable fonts, although admittedly too expensive to be a home computer.

In some sense BK-0010 looks like a stripped-down Terak (no text mode, less RAM, etc.), but the main feature - the scrollable frame buffer - is virtually identical.

4

There were one-chip versions both of the PDP-8/LSI-8 (the models are called DECmate, using the Intersil/Harris 6100 chip) and the PDP-11/LSI-11 (J-11 or "Jaws" chip, used in a range of PDP-11 models).

Also, these were hardly "home computers" due to their price point. They would be typically be bought by labs or research institutions, both in academia and industry, and not owned by a single person or family.

  • 5
    No, the CPU used in Electronika BK was completely different, and the PDP-11 instruction set was microprogrammed into it as a later design decision. As a matter of fact, the quality of implementation was not that good. The 3 MHz CPU was quoted as capable of about 300K instructions per second. – Leo B. Dec 17 '17 at 11:06
1

Back in to 80's DEC themselves had the PDT 150 which was an office version of our Lab PDP11/03's

http://www.computerhistory.org/collections/catalog/102681914

It could have been used as a home computer as its smaller than some current PC Cube cases and I believe DEC sold some to employees

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.