Hello All!
For early pre-micro CPU's i.e., before the Intel 8080 suggest you look at the Intel 4004 and 4040 (assuming that one is listed on any site.
Multi 4004's made up the basic 8080 allthough other logic had to be added. This around 1973 - 4 - sorry my memory is getting a bit dull these days and yes was involved with them and the very early 808, 6502, 68000 (hmm may be not that one) etc.
Vincent
Hello Ahem!
Sunday February 27 2022 16:38, you wrote to me:
I was referring to the internals not the physical and I was told by Intel in the early 70's of some of the details but have not kept the document's - at least to my knowledge.
I did play with a very early micro based board around that time and it consisted of multi 4004's and some other elements, seems to recall a small board say of less than a foot square, possibly a lot less - sorry to long ago and I had a lot of others systems to look after or was playing with.
My first board was (going by my poor memory) a 6502 followed around the same time of a proto 8080 with a mind bending Ram size of 4K but I could connect it to a teletype 33 for I/O a later one had a link to a tape drive - that's a Philips device.
About the same time as I had a PDP 8 and PDP 11 in my spare bedroom. [ Yes, hard work to get them up the stairs ]
Vincent
Those are both microprocessors - the 4004 is usually considered the first microprocessor. For a construction project for a home computer based on the 4040 Practical Electronics published the CHAMP project in 1977 - I did once find reprints online but not this time round.
Er no, the 8080 was not multiple 4004s or even 4040s it was an 8 bit CPU spread over a three chip set, CPU, Bus interface and Clock generator, it also required 12V, 5V and -5V supplies. the Z80 won out over the 8080 more for it's single chip 5V only convenience than the extra instructions that hardly anyone used.
On Sun, 27 Feb 2022 16:38:27 +0000, Ahem A Rivet's Shot snipped-for-privacy@eircom.net declaimed the following:
The 8228 "system controller" chip wasn't a required chip -- I've /seen/ systems built using five 8212 8-bit tri-state buffers (heck, I still have an Intel $20 evaluation kit that shipped with 8080A, clock generator, the said five 8212s, some memory chips -- 8x 2102 [1K by 1bit] static RAM, and
1KB worth of UV ROM [1702 256 by 8bit, I recall]; I subsequently sourced an 8228 and some other memory chips... Never did finish the Wirewrap card for it -- nor test the "monitor" program I'd written to handle a calculator keypad and 2-digit LED display.).Two 8212s used to latch/buffer the 16 address lines. One 8212 used to latch/buffer the data OUT One 8212 used to latch the data IN One 8212 (as I recall) used to latch/buffer the control signals (after they'd gone through some AND/OR/etc. gates to combine things like "IO/MEM" & "RD/WR" -- Intel influenced chips support separate I/O "ports" rather than relying upon memory-mapped I/O as with 6800/6502 systems). The 8228 took care of the bottom three 8212s, but did not handle the address lines.
Thank you, system controller not bus interface - it's been a while.
I'm guessing that way was cheaper in chip cost and more expensive in board area, hole count etc - the kind of things that only matter in production.
I was very definitely one of your "hardly anyone".
David
I tried to use them but not sure I found them very convenient.
It depends what extra instructions.
DJNZ and the JR family got tons of use.
On Sun, 27 Feb 2022 20:10:03 +0000, Ahem A Rivet's Shot snipped-for-privacy@eircom.net declaimed the following:
The "system" I'd seen in use with the mass of 8212s, if you'd believe it, was the floppy disk controller board used on a CRDS LSI-11. The two computers were kept in a narrow room which tended to overheat, so in summer/fall the covers were often slid back to let a fan blow into the boxes. They were used for (UCSD) Pascal, and for the OS course (can't risk letting us barbarians into the campus mainframe for that).
The "evaluation kit" was just the reference manual, and a bunch of "cosmetic defect" chips in anti-static foam. A number of these were bought by the students in the digital electronics III course (at the time, CompSci was a MATH major, exposure to digital stuff was from the PHYSICS department <G>). However, as a two-session per week, 11 week course, we never really got far in wiring up candidate systems (S-100 covered with wire-wrap sockets, anyone?).
Wow, but yeah controller boards often seemed to get odd chip choices that probably make good sense when you know the whole story.
The Newbrain was first prototyped on double eurocard wire-wrap boards, I got very slick with a wire-wrap gun on that job.
On Mon, 28 Feb 2022 17:23:52 +0000, Ahem A Rivet's Shot snipped-for-privacy@eircom.net declaimed the following:
<envy> All I had (and should have in a box of chips somewhere) was the double-ended (wrap/unwrap) "pencil".
My own wire-wrap tool was one of those golden barreled things with a stripper (wire) in the middle. The Gardner-Denver tools Newbury labs sprung for were *expensive*, but pretty much essential if you're going to get a board built and tested in a day - along with endless tubes of pre-cut, pre-stripped wire in several lengths (a very exoensive
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.