Shop OBEX P1 Docs P2 Docs Learn Events
Old Micro Discussions — Parallax Forums

Old Micro Discussions

Cluso99Cluso99 Posts: 18,066
edited 2018-03-20 08:14 in General Discussion
Thought I would start a new thread rather than continue the OT posts on Rayman's RL4000 thread.

The 70's and 80's were certainly an interesting time for microcontrollers, microprocessors, the home-brew designs, and the Apple/MacIntosh/PCs.

Some of my early microprocessor designs are up on my website www.clusos.com under My Old Designs. I certainly had fun designing with the micros of the day.

FWIW, some of them were extremely expensive. For instance, I used a pair of MC68705P3S in a design in 1980 (which is when they were released). They were US$150 ea, or about $3,000 in todays currency. By 1986 they were $8 ea !!!

Comments

  • Heater.Heater. Posts: 21,230
    Oh yeah, I had a box of 8751's on my desk for a while. 40 pin ceramic package, quartz window to erase the EPROM, gold plated pins. I remember being shocked that they were something like 200 GBP a pop. They never got used, ended up in a junk box in my cellar.
  • Cluso99Cluso99 Posts: 18,066
    edited 2018-03-20 11:06
    In the heyday, we were using 70K+ a year of 68705U3S around 1987 before we switched to Z8671's. Of course the switch was more expensive but we ran out of code space which was only 3.7KB.

    When I looked at the 68705 I also considered the 8751. Motorola was better represented in Oz in those days so this also helped too.

    Hey, I have a pair of Motorola RISC 88100 processor chips (engineering samples from 1988 direct from Motorola Austin TX) in big PGA format.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2018-03-20 12:08
    I have worked with many old micros but the one that I have worked with the longest, now that's something.
    This particular one in question I've worked with for almost 12 years now. Would you believe that! That's unheard of in this day and age :-D
    I even have a couple of original DIP40 packs too although I'm led to believe that these are still available!

  • yetiyeti Posts: 818
    edited 2018-03-20 12:53
    I still miss my SC/MP.
    Maybe that will be a FPGA project someday...
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2018-03-20 13:07
    yeti wrote: »
    I still miss my SC/MP.
    Maybe that will be a FPGA project someday...

    I think you can emulate the actual chip itself (pins and all) with the Prop, and probably faster as well! The SC/MP was very slow and the instruction set very simple. Now there's a project that you can do today instead of someday!

  • But learning FPGA is on my to do list...
    Learning how to spin is not.
  • kwinnkwinn Posts: 8,697
    yeti wrote: »
    But learning FPGA is on my to do list...
    Learning how to spin is not.

    When emulating SC/MP on a Propeller you wouldn't spin, you would s'PASM.
  • yetiyeti Posts: 818
    edited 2018-03-20 13:52
    Translating PASM is included in all Spin compilers I know in the Propeller1 universe and there is no thing like a pure PASM program because it gets launched by Spin.

    So for me talking about Spin includes PASM.
  • Cluso99Cluso99 Posts: 18,066
    edited 2018-03-20 14:03
    I have worked with many old micros but the one that I have worked with the longest, now that's something.
    This particular one in question I've worked with for almost 12 years now. Would you believe that! That's unheard of in this day and age :-D
    I even have a couple of original DIP40 packs too although I'm led to believe that these are still available!
    Next month will be 10 years for me :)

    But I have one up on you Peter. I also have a DIP32 and a DIP24 version too ;) They are collectors items!
  • I remember using the 68705 around that time with its quartz window.

    We also used another micro at that time which had a set of IC pin sockets on top of the ceramic package which broke out the internal address & data bus into a 2716 (or was it 2764) style EPROM pin-out. It was horrendously expensive. Hooking it up to my home made EPROM emulator give me a really useful development system. In-circuit emulators (ICE) cost upwards of £20K at the time.
  • I used the usual 6502, Z80, 6809, etc back in the 80s, and I was fond of the Hitachi H8 family in the 90s. I have a couple of Motorola 68000 processors in the lovely 64-pin DIP package - as far as I know they never made dual in-line chips with more pins than that.
  • Heater.Heater. Posts: 21,230
    I picked up a couple of 68000 from our local electronics store a year or so a go. They were selling them off for 1 Euro a piece so I just had to have some. Lovely, as you say.

    They just need a board to put them in get them running...
  • Cluso99Cluso99 Posts: 18,066
    edited 2018-03-20 17:17
    I designed modems using Rockwell QUIP64 chips. These were wide DIPs with 0.05" pin spacing but the pins were bent into 2 rows 0.1" pitch per side giving a quad row of pins. The pins were quite fragile. Originally we socketed them but the sockets caused more problems than they solved so we ditched them.
  • Cluso99Cluso99 Posts: 18,066
    TonyD wrote: »
    I remember using the 68705 around that time with its quartz window.

    We also used another micro at that time which had a set of IC pin sockets on top of the ceramic package which broke out the internal address & data bus into a 2716 (or was it 2764) style EPROM pin-out. It was horrendously expensive. Hooking it up to my home made EPROM emulator give me a really useful development system. In-circuit emulators (ICE) cost upwards of £20K at the time.
    I recall the chip with a socket on top. Cannot recall who made them.

    Dallas made a battery backed RAM in a 2732 footprint. Quite a thick chip. They gave a 10 year life but we had catastrophic field failures after 5 years. Fortunately my boxes were softloadable remotely over leased lines, so until we replaced the EPROMS were were at least able to reload the code after a power failure.
  • Cluso99Cluso99 Posts: 18,066
    Never did a 68000 design but I did do a 68030 design with multiple serial ports for PCs.
  • Cluso99 wrote: »
    I recall the chip with a socket on top. Cannot recall who made them.

    Dallas made a battery backed RAM in a 2732 footprint. ...
    Thinking back, I'm sure it was a second-source alternative to the Motorola 6805's. Perhaps Mitsubishi or another one of the Japanese chip makers.

    We used the Dallas chips as well but mainly as dev chips to save time by avoiding the EPROM erase cycle.
  • Cluso99Cluso99 Posts: 18,066
    Might have been Hitachi. They also made a 68xxx in a DIP64 @ 0.7" pitch. IIRC it was still 0.6" width, so virtually same footprint as a DIP40.
  • Heater.Heater. Posts: 21,230
    Cluso,
    I designed modems using Rockwell QUIP64 chips.
    What, you mean like these: https://en.wikichip.org/wiki/rockwell_international/pps-4

    I was offered a whole box of those when Marconi Radar dev labs was clearing out old junk in 1981 or so. I could not find any data books about the place to go with them at the time so I declined.

  • Cluso99Cluso99 Posts: 18,066
    Yes although those are much smaller ~Quip40.
    They were Rockwell DSP chips.
  • We used those QUIP 64 chips back in the 80's. They were a pain as they were always socketed. The assembly girls almost always bent a pin upon insertion, and we spent so much time in final test fixing them.
  • Cluso99Cluso99 Posts: 18,066
    Once we stopped using sockets most of our production and field problems disappeared. Add to that, we saved the expense of the pair of sockets per board which were not cheap.
  • I used to use those battery-backed-RAM 2732-EPROM-equivalent chips back in the 80s. The company I worked for was using Commodore PET main boards in industrial applications and used to replace the on-board ROMs that held the O/S (if you can call it that) with 2732s containing modified code. We also had a pager board that could switch one of eight EPROMS into one socket on the PET board and we used to store a 32kb program in there for downloading to the PET RAM at start up.

    At least I think it was 2732s but I also remember using 2532s. I forget what the difference was between the 2532 and 2732 - but they were very similar - both 4kb x 8bit EPROMS. There was also a 2716 which was a 2kb instead of 4kb version which I think the PET used in one of its ROM slots.

    I remember the RAM replacement modules had the same footprint as a 2732 but were much taller - about the same height as they were wide. I think they were probably made by using a small plastic box containing the RAM and battery with a 24-pin DIL socket base and then potting the whole thing. I vaguely remember they had yellow labels stuck on them by the manufacturer.

    Of course, the EPROMs didn't have a read/write pin (being read only) so the write enable pin on the battery-backed RAM replacements was something that stuck up vertically at one end of the 'chip' and you had to fit a separate jumper wire to those pins from the 6502 processor's R/W line when you wanted to write to the modules.

    Often we used the battery backed modules to develop a program and then copied the contents to EPROMs using an EPROM reader/burner before deploying out in the field. But sometimes the RAM modules got left in customer's units when we weren't sure whether more changes might be required to their programs. The modules used to retain their contents for several years and were reasonably reliable.
  • Cluso99Cluso99 Posts: 18,066
    edited 2018-03-23 05:50
    Yes, I remember the yellow label. Was it one or two pins at the end? My memory is suggesting two, but the second pin would only be ground if it existed. They were very expensive initially, so I only had one. These were before the Dallas (and a second source?) came out which were nowhere as thick.

    I wrote a bootloader that I use to store in EPROM, so I just downloaded code into RAM and jumped to it after the download. Saved cheapskate of development time.
  • I'm not sure about the number of pins at the end either. I also think it was more than one - possibly it was three pins? I have a (very) vague memory of attaching the 'read/write wire' to the 'centre pin' but that could just be the pin closest to the centre of the chip, rather than the centre one of three pins. I looked through Google images to see if I could find any photographs of those early devices, but I was unsuccessful.

    A colleague at the place I worked wrote some small upload/download routines in machine code that saved/restored the program in Commodore PET RAM to our paged Battery-backed RAM / EPROM board. The routines were small enough to cram into a small unused area in a copy of one of the PET's ROMs. You called the upload routine with a SYS command. The download routine was automatically hooked into some code that ran in the standard ROM at start up. My colleague invented some cool acronyms for his special routines: the uploader was named PULSE for 'program upload save and end' and the downloader was named POLAR for 'power on load and run'. The combination of the two was therefore known as the 'Pulsar ROM'. We fitted this pulsar ROM into the PET, replacing one of its standard ROMS and when you switched on the PET the polar routine would detect if a pager board was present and if so would automatically download and boot whatever program was stored in it.
  • Cluso99 wrote: »
    TonyD wrote: »
    I remember using the 68705 around that time with its quartz window.

    We also used another micro at that time which had a set of IC pin sockets on top of the ceramic package which broke out the internal address & data bus into a 2716 (or was it 2764) style EPROM pin-out. It was horrendously expensive. Hooking it up to my home made EPROM emulator give me a really useful development system. In-circuit emulators (ICE) cost upwards of £20K at the time.
    I recall the chip with a socket on top. Cannot recall who made them.

    Dallas made a battery backed RAM in a 2732 footprint. Quite a thick chip. They gave a 10 year life but we had catastrophic field failures after 5 years. Fortunately my boxes were softloadable remotely over leased lines, so until we replaced the EPROMS were were at least able to reload the code after a power failure.

    Hello!
    Back when I was based in Westchester County, I was just getting into the wonders of applying computers to do the near impossible, and promptly started reading nearly everything the local library system had on that subject.

    One library had on its shelves a dandy book on the Z-80, and on its pages it showed a development device for the Z8, and it wore a 2716 (or a 2764 depending on the design cycle) on its back.
Sign In or Register to comment.