VCF Museum - Microcomputer slots, S-100 to Apple II to IBM PC, Lou Eggebrecht
This is the first in what I expect to be a series of posts that highlight subject I/We are thinking of at the museum. In the museum, one of our talking points- a "sub tour" in the museum, highlights concepts of interconnects, from back planes to buses. Bus slots are best represented currently in the micro era exhibits on S-100, Apple II, and the PC. One idea that is probably coming is to have a whole slew of labeled add-in boards on the wall to show how significant the ability and the practice of upgrading a microcomputer was. Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too. I'd like to share a bit of lore on the IBM PC bus, as this, I find, is lesser known in the hobby. The father of the IBM PC ISA bus is Lou Eggebrecht, and its a great story. He wrote the book, literally, on the IBM PC ISA Bus. https://archive.org/details/interfacingtoibm00egge I just bought a copy. We'll have this in the museum. After the PC was launched, a situation happened between him and IBM and they "parted ways". A short interpretation might go along the lines that IBM perceived Mr. Eggebrecht was capitalizing too much personally on the ISA bus. My reference is "Blue Magic - The People, Power and Politics Behind the IBM Personal Computer" by James Chposky and Ted Leonsis. I don't know his whole career after IBM. I do know he went on to be an influential person at Commodore, ICS and Broadcom. Here's a period piece on him at Commodore (https://amr.abime.net/review_51617) He seemed to follow along with moves of Mr. Hock Tan. Hock Tan is the CEO that presided at Commodore at the end of line, selling off the assets I think. Hock Tan then came to ICS (Integrated Circuit Systems), a Valley Forge-based fabless mixed signal semiconductor designer with products revolving around phase-locked loops. Of interest to most of our hobby--- ICS was started by Ex-MOS process engineer Ed Arnold in the late 1970s. ICS had a number of ex-Commodore personnel on staff. Mr. Eggebrecht I believe was on the board of ICS for while, and this is where I had a chance to meet him in the late 1990s. At the time I only knew he had some involvement with the IBM PC. He was very likeable; humble, warm, and gracious. Mr. Eggebrecht was ideal to have on the board at ICS - who's chief products were the timing generators for the Intel processors and chipsets at the time. Hock Tan went on to IDT (as did I) when he sold ICS to IDT and then to Avago (I did not). Hock Tan orchestrated the huge merger of Avago and Broadcom. I mention this, because, not surprisingly, Lou Eggebrecht served on the board of Broadcom for some time too. He would be a great speaker at a VCF! :-O I hope his oral history gets taken too.
I think an exhibit on computer bus "slots" should at least mention that computers are still designed that way, and show examples of ISA 8 bit, ISA 16 bit, and then the successors to those (PCI etc.). This is just to show that the story of "slots" isn't over, and the interfaces continue to evolve. Bill Dudley This email is free of malware because I run Linux. On Mon, May 22, 2023 at 11:58 AM Douglas Crawford via vcf-midatlantic < vcf-midatlantic@lists.vcfed.org> wrote:
This is the first in what I expect to be a series of posts that highlight subject I/We are thinking of at the museum.
In the museum, one of our talking points- a "sub tour" in the museum, highlights concepts of interconnects, from back planes to buses. Bus slots are best represented currently in the micro era exhibits on S-100, Apple II, and the PC. One idea that is probably coming is to have a whole slew of labeled add-in boards on the wall to show how significant the ability and the practice of upgrading a microcomputer was.
Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too.
I'd like to share a bit of lore on the IBM PC bus, as this, I find, is lesser known in the hobby.
The father of the IBM PC ISA bus is Lou Eggebrecht, and its a great story. He wrote the book, literally, on the IBM PC ISA Bus. https://archive.org/details/interfacingtoibm00egge I just bought a copy. We'll have this in the museum.
After the PC was launched, a situation happened between him and IBM and they "parted ways". A short interpretation might go along the lines that IBM perceived Mr. Eggebrecht was capitalizing too much personally on the ISA bus. My reference is "Blue Magic - The People, Power and Politics Behind the IBM Personal Computer" by James Chposky and Ted Leonsis.
I don't know his whole career after IBM. I do know he went on to be an influential person at Commodore, ICS and Broadcom. Here's a period piece on him at Commodore (https://amr.abime.net/review_51617) He seemed to follow along with moves of Mr. Hock Tan. Hock Tan is the CEO that presided at Commodore at the end of line, selling off the assets I think. Hock Tan then came to ICS (Integrated Circuit Systems), a Valley Forge-based fabless mixed signal semiconductor designer with products revolving around phase-locked loops. Of interest to most of our hobby--- ICS was started by Ex-MOS process engineer Ed Arnold in the late 1970s. ICS had a number of ex-Commodore personnel on staff.
Mr. Eggebrecht I believe was on the board of ICS for while, and this is where I had a chance to meet him in the late 1990s. At the time I only knew he had some involvement with the IBM PC. He was very likeable; humble, warm, and gracious. Mr. Eggebrecht was ideal to have on the board at ICS - who's chief products were the timing generators for the Intel processors and chipsets at the time.
Hock Tan went on to IDT (as did I) when he sold ICS to IDT and then to Avago (I did not). Hock Tan orchestrated the huge merger of Avago and Broadcom. I mention this, because, not surprisingly, Lou Eggebrecht served on the board of Broadcom for some time too.
He would be a great speaker at a VCF! :-O I hope his oral history gets taken too.
Hello! Doug I have a copy here of that same book he wrote on the IBM PC backplane. It seems that every time I come up with a great idea to build the ultimate something else that would be used inside one, it seems that the PC wouldn't be the right platform. Oddly enough the right platform is the Apple, because prior to my switching sides, I was extremely good at programming on it, and even understanding the logic behind how the system did things. (Ian L will confirm that.) For the purposes of this messaging platform I won't go into detail, but suffice to say it has to do with my technology interests. Incidentally the raptor has left the grounds, I think he's in the pine hills. ----- Gregg C Levine gregg.drwho8@gmail.com "This signature fought the Time Wars, time and again." On Mon, May 22, 2023 at 11:58 AM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
This is the first in what I expect to be a series of posts that highlight subject I/We are thinking of at the museum.
In the museum, one of our talking points- a "sub tour" in the museum, highlights concepts of interconnects, from back planes to buses. Bus slots are best represented currently in the micro era exhibits on S-100, Apple II, and the PC. One idea that is probably coming is to have a whole slew of labeled add-in boards on the wall to show how significant the ability and the practice of upgrading a microcomputer was.
Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too.
I'd like to share a bit of lore on the IBM PC bus, as this, I find, is lesser known in the hobby.
The father of the IBM PC ISA bus is Lou Eggebrecht, and its a great story. He wrote the book, literally, on the IBM PC ISA Bus. https://archive.org/details/interfacingtoibm00egge I just bought a copy. We'll have this in the museum.
After the PC was launched, a situation happened between him and IBM and they "parted ways". A short interpretation might go along the lines that IBM perceived Mr. Eggebrecht was capitalizing too much personally on the ISA bus. My reference is "Blue Magic - The People, Power and Politics Behind the IBM Personal Computer" by James Chposky and Ted Leonsis.
I don't know his whole career after IBM. I do know he went on to be an influential person at Commodore, ICS and Broadcom. Here's a period piece on him at Commodore (https://amr.abime.net/review_51617) He seemed to follow along with moves of Mr. Hock Tan. Hock Tan is the CEO that presided at Commodore at the end of line, selling off the assets I think. Hock Tan then came to ICS (Integrated Circuit Systems), a Valley Forge-based fabless mixed signal semiconductor designer with products revolving around phase-locked loops. Of interest to most of our hobby--- ICS was started by Ex-MOS process engineer Ed Arnold in the late 1970s. ICS had a number of ex-Commodore personnel on staff.
Mr. Eggebrecht I believe was on the board of ICS for while, and this is where I had a chance to meet him in the late 1990s. At the time I only knew he had some involvement with the IBM PC. He was very likeable; humble, warm, and gracious. Mr. Eggebrecht was ideal to have on the board at ICS - who's chief products were the timing generators for the Intel processors and chipsets at the time.
Hock Tan went on to IDT (as did I) when he sold ICS to IDT and then to Avago (I did not). Hock Tan orchestrated the huge merger of Avago and Broadcom. I mention this, because, not surprisingly, Lou Eggebrecht served on the board of Broadcom for some time too.
He would be a great speaker at a VCF! :-O I hope his oral history gets taken too.
On Mon, May 22, 2023 at 11:58 AM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too.
Depends on the specific minicomputer, and to an extent, the era, but starting in 1970, DEC went from random-wired logic to a bus system with the Unibus in the PDP-11/20. They started with backplane-scale "system modules" that were bridged together such that the Unibus was at the start and end of a backplane block (and the middle could still be random-wired logic) but that quickly (1972) evolved to one slot after another just being "Unibus" (there are some power supply pin variations but the basic signals are the same from slot to slot). The CPU goes at the top but I/O cards follow a set of priority rules as to the order that follows. From the top side, the slots all look the same, like an embedded PC backplane that is 100% ISA slots with the CPU on an ISA card. Following from the PDP-11 Unibus is the PDP-8 Omnibus (a bus of busses) where all the slots are exactly identical. Then the Qbus came along and though there are a couple different wiring schemes (Q18, Q22, CD-slots, straight-down vs serpentine) but it's not a free-for-all - there are really only a handful of arrangements one encounters. There are details that one might recognize from microcomputer "slots" and there are some differences. In broad terms, with minicomputers, it's not unusual to see the CPU and memory and peripherals either on the same bus structure, or sometimes there are dedicated memory slots and dedicated peripheral slots that may or may not use the same connectors. It's not as straightforward as, say, the ISA bus or the Apple II slot scheme. The closest examples of structural resemblances that come to mine are, for one, the Omnibus and the S-100 bus. This is just one vendor. There is something visually similar with Data General Nova systems but I'm not experienced enough to comment on them. There's more but I wanted to at least throw some words out there about some of the more popular minis. -ethan
Data General's full-size Novas and Eclipse series machines use two of the same 100 pin connectors found on S-100. Many of the slots are keyed to accept only certain types of cards, but they're more or less wired alike. The CPU for example can only be inserted in the bottom-most slot on a 1200, with either memory or a CPU feature expansion card directly above that. An 800 splits the CPU cards between the two bottom slots. I can't speak to the backplane design of the Eclipse series or the later Novas, but it wouldn't surprise me if they're similarly keyed. -Alexander 'Z' Pierson>This is just one vendor. There is something visually similar with Data General Nova systems but I'm not experienced enough to comment on them.
Thanks that a great short overview of the DEC side, thanks. Did the minicomputers gather much 3rd party support for cards on their buses? I'm going to guess some, but not near as much as what happened in microcomputers, perhaps simply due to installed base/market size being relatively larger on the microcomputer platforms. On 5/22/2023 8:00 PM, Ethan Dicks via vcf-midatlantic wrote:
On Mon, May 22, 2023 at 11:58 AM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too.
Depends on the specific minicomputer, and to an extent, the era, but starting in 1970, DEC went from random-wired logic to a bus system with the Unibus in the PDP-11/20. They started with backplane-scale "system modules" that were bridged together such that the Unibus was at the start and end of a backplane block (and the middle could still be random-wired logic) but that quickly (1972) evolved to one slot after another just being "Unibus" (there are some power supply pin variations but the basic signals are the same from slot to slot). The CPU goes at the top but I/O cards follow a set of priority rules as to the order that follows. From the top side, the slots all look the same, like an embedded PC backplane that is 100% ISA slots with the CPU on an ISA card.
Following from the PDP-11 Unibus is the PDP-8 Omnibus (a bus of busses) where all the slots are exactly identical.
Then the Qbus came along and though there are a couple different wiring schemes (Q18, Q22, CD-slots, straight-down vs serpentine) but it's not a free-for-all - there are really only a handful of arrangements one encounters.
There are details that one might recognize from microcomputer "slots" and there are some differences. In broad terms, with minicomputers, it's not unusual to see the CPU and memory and peripherals either on the same bus structure, or sometimes there are dedicated memory slots and dedicated peripheral slots that may or may not use the same connectors. It's not as straightforward as, say, the ISA bus or the Apple II slot scheme. The closest examples of structural resemblances that come to mine are, for one, the Omnibus and the S-100 bus.
This is just one vendor. There is something visually similar with Data General Nova systems but I'm not experienced enough to comment on them.
There's more but I wanted to at least throw some words out there about some of the more popular minis.
-ethan
Lots and lots! Many of my (personal) VAXen and PDP-11s have cards from Emulex, Dilog, etc. One of my PDP-8s has a card from Data Systems Design. That's just some DECworld examples, but there were tons of third-party peripherals for most minicomputers. -Dave On 5/22/23 22:09, Douglas Crawford via vcf-midatlantic wrote:
Thanks that a great short overview of the DEC side, thanks. Did the minicomputers gather much 3rd party support for cards on their buses? I'm going to guess some, but not near as much as what happened in microcomputers, perhaps simply due to installed base/market size being relatively larger on the microcomputer platforms.
On 5/22/2023 8:00 PM, Ethan Dicks via vcf-midatlantic wrote:
On Mon, May 22, 2023 at 11:58 AM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
Slots were important to make these machines versatile and created entire sub industries around each machine that had slots. I'm less versed in the slot systems of the minicomputers; I imagine a certain amount of this dynamic occurred there too.
Depends on the specific minicomputer, and to an extent, the era, but starting in 1970, DEC went from random-wired logic to a bus system with the Unibus in the PDP-11/20. They started with backplane-scale "system modules" that were bridged together such that the Unibus was at the start and end of a backplane block (and the middle could still be random-wired logic) but that quickly (1972) evolved to one slot after another just being "Unibus" (there are some power supply pin variations but the basic signals are the same from slot to slot). The CPU goes at the top but I/O cards follow a set of priority rules as to the order that follows. From the top side, the slots all look the same, like an embedded PC backplane that is 100% ISA slots with the CPU on an ISA card.
Following from the PDP-11 Unibus is the PDP-8 Omnibus (a bus of busses) where all the slots are exactly identical.
Then the Qbus came along and though there are a couple different wiring schemes (Q18, Q22, CD-slots, straight-down vs serpentine) but it's not a free-for-all - there are really only a handful of arrangements one encounters.
There are details that one might recognize from microcomputer "slots" and there are some differences. In broad terms, with minicomputers, it's not unusual to see the CPU and memory and peripherals either on the same bus structure, or sometimes there are dedicated memory slots and dedicated peripheral slots that may or may not use the same connectors. It's not as straightforward as, say, the ISA bus or the Apple II slot scheme. The closest examples of structural resemblances that come to mine are, for one, the Omnibus and the S-100 bus.
This is just one vendor. There is something visually similar with Data General Nova systems but I'm not experienced enough to comment on them.
There's more but I wanted to at least throw some words out there about some of the more popular minis.
-ethan
-- Dave McGuire, AK4HZ New Kensington, PA
On Mon, May 22, 2023 at 10:09 PM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
Thanks that a great short overview of the DEC side, thanks.
You bet!
Did the minicomputers gather much 3rd party support for cards on their buses? I'm going to guess some, but not near as much as what happened in microcomputers, perhaps simply due to installed base/market size being relatively larger on the microcomputer platforms.
As Dave said, lots and lots! My employer for most of 1984-1994 _made_ Qbus, Unibus, and VAXBI boards (COMBOARDs - m68k + USART + RAM to run IBM protocol emulators). Emulex, Dilog, Systems Industries, Western Peripherals, Peritek... so many 3rd-party board vendors for DEC boxes. Memory, and disk controllers, and tape controllers were made in abundance by 3rd parties. CPU boards, not so much. Even in Heathkit boxes where you were soldering your own serial/parallel cards and the backplane, it took a DEC-made CPU board. We used to buy "DEC Kit" bus chip packages directly from Digital. Outside of the VAXBI, they were very generous with information and OEM parts. -ethan
So interesting! On 5/24/2023 2:02 PM, Ethan Dicks via vcf-midatlantic wrote:
On Mon, May 22, 2023 at 10:09 PM Douglas Crawford via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
Thanks that a great short overview of the DEC side, thanks.
You bet!
Did the minicomputers gather much 3rd party support for cards on their buses? I'm going to guess some, but not near as much as what happened in microcomputers, perhaps simply due to installed base/market size being relatively larger on the microcomputer platforms.
As Dave said, lots and lots! My employer for most of 1984-1994 _made_ Qbus, Unibus, and VAXBI boards (COMBOARDs - m68k + USART + RAM to run IBM protocol emulators). Emulex, Dilog, Systems Industries, Western Peripherals, Peritek... so many 3rd-party board vendors for DEC boxes. Memory, and disk controllers, and tape controllers were made in abundance by 3rd parties. CPU boards, not so much. Even in Heathkit boxes where you were soldering your own serial/parallel cards and the backplane, it took a DEC-made CPU board.
We used to buy "DEC Kit" bus chip packages directly from Digital. Outside of the VAXBI, they were very generous with information and OEM parts.
-ethan
participants (6)
-
Alexander Pierson -
Dave McGuire -
Douglas Crawford -
Ethan Dicks -
Gregg Levine -
William Dudley